diff --git a/.gitignore b/.gitignore
index 7540a95..df50c22 100644
--- a/.gitignore
+++ b/.gitignore
@@ -4,3 +4,12 @@ memristor_utils_g_extradim.py
 memristor_utils_g_posneg.py
 *.log
 *.pdf
+/main.py
+*.h5
+*.pkl
+*.csv
+*.zip
+*.mat
+*.pickle
+*.xlsx
+*.~lock*
diff --git a/LICENSE b/LICENSE
new file mode 100644
index 0000000..f288702
--- /dev/null
+++ b/LICENSE
@@ -0,0 +1,674 @@
+                    GNU GENERAL PUBLIC LICENSE
+                       Version 3, 29 June 2007
+
+ Copyright (C) 2007 Free Software Foundation, Inc. <https://fsf.org/>
+ Everyone is permitted to copy and distribute verbatim copies
+ of this license document, but changing it is not allowed.
+
+                            Preamble
+
+  The GNU General Public License is a free, copyleft license for
+software and other kinds of works.
+
+  The licenses for most software and other practical works are designed
+to take away your freedom to share and change the works.  By contrast,
+the GNU General Public License is intended to guarantee your freedom to
+share and change all versions of a program--to make sure it remains free
+software for all its users.  We, the Free Software Foundation, use the
+GNU General Public License for most of our software; it applies also to
+any other work released this way by its authors.  You can apply it to
+your programs, too.
+
+  When we speak of free software, we are referring to freedom, not
+price.  Our General Public Licenses are designed to make sure that you
+have the freedom to distribute copies of free software (and charge for
+them if you wish), that you receive source code or can get it if you
+want it, that you can change the software or use pieces of it in new
+free programs, and that you know you can do these things.
+
+  To protect your rights, we need to prevent others from denying you
+these rights or asking you to surrender the rights.  Therefore, you have
+certain responsibilities if you distribute copies of the software, or if
+you modify it: responsibilities to respect the freedom of others.
+
+  For example, if you distribute copies of such a program, whether
+gratis or for a fee, you must pass on to the recipients the same
+freedoms that you received.  You must make sure that they, too, receive
+or can get the source code.  And you must show them these terms so they
+know their rights.
+
+  Developers that use the GNU GPL protect your rights with two steps:
+(1) assert copyright on the software, and (2) offer you this License
+giving you legal permission to copy, distribute and/or modify it.
+
+  For the developers' and authors' protection, the GPL clearly explains
+that there is no warranty for this free software.  For both users' and
+authors' sake, the GPL requires that modified versions be marked as
+changed, so that their problems will not be attributed erroneously to
+authors of previous versions.
+
+  Some devices are designed to deny users access to install or run
+modified versions of the software inside them, although the manufacturer
+can do so.  This is fundamentally incompatible with the aim of
+protecting users' freedom to change the software.  The systematic
+pattern of such abuse occurs in the area of products for individuals to
+use, which is precisely where it is most unacceptable.  Therefore, we
+have designed this version of the GPL to prohibit the practice for those
+products.  If such problems arise substantially in other domains, we
+stand ready to extend this provision to those domains in future versions
+of the GPL, as needed to protect the freedom of users.
+
+  Finally, every program is threatened constantly by software patents.
+States should not allow patents to restrict development and use of
+software on general-purpose computers, but in those that do, we wish to
+avoid the special danger that patents applied to a free program could
+make it effectively proprietary.  To prevent this, the GPL assures that
+patents cannot be used to render the program non-free.
+
+  The precise terms and conditions for copying, distribution and
+modification follow.
+
+                       TERMS AND CONDITIONS
+
+  0. Definitions.
+
+  "This License" refers to version 3 of the GNU General Public License.
+
+  "Copyright" also means copyright-like laws that apply to other kinds of
+works, such as semiconductor masks.
+
+  "The Program" refers to any copyrightable work licensed under this
+License.  Each licensee is addressed as "you".  "Licensees" and
+"recipients" may be individuals or organizations.
+
+  To "modify" a work means to copy from or adapt all or part of the work
+in a fashion requiring copyright permission, other than the making of an
+exact copy.  The resulting work is called a "modified version" of the
+earlier work or a work "based on" the earlier work.
+
+  A "covered work" means either the unmodified Program or a work based
+on the Program.
+
+  To "propagate" a work means to do anything with it that, without
+permission, would make you directly or secondarily liable for
+infringement under applicable copyright law, except executing it on a
+computer or modifying a private copy.  Propagation includes copying,
+distribution (with or without modification), making available to the
+public, and in some countries other activities as well.
+
+  To "convey" a work means any kind of propagation that enables other
+parties to make or receive copies.  Mere interaction with a user through
+a computer network, with no transfer of a copy, is not conveying.
+
+  An interactive user interface displays "Appropriate Legal Notices"
+to the extent that it includes a convenient and prominently visible
+feature that (1) displays an appropriate copyright notice, and (2)
+tells the user that there is no warranty for the work (except to the
+extent that warranties are provided), that licensees may convey the
+work under this License, and how to view a copy of this License.  If
+the interface presents a list of user commands or options, such as a
+menu, a prominent item in the list meets this criterion.
+
+  1. Source Code.
+
+  The "source code" for a work means the preferred form of the work
+for making modifications to it.  "Object code" means any non-source
+form of a work.
+
+  A "Standard Interface" means an interface that either is an official
+standard defined by a recognized standards body, or, in the case of
+interfaces specified for a particular programming language, one that
+is widely used among developers working in that language.
+
+  The "System Libraries" of an executable work include anything, other
+than the work as a whole, that (a) is included in the normal form of
+packaging a Major Component, but which is not part of that Major
+Component, and (b) serves only to enable use of the work with that
+Major Component, or to implement a Standard Interface for which an
+implementation is available to the public in source code form.  A
+"Major Component", in this context, means a major essential component
+(kernel, window system, and so on) of the specific operating system
+(if any) on which the executable work runs, or a compiler used to
+produce the work, or an object code interpreter used to run it.
+
+  The "Corresponding Source" for a work in object code form means all
+the source code needed to generate, install, and (for an executable
+work) run the object code and to modify the work, including scripts to
+control those activities.  However, it does not include the work's
+System Libraries, or general-purpose tools or generally available free
+programs which are used unmodified in performing those activities but
+which are not part of the work.  For example, Corresponding Source
+includes interface definition files associated with source files for
+the work, and the source code for shared libraries and dynamically
+linked subprograms that the work is specifically designed to require,
+such as by intimate data communication or control flow between those
+subprograms and other parts of the work.
+
+  The Corresponding Source need not include anything that users
+can regenerate automatically from other parts of the Corresponding
+Source.
+
+  The Corresponding Source for a work in source code form is that
+same work.
+
+  2. Basic Permissions.
+
+  All rights granted under this License are granted for the term of
+copyright on the Program, and are irrevocable provided the stated
+conditions are met.  This License explicitly affirms your unlimited
+permission to run the unmodified Program.  The output from running a
+covered work is covered by this License only if the output, given its
+content, constitutes a covered work.  This License acknowledges your
+rights of fair use or other equivalent, as provided by copyright law.
+
+  You may make, run and propagate covered works that you do not
+convey, without conditions so long as your license otherwise remains
+in force.  You may convey covered works to others for the sole purpose
+of having them make modifications exclusively for you, or provide you
+with facilities for running those works, provided that you comply with
+the terms of this License in conveying all material for which you do
+not control copyright.  Those thus making or running the covered works
+for you must do so exclusively on your behalf, under your direction
+and control, on terms that prohibit them from making any copies of
+your copyrighted material outside their relationship with you.
+
+  Conveying under any other circumstances is permitted solely under
+the conditions stated below.  Sublicensing is not allowed; section 10
+makes it unnecessary.
+
+  3. Protecting Users' Legal Rights From Anti-Circumvention Law.
+
+  No covered work shall be deemed part of an effective technological
+measure under any applicable law fulfilling obligations under article
+11 of the WIPO copyright treaty adopted on 20 December 1996, or
+similar laws prohibiting or restricting circumvention of such
+measures.
+
+  When you convey a covered work, you waive any legal power to forbid
+circumvention of technological measures to the extent such circumvention
+is effected by exercising rights under this License with respect to
+the covered work, and you disclaim any intention to limit operation or
+modification of the work as a means of enforcing, against the work's
+users, your or third parties' legal rights to forbid circumvention of
+technological measures.
+
+  4. Conveying Verbatim Copies.
+
+  You may convey verbatim copies of the Program's source code as you
+receive it, in any medium, provided that you conspicuously and
+appropriately publish on each copy an appropriate copyright notice;
+keep intact all notices stating that this License and any
+non-permissive terms added in accord with section 7 apply to the code;
+keep intact all notices of the absence of any warranty; and give all
+recipients a copy of this License along with the Program.
+
+  You may charge any price or no price for each copy that you convey,
+and you may offer support or warranty protection for a fee.
+
+  5. Conveying Modified Source Versions.
+
+  You may convey a work based on the Program, or the modifications to
+produce it from the Program, in the form of source code under the
+terms of section 4, provided that you also meet all of these conditions:
+
+    a) The work must carry prominent notices stating that you modified
+    it, and giving a relevant date.
+
+    b) The work must carry prominent notices stating that it is
+    released under this License and any conditions added under section
+    7.  This requirement modifies the requirement in section 4 to
+    "keep intact all notices".
+
+    c) You must license the entire work, as a whole, under this
+    License to anyone who comes into possession of a copy.  This
+    License will therefore apply, along with any applicable section 7
+    additional terms, to the whole of the work, and all its parts,
+    regardless of how they are packaged.  This License gives no
+    permission to license the work in any other way, but it does not
+    invalidate such permission if you have separately received it.
+
+    d) If the work has interactive user interfaces, each must display
+    Appropriate Legal Notices; however, if the Program has interactive
+    interfaces that do not display Appropriate Legal Notices, your
+    work need not make them do so.
+
+  A compilation of a covered work with other separate and independent
+works, which are not by their nature extensions of the covered work,
+and which are not combined with it such as to form a larger program,
+in or on a volume of a storage or distribution medium, is called an
+"aggregate" if the compilation and its resulting copyright are not
+used to limit the access or legal rights of the compilation's users
+beyond what the individual works permit.  Inclusion of a covered work
+in an aggregate does not cause this License to apply to the other
+parts of the aggregate.
+
+  6. Conveying Non-Source Forms.
+
+  You may convey a covered work in object code form under the terms
+of sections 4 and 5, provided that you also convey the
+machine-readable Corresponding Source under the terms of this License,
+in one of these ways:
+
+    a) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by the
+    Corresponding Source fixed on a durable physical medium
+    customarily used for software interchange.
+
+    b) Convey the object code in, or embodied in, a physical product
+    (including a physical distribution medium), accompanied by a
+    written offer, valid for at least three years and valid for as
+    long as you offer spare parts or customer support for that product
+    model, to give anyone who possesses the object code either (1) a
+    copy of the Corresponding Source for all the software in the
+    product that is covered by this License, on a durable physical
+    medium customarily used for software interchange, for a price no
+    more than your reasonable cost of physically performing this
+    conveying of source, or (2) access to copy the
+    Corresponding Source from a network server at no charge.
+
+    c) Convey individual copies of the object code with a copy of the
+    written offer to provide the Corresponding Source.  This
+    alternative is allowed only occasionally and noncommercially, and
+    only if you received the object code with such an offer, in accord
+    with subsection 6b.
+
+    d) Convey the object code by offering access from a designated
+    place (gratis or for a charge), and offer equivalent access to the
+    Corresponding Source in the same way through the same place at no
+    further charge.  You need not require recipients to copy the
+    Corresponding Source along with the object code.  If the place to
+    copy the object code is a network server, the Corresponding Source
+    may be on a different server (operated by you or a third party)
+    that supports equivalent copying facilities, provided you maintain
+    clear directions next to the object code saying where to find the
+    Corresponding Source.  Regardless of what server hosts the
+    Corresponding Source, you remain obligated to ensure that it is
+    available for as long as needed to satisfy these requirements.
+
+    e) Convey the object code using peer-to-peer transmission, provided
+    you inform other peers where the object code and Corresponding
+    Source of the work are being offered to the general public at no
+    charge under subsection 6d.
+
+  A separable portion of the object code, whose source code is excluded
+from the Corresponding Source as a System Library, need not be
+included in conveying the object code work.
+
+  A "User Product" is either (1) a "consumer product", which means any
+tangible personal property which is normally used for personal, family,
+or household purposes, or (2) anything designed or sold for incorporation
+into a dwelling.  In determining whether a product is a consumer product,
+doubtful cases shall be resolved in favor of coverage.  For a particular
+product received by a particular user, "normally used" refers to a
+typical or common use of that class of product, regardless of the status
+of the particular user or of the way in which the particular user
+actually uses, or expects or is expected to use, the product.  A product
+is a consumer product regardless of whether the product has substantial
+commercial, industrial or non-consumer uses, unless such uses represent
+the only significant mode of use of the product.
+
+  "Installation Information" for a User Product means any methods,
+procedures, authorization keys, or other information required to install
+and execute modified versions of a covered work in that User Product from
+a modified version of its Corresponding Source.  The information must
+suffice to ensure that the continued functioning of the modified object
+code is in no case prevented or interfered with solely because
+modification has been made.
+
+  If you convey an object code work under this section in, or with, or
+specifically for use in, a User Product, and the conveying occurs as
+part of a transaction in which the right of possession and use of the
+User Product is transferred to the recipient in perpetuity or for a
+fixed term (regardless of how the transaction is characterized), the
+Corresponding Source conveyed under this section must be accompanied
+by the Installation Information.  But this requirement does not apply
+if neither you nor any third party retains the ability to install
+modified object code on the User Product (for example, the work has
+been installed in ROM).
+
+  The requirement to provide Installation Information does not include a
+requirement to continue to provide support service, warranty, or updates
+for a work that has been modified or installed by the recipient, or for
+the User Product in which it has been modified or installed.  Access to a
+network may be denied when the modification itself materially and
+adversely affects the operation of the network or violates the rules and
+protocols for communication across the network.
+
+  Corresponding Source conveyed, and Installation Information provided,
+in accord with this section must be in a format that is publicly
+documented (and with an implementation available to the public in
+source code form), and must require no special password or key for
+unpacking, reading or copying.
+
+  7. Additional Terms.
+
+  "Additional permissions" are terms that supplement the terms of this
+License by making exceptions from one or more of its conditions.
+Additional permissions that are applicable to the entire Program shall
+be treated as though they were included in this License, to the extent
+that they are valid under applicable law.  If additional permissions
+apply only to part of the Program, that part may be used separately
+under those permissions, but the entire Program remains governed by
+this License without regard to the additional permissions.
+
+  When you convey a copy of a covered work, you may at your option
+remove any additional permissions from that copy, or from any part of
+it.  (Additional permissions may be written to require their own
+removal in certain cases when you modify the work.)  You may place
+additional permissions on material, added by you to a covered work,
+for which you have or can give appropriate copyright permission.
+
+  Notwithstanding any other provision of this License, for material you
+add to a covered work, you may (if authorized by the copyright holders of
+that material) supplement the terms of this License with terms:
+
+    a) Disclaiming warranty or limiting liability differently from the
+    terms of sections 15 and 16 of this License; or
+
+    b) Requiring preservation of specified reasonable legal notices or
+    author attributions in that material or in the Appropriate Legal
+    Notices displayed by works containing it; or
+
+    c) Prohibiting misrepresentation of the origin of that material, or
+    requiring that modified versions of such material be marked in
+    reasonable ways as different from the original version; or
+
+    d) Limiting the use for publicity purposes of names of licensors or
+    authors of the material; or
+
+    e) Declining to grant rights under trademark law for use of some
+    trade names, trademarks, or service marks; or
+
+    f) Requiring indemnification of licensors and authors of that
+    material by anyone who conveys the material (or modified versions of
+    it) with contractual assumptions of liability to the recipient, for
+    any liability that these contractual assumptions directly impose on
+    those licensors and authors.
+
+  All other non-permissive additional terms are considered "further
+restrictions" within the meaning of section 10.  If the Program as you
+received it, or any part of it, contains a notice stating that it is
+governed by this License along with a term that is a further
+restriction, you may remove that term.  If a license document contains
+a further restriction but permits relicensing or conveying under this
+License, you may add to a covered work material governed by the terms
+of that license document, provided that the further restriction does
+not survive such relicensing or conveying.
+
+  If you add terms to a covered work in accord with this section, you
+must place, in the relevant source files, a statement of the
+additional terms that apply to those files, or a notice indicating
+where to find the applicable terms.
+
+  Additional terms, permissive or non-permissive, may be stated in the
+form of a separately written license, or stated as exceptions;
+the above requirements apply either way.
+
+  8. Termination.
+
+  You may not propagate or modify a covered work except as expressly
+provided under this License.  Any attempt otherwise to propagate or
+modify it is void, and will automatically terminate your rights under
+this License (including any patent licenses granted under the third
+paragraph of section 11).
+
+  However, if you cease all violation of this License, then your
+license from a particular copyright holder is reinstated (a)
+provisionally, unless and until the copyright holder explicitly and
+finally terminates your license, and (b) permanently, if the copyright
+holder fails to notify you of the violation by some reasonable means
+prior to 60 days after the cessation.
+
+  Moreover, your license from a particular copyright holder is
+reinstated permanently if the copyright holder notifies you of the
+violation by some reasonable means, this is the first time you have
+received notice of violation of this License (for any work) from that
+copyright holder, and you cure the violation prior to 30 days after
+your receipt of the notice.
+
+  Termination of your rights under this section does not terminate the
+licenses of parties who have received copies or rights from you under
+this License.  If your rights have been terminated and not permanently
+reinstated, you do not qualify to receive new licenses for the same
+material under section 10.
+
+  9. Acceptance Not Required for Having Copies.
+
+  You are not required to accept this License in order to receive or
+run a copy of the Program.  Ancillary propagation of a covered work
+occurring solely as a consequence of using peer-to-peer transmission
+to receive a copy likewise does not require acceptance.  However,
+nothing other than this License grants you permission to propagate or
+modify any covered work.  These actions infringe copyright if you do
+not accept this License.  Therefore, by modifying or propagating a
+covered work, you indicate your acceptance of this License to do so.
+
+  10. Automatic Licensing of Downstream Recipients.
+
+  Each time you convey a covered work, the recipient automatically
+receives a license from the original licensors, to run, modify and
+propagate that work, subject to this License.  You are not responsible
+for enforcing compliance by third parties with this License.
+
+  An "entity transaction" is a transaction transferring control of an
+organization, or substantially all assets of one, or subdividing an
+organization, or merging organizations.  If propagation of a covered
+work results from an entity transaction, each party to that
+transaction who receives a copy of the work also receives whatever
+licenses to the work the party's predecessor in interest had or could
+give under the previous paragraph, plus a right to possession of the
+Corresponding Source of the work from the predecessor in interest, if
+the predecessor has it or can get it with reasonable efforts.
+
+  You may not impose any further restrictions on the exercise of the
+rights granted or affirmed under this License.  For example, you may
+not impose a license fee, royalty, or other charge for exercise of
+rights granted under this License, and you may not initiate litigation
+(including a cross-claim or counterclaim in a lawsuit) alleging that
+any patent claim is infringed by making, using, selling, offering for
+sale, or importing the Program or any portion of it.
+
+  11. Patents.
+
+  A "contributor" is a copyright holder who authorizes use under this
+License of the Program or a work on which the Program is based.  The
+work thus licensed is called the contributor's "contributor version".
+
+  A contributor's "essential patent claims" are all patent claims
+owned or controlled by the contributor, whether already acquired or
+hereafter acquired, that would be infringed by some manner, permitted
+by this License, of making, using, or selling its contributor version,
+but do not include claims that would be infringed only as a
+consequence of further modification of the contributor version.  For
+purposes of this definition, "control" includes the right to grant
+patent sublicenses in a manner consistent with the requirements of
+this License.
+
+  Each contributor grants you a non-exclusive, worldwide, royalty-free
+patent license under the contributor's essential patent claims, to
+make, use, sell, offer for sale, import and otherwise run, modify and
+propagate the contents of its contributor version.
+
+  In the following three paragraphs, a "patent license" is any express
+agreement or commitment, however denominated, not to enforce a patent
+(such as an express permission to practice a patent or covenant not to
+sue for patent infringement).  To "grant" such a patent license to a
+party means to make such an agreement or commitment not to enforce a
+patent against the party.
+
+  If you convey a covered work, knowingly relying on a patent license,
+and the Corresponding Source of the work is not available for anyone
+to copy, free of charge and under the terms of this License, through a
+publicly available network server or other readily accessible means,
+then you must either (1) cause the Corresponding Source to be so
+available, or (2) arrange to deprive yourself of the benefit of the
+patent license for this particular work, or (3) arrange, in a manner
+consistent with the requirements of this License, to extend the patent
+license to downstream recipients.  "Knowingly relying" means you have
+actual knowledge that, but for the patent license, your conveying the
+covered work in a country, or your recipient's use of the covered work
+in a country, would infringe one or more identifiable patents in that
+country that you have reason to believe are valid.
+
+  If, pursuant to or in connection with a single transaction or
+arrangement, you convey, or propagate by procuring conveyance of, a
+covered work, and grant a patent license to some of the parties
+receiving the covered work authorizing them to use, propagate, modify
+or convey a specific copy of the covered work, then the patent license
+you grant is automatically extended to all recipients of the covered
+work and works based on it.
+
+  A patent license is "discriminatory" if it does not include within
+the scope of its coverage, prohibits the exercise of, or is
+conditioned on the non-exercise of one or more of the rights that are
+specifically granted under this License.  You may not convey a covered
+work if you are a party to an arrangement with a third party that is
+in the business of distributing software, under which you make payment
+to the third party based on the extent of your activity of conveying
+the work, and under which the third party grants, to any of the
+parties who would receive the covered work from you, a discriminatory
+patent license (a) in connection with copies of the covered work
+conveyed by you (or copies made from those copies), or (b) primarily
+for and in connection with specific products or compilations that
+contain the covered work, unless you entered into that arrangement,
+or that patent license was granted, prior to 28 March 2007.
+
+  Nothing in this License shall be construed as excluding or limiting
+any implied license or other defenses to infringement that may
+otherwise be available to you under applicable patent law.
+
+  12. No Surrender of Others' Freedom.
+
+  If conditions are imposed on you (whether by court order, agreement or
+otherwise) that contradict the conditions of this License, they do not
+excuse you from the conditions of this License.  If you cannot convey a
+covered work so as to satisfy simultaneously your obligations under this
+License and any other pertinent obligations, then as a consequence you may
+not convey it at all.  For example, if you agree to terms that obligate you
+to collect a royalty for further conveying from those to whom you convey
+the Program, the only way you could satisfy both those terms and this
+License would be to refrain entirely from conveying the Program.
+
+  13. Use with the GNU Affero General Public License.
+
+  Notwithstanding any other provision of this License, you have
+permission to link or combine any covered work with a work licensed
+under version 3 of the GNU Affero General Public License into a single
+combined work, and to convey the resulting work.  The terms of this
+License will continue to apply to the part which is the covered work,
+but the special requirements of the GNU Affero General Public License,
+section 13, concerning interaction through a network will apply to the
+combination as such.
+
+  14. Revised Versions of this License.
+
+  The Free Software Foundation may publish revised and/or new versions of
+the GNU General Public License from time to time.  Such new versions will
+be similar in spirit to the present version, but may differ in detail to
+address new problems or concerns.
+
+  Each version is given a distinguishing version number.  If the
+Program specifies that a certain numbered version of the GNU General
+Public License "or any later version" applies to it, you have the
+option of following the terms and conditions either of that numbered
+version or of any later version published by the Free Software
+Foundation.  If the Program does not specify a version number of the
+GNU General Public License, you may choose any version ever published
+by the Free Software Foundation.
+
+  If the Program specifies that a proxy can decide which future
+versions of the GNU General Public License can be used, that proxy's
+public statement of acceptance of a version permanently authorizes you
+to choose that version for the Program.
+
+  Later license versions may give you additional or different
+permissions.  However, no additional obligations are imposed on any
+author or copyright holder as a result of your choosing to follow a
+later version.
+
+  15. Disclaimer of Warranty.
+
+  THERE IS NO WARRANTY FOR THE PROGRAM, TO THE EXTENT PERMITTED BY
+APPLICABLE LAW.  EXCEPT WHEN OTHERWISE STATED IN WRITING THE COPYRIGHT
+HOLDERS AND/OR OTHER PARTIES PROVIDE THE PROGRAM "AS IS" WITHOUT WARRANTY
+OF ANY KIND, EITHER EXPRESSED OR IMPLIED, INCLUDING, BUT NOT LIMITED TO,
+THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A PARTICULAR
+PURPOSE.  THE ENTIRE RISK AS TO THE QUALITY AND PERFORMANCE OF THE PROGRAM
+IS WITH YOU.  SHOULD THE PROGRAM PROVE DEFECTIVE, YOU ASSUME THE COST OF
+ALL NECESSARY SERVICING, REPAIR OR CORRECTION.
+
+  16. Limitation of Liability.
+
+  IN NO EVENT UNLESS REQUIRED BY APPLICABLE LAW OR AGREED TO IN WRITING
+WILL ANY COPYRIGHT HOLDER, OR ANY OTHER PARTY WHO MODIFIES AND/OR CONVEYS
+THE PROGRAM AS PERMITTED ABOVE, BE LIABLE TO YOU FOR DAMAGES, INCLUDING ANY
+GENERAL, SPECIAL, INCIDENTAL OR CONSEQUENTIAL DAMAGES ARISING OUT OF THE
+USE OR INABILITY TO USE THE PROGRAM (INCLUDING BUT NOT LIMITED TO LOSS OF
+DATA OR DATA BEING RENDERED INACCURATE OR LOSSES SUSTAINED BY YOU OR THIRD
+PARTIES OR A FAILURE OF THE PROGRAM TO OPERATE WITH ANY OTHER PROGRAMS),
+EVEN IF SUCH HOLDER OR OTHER PARTY HAS BEEN ADVISED OF THE POSSIBILITY OF
+SUCH DAMAGES.
+
+  17. Interpretation of Sections 15 and 16.
+
+  If the disclaimer of warranty and limitation of liability provided
+above cannot be given local legal effect according to their terms,
+reviewing courts shall apply local law that most closely approximates
+an absolute waiver of all civil liability in connection with the
+Program, unless a warranty or assumption of liability accompanies a
+copy of the Program in return for a fee.
+
+                     END OF TERMS AND CONDITIONS
+
+            How to Apply These Terms to Your New Programs
+
+  If you develop a new program, and you want it to be of the greatest
+possible use to the public, the best way to achieve this is to make it
+free software which everyone can redistribute and change under these terms.
+
+  To do so, attach the following notices to the program.  It is safest
+to attach them to the start of each source file to most effectively
+state the exclusion of warranty; and each file should have at least
+the "copyright" line and a pointer to where the full notice is found.
+
+    <one line to give the program's name and a brief idea of what it does.>
+    Copyright (C) <year>  <name of author>
+
+    This program is free software: you can redistribute it and/or modify
+    it under the terms of the GNU General Public License as published by
+    the Free Software Foundation, either version 3 of the License, or
+    (at your option) any later version.
+
+    This program is distributed in the hope that it will be useful,
+    but WITHOUT ANY WARRANTY; without even the implied warranty of
+    MERCHANTABILITY or FITNESS FOR A PARTICULAR PURPOSE.  See the
+    GNU General Public License for more details.
+
+    You should have received a copy of the GNU General Public License
+    along with this program.  If not, see <https://www.gnu.org/licenses/>.
+
+Also add information on how to contact you by electronic and paper mail.
+
+  If the program does terminal interaction, make it output a short
+notice like this when it starts in an interactive mode:
+
+    <program>  Copyright (C) <year>  <name of author>
+    This program comes with ABSOLUTELY NO WARRANTY; for details type `show w'.
+    This is free software, and you are welcome to redistribute it
+    under certain conditions; type `show c' for details.
+
+The hypothetical commands `show w' and `show c' should show the appropriate
+parts of the General Public License.  Of course, your program's commands
+might be different; for a GUI interface, you would use an "about box".
+
+  You should also get your employer (if you work as a programmer) or school,
+if any, to sign a "copyright disclaimer" for the program, if necessary.
+For more information on this, and how to apply and follow the GNU GPL, see
+<https://www.gnu.org/licenses/>.
+
+  The GNU General Public License does not permit incorporating your program
+into proprietary programs.  If your program is a subroutine library, you
+may consider it more useful to permit linking proprietary applications with
+the library.  If this is what you want to do, use the GNU Lesser General
+Public License instead of this License.  But first, please read
+<https://www.gnu.org/licenses/why-not-lgpl.html>.
diff --git a/MNIST/Train.py b/MNIST/Train.py
deleted file mode 100755
index 8a8b846..0000000
--- a/MNIST/Train.py
+++ /dev/null
@@ -1,162 +0,0 @@
-import numpy as np
-import pickle
-import math
-import tensorflow as tf
-from tensorflow import keras
-from tensorflow.keras.datasets import cifar10, mnist
-from tensorflow.keras.optimizers import SGD
-from tensorflow.keras.preprocessing.image import ImageDataGenerator
-import os
-import sys
-sys.path.insert(0, '..')
-from model_architectures import get_model
-
-
-def load_svhn(path_to_dataset):
-    import scipy.io as sio
-    train = sio.loadmat(path_to_dataset+'/train.mat')
-    test = sio.loadmat(path_to_dataset+'/test.mat')
-    extra = sio.loadmat(path_to_dataset+'/extra.mat')
-    x_train = np.transpose(train['X'], [3, 0, 1, 2])
-    y_train = train['y']-1
-
-    x_test = np.transpose(test['X'], [3, 0, 1, 2])
-    y_test = test['y']-1
-
-    x_extra = np.transpose(extra['X'], [3, 0, 1, 2])
-    y_extra=extra['y']-1
-
-    x_train = np.concatenate((x_train, x_extra), axis=0)
-    y_train = np.concatenate((y_train, y_extra), axis=0)
-
-    return (x_train, y_train), (x_test, y_test)
-
-def get_examples(dataset):
-    if dataset == "MNIST":
-        (x_train, y_train), (x_test, y_test) = mnist.load_data()
-        # convert class vectors to binary class matrices
-        x_train = x_train.reshape(-1,784)
-        x_test = x_test.reshape(-1,784)
-        use_generator = False
-    elif dataset == "CIFAR-10" or dataset == "binarynet":
-        use_generator = True
-        (x_train, y_train), (x_test, y_test) = cifar10.load_data()
-    elif dataset == "SVHN" or dataset == "binarynet-svhn":
-        use_generator = True
-        (x_train, y_train), (x_test, y_test) = load_svhn('./svhn_data')
-    else:
-        raise("dataset should be one of the following: [MNIST, CIFAR-10, SVHN, binarynet, binarynet-svhn].")
-
-    x_train = x_train.astype(np.float32)
-    x_test = x_test.astype(np.float32)
-    x_train /= 255
-    x_test /= 255
-
-    print('x_train shape:', x_train.shape)
-    print(x_train.shape[0], 'train samples')
-    print(x_test.shape[0], 'test samples')
-
-    return x_train, y_train, x_test, y_test, use_generator
-
-# learning rate schedule
-def step_decay(epoch):
-    initial_lrate = 0.025
-    drop = 0.5
-    epochs_drop = 50.0
-    lrate = initial_lrate * math.pow(drop, math.floor((1+epoch)/epochs_drop))
-    return lrate
-
-def train_network(dir_path, dataset, x_train, y_train, num_epochs, use_generator, batch_size, group_idx=None, is_regularized=False):
-    os.makedirs(dir_path, exist_ok=True)
-
-    model = get_model(dataset, batch_size, group_idx=group_idx, is_regularized=is_regularized)
-
-    lr = 0.01
-    opt = keras.optimizers.SGD(lr=lr)
-    model.compile(
-            loss='sparse_categorical_crossentropy',optimizer=opt, metrics=['accuracy'])
-
-    weights_path = dir_path + "/output_model.h5"
-    cback=keras.callbacks.ModelCheckpoint(
-            weights_path, monitor='val_accuracy', save_best_only=True)
-
-    if use_generator:
-        if dataset=="CIFAR-10" or dataset=="binarynet":
-            horizontal_flip=True
-        if dataset=="SVHN" or dataset=="binarynet-svhn":
-            horizontal_flip=False
-
-        datagen = ImageDataGenerator(
-                width_shift_range=0.15,  # randomly shift images horizontally (fraction of total width)
-                height_shift_range=0.15,  # randomly shift images vertically (fraction of total height)
-                horizontal_flip=horizontal_flip)  # randomly flip images
-        if keras.__version__[0]=='2':
-            history=model.fit_generator(
-                    datagen.flow(x_train, y_train, batch_size=batch_size),
-                    steps_per_epoch=x_train.shape[0]/batch_size,
-                    nb_epoch=num_epochs, validation_split=0.1,
-                    verbose=2, callbacks=[cback])
-        if keras.__version__[0]=='1':
-            history=model.fit_generator(
-                    datagen.flow(x_train, y_train,batch_size=batch_size),
-                    samples_per_epoch=x_train.shape[0], nb_epoch=num_epochs,
-                    verbose=2, validation_split=0.1, callbacks=[cback])
-    else:
-        if keras.__version__[0]=='2':
-            history=model.fit(x_train, y_train, batch_size=batch_size,validation_split=0.1, verbose=2, epochs=num_epochs,callbacks=[cback])
-        if keras.__version__[0]=='1':
-            history=model.fit(x_train, y_train,batch_size=batch_size,validation_split=0.1, verbose=2,nb_epoch=num_epochs,callbacks=[cback])
-
-    dic={'hard': history.history}
-    history_path = dir_path + "/history_output_model.pkl"
-    with open(history_path,'wb') as handle:
-        pickle.dump(dic, handle)
-
-def evaluate_network(dir_path, dataset, x_test, y_test, batch_size, group_idx=None, is_regularized=True, log_dir_full_path=None):
-    weights_path= dir_path + "/output_model.h5"
-    model = get_model(dataset, batch_size, group_idx=group_idx, is_regularized=is_regularized, log_dir_full_path=log_dir_full_path)
-    model.load_weights(weights_path)
-    opt = keras.optimizers.SGD()
-    model.compile(loss='sparse_categorical_crossentropy', optimizer=opt, metrics=['accuracy'])
-    score = model.evaluate(x_test, y_test, verbose=0, batch_size=batch_size)
-    if log_dir_full_path is None:
-        print("Test loss was %0.4f, test accuracy was %0.4f"%(score[0], score[1]))
-    else:
-        log_file_full_path = "{}/accuracy.csv".format(log_dir_full_path)
-        open(log_file_full_path, "a").close()
-        tf.print(score[1], output_stream="file://{}".format(log_file_full_path))
-
-
-
-dataset = "MNIST"
-path_to_project = "path/to/project"
-Train = True
-Evaluate = True
-if Train:
-    batch_size = 100
-else:
-    batch_size = 10000
-num_epochs = 1000
-x_train, y_train, x_test, y_test, use_generator = get_examples(dataset)
-
-group_idxs = [0, 2, 1]
-num_training_repeats = 5
-num_inference_repeats = 25
-
-for is_regularized in [True, False]:
-    if is_regularized:
-        regularized_label = "regularized"
-    else:
-        regularized_label = "non-regularized"
-    for group_idx in group_idxs:
-        log_dir_full_path = "{}/MNIST/models/{}/{}/group-{}".format(
-                path_to_project, dataset, regularized_label, group_idx)
-        for training_repeat_idx in range(num_training_repeats):
-            dir_path = "models/{}/{}/group-{}/network-{}".format(
-                    dataset, regularized_label, group_idx, training_repeat_idx)
-            if Train:
-                train_network(dir_path, dataset, x_train, y_train, num_epochs, use_generator, batch_size, group_idx=group_idx, is_regularized=is_regularized)
-            if Evaluate:
-                for inference_repeat_idx in range(num_inference_repeats):
-                    evaluate_network(dir_path, dataset, x_test, y_test, batch_size, group_idx=group_idx, is_regularized=is_regularized, log_dir_full_path=log_dir_full_path)
-
diff --git a/MNIST/nonlinear_autograd.txt b/MNIST/nonlinear_autograd.txt
deleted file mode 100644
index d58a3e3..0000000
--- a/MNIST/nonlinear_autograd.txt
+++ /dev/null
@@ -1,2005 +0,0 @@
-X_train shape: (60000, 784)
-60000 train samples
-10000 test samples
-Train on 60000 samples, validate on 10000 samples
-Epoch 1/1000
-60000/60000 - 9s - loss: 0.7709 - accuracy: 0.7652 - val_loss: 0.4542 - val_accuracy: 0.8667
-Epoch 2/1000
-60000/60000 - 6s - loss: 0.4109 - accuracy: 0.8801 - val_loss: 0.3558 - val_accuracy: 0.8991
-Epoch 3/1000
-60000/60000 - 6s - loss: 0.3375 - accuracy: 0.9016 - val_loss: 0.3199 - val_accuracy: 0.9029
-Epoch 4/1000
-60000/60000 - 6s - loss: 0.2981 - accuracy: 0.9125 - val_loss: 0.2970 - val_accuracy: 0.9117
-Epoch 5/1000
-60000/60000 - 6s - loss: 0.2747 - accuracy: 0.9191 - val_loss: 0.2739 - val_accuracy: 0.9171
-Epoch 6/1000
-60000/60000 - 6s - loss: 0.2559 - accuracy: 0.9245 - val_loss: 0.2680 - val_accuracy: 0.9189
-Epoch 7/1000
-60000/60000 - 6s - loss: 0.2403 - accuracy: 0.9283 - val_loss: 0.2660 - val_accuracy: 0.9202
-Epoch 8/1000
-60000/60000 - 6s - loss: 0.2273 - accuracy: 0.9327 - val_loss: 0.2586 - val_accuracy: 0.9200
-Epoch 9/1000
-60000/60000 - 6s - loss: 0.2156 - accuracy: 0.9370 - val_loss: 0.2522 - val_accuracy: 0.9237
-Epoch 10/1000
-60000/60000 - 6s - loss: 0.2059 - accuracy: 0.9388 - val_loss: 0.2586 - val_accuracy: 0.9228
-Epoch 11/1000
-60000/60000 - 6s - loss: 0.1988 - accuracy: 0.9414 - val_loss: 0.2469 - val_accuracy: 0.9263
-Epoch 12/1000
-60000/60000 - 6s - loss: 0.1920 - accuracy: 0.9438 - val_loss: 0.2464 - val_accuracy: 0.9263
-Epoch 13/1000
-60000/60000 - 6s - loss: 0.1854 - accuracy: 0.9453 - val_loss: 0.2431 - val_accuracy: 0.9281
-Epoch 14/1000
-60000/60000 - 6s - loss: 0.1802 - accuracy: 0.9469 - val_loss: 0.2466 - val_accuracy: 0.9252
-Epoch 15/1000
-60000/60000 - 6s - loss: 0.1742 - accuracy: 0.9486 - val_loss: 0.2430 - val_accuracy: 0.9294
-Epoch 16/1000
-60000/60000 - 6s - loss: 0.1711 - accuracy: 0.9494 - val_loss: 0.2393 - val_accuracy: 0.9300
-Epoch 17/1000
-60000/60000 - 6s - loss: 0.1670 - accuracy: 0.9502 - val_loss: 0.2432 - val_accuracy: 0.9287
-Epoch 18/1000
-60000/60000 - 6s - loss: 0.1619 - accuracy: 0.9522 - val_loss: 0.2384 - val_accuracy: 0.9324
-Epoch 19/1000
-60000/60000 - 6s - loss: 0.1577 - accuracy: 0.9532 - val_loss: 0.2420 - val_accuracy: 0.9298
-Epoch 20/1000
-60000/60000 - 6s - loss: 0.1533 - accuracy: 0.9552 - val_loss: 0.2388 - val_accuracy: 0.9304
-Epoch 21/1000
-60000/60000 - 6s - loss: 0.1508 - accuracy: 0.9551 - val_loss: 0.2442 - val_accuracy: 0.9279
-Epoch 22/1000
-60000/60000 - 6s - loss: 0.1477 - accuracy: 0.9562 - val_loss: 0.2417 - val_accuracy: 0.9285
-Epoch 23/1000
-60000/60000 - 6s - loss: 0.1457 - accuracy: 0.9567 - val_loss: 0.2384 - val_accuracy: 0.9273
-Epoch 24/1000
-60000/60000 - 6s - loss: 0.1429 - accuracy: 0.9579 - val_loss: 0.2403 - val_accuracy: 0.9312
-Epoch 25/1000
-60000/60000 - 6s - loss: 0.1400 - accuracy: 0.9585 - val_loss: 0.2350 - val_accuracy: 0.9310
-Epoch 26/1000
-60000/60000 - 6s - loss: 0.1374 - accuracy: 0.9596 - val_loss: 0.2388 - val_accuracy: 0.9293
-Epoch 27/1000
-60000/60000 - 6s - loss: 0.1342 - accuracy: 0.9603 - val_loss: 0.2398 - val_accuracy: 0.9297
-Epoch 28/1000
-60000/60000 - 6s - loss: 0.1301 - accuracy: 0.9621 - val_loss: 0.2356 - val_accuracy: 0.9310
-Epoch 29/1000
-60000/60000 - 6s - loss: 0.1296 - accuracy: 0.9622 - val_loss: 0.2326 - val_accuracy: 0.9322
-Epoch 30/1000
-60000/60000 - 6s - loss: 0.1269 - accuracy: 0.9629 - val_loss: 0.2450 - val_accuracy: 0.9283
-Epoch 31/1000
-60000/60000 - 6s - loss: 0.1249 - accuracy: 0.9641 - val_loss: 0.2444 - val_accuracy: 0.9281
-Epoch 32/1000
-60000/60000 - 6s - loss: 0.1237 - accuracy: 0.9640 - val_loss: 0.2431 - val_accuracy: 0.9308
-Epoch 33/1000
-60000/60000 - 6s - loss: 0.1227 - accuracy: 0.9644 - val_loss: 0.2406 - val_accuracy: 0.9310
-Epoch 34/1000
-60000/60000 - 6s - loss: 0.1220 - accuracy: 0.9641 - val_loss: 0.2445 - val_accuracy: 0.9302
-Epoch 35/1000
-60000/60000 - 6s - loss: 0.1195 - accuracy: 0.9652 - val_loss: 0.2401 - val_accuracy: 0.9319
-Epoch 36/1000
-60000/60000 - 6s - loss: 0.1181 - accuracy: 0.9658 - val_loss: 0.2423 - val_accuracy: 0.9300
-Epoch 37/1000
-60000/60000 - 6s - loss: 0.1166 - accuracy: 0.9665 - val_loss: 0.2451 - val_accuracy: 0.9294
-Epoch 38/1000
-60000/60000 - 6s - loss: 0.1152 - accuracy: 0.9665 - val_loss: 0.2485 - val_accuracy: 0.9286
-Epoch 39/1000
-60000/60000 - 6s - loss: 0.1123 - accuracy: 0.9673 - val_loss: 0.2440 - val_accuracy: 0.9290
-Epoch 40/1000
-60000/60000 - 6s - loss: 0.1121 - accuracy: 0.9678 - val_loss: 0.2479 - val_accuracy: 0.9301
-Epoch 41/1000
-60000/60000 - 6s - loss: 0.1095 - accuracy: 0.9686 - val_loss: 0.2423 - val_accuracy: 0.9298
-Epoch 42/1000
-60000/60000 - 6s - loss: 0.1086 - accuracy: 0.9689 - val_loss: 0.2558 - val_accuracy: 0.9298
-Epoch 43/1000
-60000/60000 - 6s - loss: 0.1093 - accuracy: 0.9687 - val_loss: 0.2535 - val_accuracy: 0.9285
-Epoch 44/1000
-60000/60000 - 6s - loss: 0.1100 - accuracy: 0.9682 - val_loss: 0.2487 - val_accuracy: 0.9291
-Epoch 45/1000
-60000/60000 - 6s - loss: 0.1062 - accuracy: 0.9694 - val_loss: 0.2486 - val_accuracy: 0.9300
-Epoch 46/1000
-60000/60000 - 6s - loss: 0.1049 - accuracy: 0.9709 - val_loss: 0.2478 - val_accuracy: 0.9310
-Epoch 47/1000
-60000/60000 - 6s - loss: 0.1032 - accuracy: 0.9706 - val_loss: 0.2462 - val_accuracy: 0.9301
-Epoch 48/1000
-60000/60000 - 6s - loss: 0.1025 - accuracy: 0.9714 - val_loss: 0.2496 - val_accuracy: 0.9292
-Epoch 49/1000
-60000/60000 - 6s - loss: 0.1007 - accuracy: 0.9719 - val_loss: 0.2462 - val_accuracy: 0.9316
-Epoch 50/1000
-60000/60000 - 6s - loss: 0.0996 - accuracy: 0.9723 - val_loss: 0.2511 - val_accuracy: 0.9300
-Epoch 51/1000
-60000/60000 - 6s - loss: 0.0986 - accuracy: 0.9725 - val_loss: 0.2544 - val_accuracy: 0.9304
-Epoch 52/1000
-60000/60000 - 6s - loss: 0.0998 - accuracy: 0.9721 - val_loss: 0.2481 - val_accuracy: 0.9321
-Epoch 53/1000
-60000/60000 - 6s - loss: 0.0988 - accuracy: 0.9726 - val_loss: 0.2531 - val_accuracy: 0.9288
-Epoch 54/1000
-60000/60000 - 6s - loss: 0.0973 - accuracy: 0.9733 - val_loss: 0.2497 - val_accuracy: 0.9310
-Epoch 55/1000
-60000/60000 - 6s - loss: 0.0988 - accuracy: 0.9724 - val_loss: 0.2524 - val_accuracy: 0.9285
-Epoch 56/1000
-60000/60000 - 6s - loss: 0.0954 - accuracy: 0.9735 - val_loss: 0.2546 - val_accuracy: 0.9313
-Epoch 57/1000
-60000/60000 - 6s - loss: 0.0943 - accuracy: 0.9737 - val_loss: 0.2547 - val_accuracy: 0.9304
-Epoch 58/1000
-60000/60000 - 6s - loss: 0.0953 - accuracy: 0.9732 - val_loss: 0.2540 - val_accuracy: 0.9289
-Epoch 59/1000
-60000/60000 - 6s - loss: 0.0954 - accuracy: 0.9737 - val_loss: 0.2603 - val_accuracy: 0.9290
-Epoch 60/1000
-60000/60000 - 6s - loss: 0.0931 - accuracy: 0.9743 - val_loss: 0.2615 - val_accuracy: 0.9287
-Epoch 61/1000
-60000/60000 - 6s - loss: 0.0928 - accuracy: 0.9743 - val_loss: 0.2574 - val_accuracy: 0.9305
-Epoch 62/1000
-60000/60000 - 6s - loss: 0.0905 - accuracy: 0.9754 - val_loss: 0.2611 - val_accuracy: 0.9290
-Epoch 63/1000
-60000/60000 - 6s - loss: 0.0905 - accuracy: 0.9752 - val_loss: 0.2633 - val_accuracy: 0.9294
-Epoch 64/1000
-60000/60000 - 6s - loss: 0.0905 - accuracy: 0.9751 - val_loss: 0.2626 - val_accuracy: 0.9288
-Epoch 65/1000
-60000/60000 - 6s - loss: 0.0891 - accuracy: 0.9760 - val_loss: 0.2588 - val_accuracy: 0.9283
-Epoch 66/1000
-60000/60000 - 6s - loss: 0.0896 - accuracy: 0.9758 - val_loss: 0.2612 - val_accuracy: 0.9305
-Epoch 67/1000
-60000/60000 - 6s - loss: 0.0888 - accuracy: 0.9759 - val_loss: 0.2625 - val_accuracy: 0.9308
-Epoch 68/1000
-60000/60000 - 6s - loss: 0.0881 - accuracy: 0.9759 - val_loss: 0.2648 - val_accuracy: 0.9307
-Epoch 69/1000
-60000/60000 - 6s - loss: 0.0876 - accuracy: 0.9764 - val_loss: 0.2605 - val_accuracy: 0.9291
-Epoch 70/1000
-60000/60000 - 6s - loss: 0.0872 - accuracy: 0.9765 - val_loss: 0.2620 - val_accuracy: 0.9288
-Epoch 71/1000
-60000/60000 - 6s - loss: 0.0871 - accuracy: 0.9767 - val_loss: 0.2585 - val_accuracy: 0.9295
-Epoch 72/1000
-60000/60000 - 6s - loss: 0.0878 - accuracy: 0.9762 - val_loss: 0.2631 - val_accuracy: 0.9288
-Epoch 73/1000
-60000/60000 - 6s - loss: 0.0862 - accuracy: 0.9771 - val_loss: 0.2665 - val_accuracy: 0.9284
-Epoch 74/1000
-60000/60000 - 6s - loss: 0.0853 - accuracy: 0.9765 - val_loss: 0.2652 - val_accuracy: 0.9278
-Epoch 75/1000
-60000/60000 - 6s - loss: 0.0840 - accuracy: 0.9778 - val_loss: 0.2651 - val_accuracy: 0.9287
-Epoch 76/1000
-60000/60000 - 6s - loss: 0.0838 - accuracy: 0.9779 - val_loss: 0.2603 - val_accuracy: 0.9310
-Epoch 77/1000
-60000/60000 - 6s - loss: 0.0841 - accuracy: 0.9776 - val_loss: 0.2692 - val_accuracy: 0.9260
-Epoch 78/1000
-60000/60000 - 6s - loss: 0.0836 - accuracy: 0.9779 - val_loss: 0.2718 - val_accuracy: 0.9279
-Epoch 79/1000
-60000/60000 - 6s - loss: 0.0837 - accuracy: 0.9780 - val_loss: 0.2644 - val_accuracy: 0.9295
-Epoch 80/1000
-60000/60000 - 6s - loss: 0.0831 - accuracy: 0.9778 - val_loss: 0.2675 - val_accuracy: 0.9283
-Epoch 81/1000
-60000/60000 - 6s - loss: 0.0811 - accuracy: 0.9780 - val_loss: 0.2663 - val_accuracy: 0.9298
-Epoch 82/1000
-60000/60000 - 6s - loss: 0.0819 - accuracy: 0.9782 - val_loss: 0.2658 - val_accuracy: 0.9287
-Epoch 83/1000
-60000/60000 - 6s - loss: 0.0819 - accuracy: 0.9784 - val_loss: 0.2709 - val_accuracy: 0.9277
-Epoch 84/1000
-60000/60000 - 6s - loss: 0.0812 - accuracy: 0.9787 - val_loss: 0.2698 - val_accuracy: 0.9280
-Epoch 85/1000
-60000/60000 - 6s - loss: 0.0810 - accuracy: 0.9790 - val_loss: 0.2667 - val_accuracy: 0.9290
-Epoch 86/1000
-60000/60000 - 6s - loss: 0.0805 - accuracy: 0.9786 - val_loss: 0.2685 - val_accuracy: 0.9280
-Epoch 87/1000
-60000/60000 - 6s - loss: 0.0804 - accuracy: 0.9785 - val_loss: 0.2727 - val_accuracy: 0.9285
-Epoch 88/1000
-60000/60000 - 6s - loss: 0.0779 - accuracy: 0.9798 - val_loss: 0.2718 - val_accuracy: 0.9292
-Epoch 89/1000
-60000/60000 - 6s - loss: 0.0781 - accuracy: 0.9797 - val_loss: 0.2716 - val_accuracy: 0.9287
-Epoch 90/1000
-60000/60000 - 6s - loss: 0.0779 - accuracy: 0.9797 - val_loss: 0.2735 - val_accuracy: 0.9286
-Epoch 91/1000
-60000/60000 - 6s - loss: 0.0783 - accuracy: 0.9796 - val_loss: 0.2752 - val_accuracy: 0.9265
-Epoch 92/1000
-60000/60000 - 6s - loss: 0.0770 - accuracy: 0.9800 - val_loss: 0.2749 - val_accuracy: 0.9291
-Epoch 93/1000
-60000/60000 - 6s - loss: 0.0768 - accuracy: 0.9802 - val_loss: 0.2756 - val_accuracy: 0.9283
-Epoch 94/1000
-60000/60000 - 6s - loss: 0.0783 - accuracy: 0.9792 - val_loss: 0.2721 - val_accuracy: 0.9287
-Epoch 95/1000
-60000/60000 - 6s - loss: 0.0778 - accuracy: 0.9797 - val_loss: 0.2725 - val_accuracy: 0.9304
-Epoch 96/1000
-60000/60000 - 6s - loss: 0.0763 - accuracy: 0.9799 - val_loss: 0.2742 - val_accuracy: 0.9291
-Epoch 97/1000
-60000/60000 - 6s - loss: 0.0772 - accuracy: 0.9798 - val_loss: 0.2783 - val_accuracy: 0.9296
-Epoch 98/1000
-60000/60000 - 6s - loss: 0.0761 - accuracy: 0.9804 - val_loss: 0.2821 - val_accuracy: 0.9276
-Epoch 99/1000
-60000/60000 - 6s - loss: 0.0767 - accuracy: 0.9798 - val_loss: 0.2818 - val_accuracy: 0.9274
-Epoch 100/1000
-60000/60000 - 6s - loss: 0.0759 - accuracy: 0.9801 - val_loss: 0.2786 - val_accuracy: 0.9274
-Epoch 101/1000
-60000/60000 - 6s - loss: 0.0752 - accuracy: 0.9804 - val_loss: 0.2767 - val_accuracy: 0.9285
-Epoch 102/1000
-60000/60000 - 6s - loss: 0.0757 - accuracy: 0.9800 - val_loss: 0.2811 - val_accuracy: 0.9271
-Epoch 103/1000
-60000/60000 - 6s - loss: 0.0742 - accuracy: 0.9808 - val_loss: 0.2777 - val_accuracy: 0.9285
-Epoch 104/1000
-60000/60000 - 6s - loss: 0.0742 - accuracy: 0.9810 - val_loss: 0.2820 - val_accuracy: 0.9290
-Epoch 105/1000
-60000/60000 - 6s - loss: 0.0736 - accuracy: 0.9806 - val_loss: 0.2806 - val_accuracy: 0.9283
-Epoch 106/1000
-60000/60000 - 6s - loss: 0.0765 - accuracy: 0.9796 - val_loss: 0.2779 - val_accuracy: 0.9282
-Epoch 107/1000
-60000/60000 - 6s - loss: 0.0756 - accuracy: 0.9799 - val_loss: 0.2811 - val_accuracy: 0.9278
-Epoch 108/1000
-60000/60000 - 6s - loss: 0.0746 - accuracy: 0.9803 - val_loss: 0.2803 - val_accuracy: 0.9283
-Epoch 109/1000
-60000/60000 - 6s - loss: 0.0752 - accuracy: 0.9802 - val_loss: 0.2857 - val_accuracy: 0.9275
-Epoch 110/1000
-60000/60000 - 6s - loss: 0.0738 - accuracy: 0.9810 - val_loss: 0.2804 - val_accuracy: 0.9286
-Epoch 111/1000
-60000/60000 - 6s - loss: 0.0748 - accuracy: 0.9804 - val_loss: 0.2841 - val_accuracy: 0.9281
-Epoch 112/1000
-60000/60000 - 6s - loss: 0.0742 - accuracy: 0.9808 - val_loss: 0.2833 - val_accuracy: 0.9298
-Epoch 113/1000
-60000/60000 - 6s - loss: 0.0744 - accuracy: 0.9806 - val_loss: 0.2834 - val_accuracy: 0.9267
-Epoch 114/1000
-60000/60000 - 6s - loss: 0.0722 - accuracy: 0.9815 - val_loss: 0.2823 - val_accuracy: 0.9280
-Epoch 115/1000
-60000/60000 - 6s - loss: 0.0716 - accuracy: 0.9815 - val_loss: 0.2818 - val_accuracy: 0.9285
-Epoch 116/1000
-60000/60000 - 6s - loss: 0.0711 - accuracy: 0.9817 - val_loss: 0.2813 - val_accuracy: 0.9290
-Epoch 117/1000
-60000/60000 - 6s - loss: 0.0713 - accuracy: 0.9818 - val_loss: 0.2849 - val_accuracy: 0.9283
-Epoch 118/1000
-60000/60000 - 6s - loss: 0.0722 - accuracy: 0.9813 - val_loss: 0.2847 - val_accuracy: 0.9277
-Epoch 119/1000
-60000/60000 - 6s - loss: 0.0729 - accuracy: 0.9807 - val_loss: 0.2849 - val_accuracy: 0.9275
-Epoch 120/1000
-60000/60000 - 6s - loss: 0.0712 - accuracy: 0.9816 - val_loss: 0.2843 - val_accuracy: 0.9274
-Epoch 121/1000
-60000/60000 - 6s - loss: 0.0705 - accuracy: 0.9819 - val_loss: 0.2890 - val_accuracy: 0.9267
-Epoch 122/1000
-60000/60000 - 6s - loss: 0.0704 - accuracy: 0.9821 - val_loss: 0.2866 - val_accuracy: 0.9276
-Epoch 123/1000
-60000/60000 - 6s - loss: 0.0696 - accuracy: 0.9821 - val_loss: 0.2856 - val_accuracy: 0.9269
-Epoch 124/1000
-60000/60000 - 6s - loss: 0.0693 - accuracy: 0.9823 - val_loss: 0.2859 - val_accuracy: 0.9271
-Epoch 125/1000
-60000/60000 - 6s - loss: 0.0698 - accuracy: 0.9819 - val_loss: 0.2848 - val_accuracy: 0.9276
-Epoch 126/1000
-60000/60000 - 6s - loss: 0.0695 - accuracy: 0.9822 - val_loss: 0.2867 - val_accuracy: 0.9270
-Epoch 127/1000
-60000/60000 - 6s - loss: 0.0688 - accuracy: 0.9824 - val_loss: 0.2850 - val_accuracy: 0.9273
-Epoch 128/1000
-60000/60000 - 6s - loss: 0.0687 - accuracy: 0.9824 - val_loss: 0.2888 - val_accuracy: 0.9286
-Epoch 129/1000
-60000/60000 - 6s - loss: 0.0693 - accuracy: 0.9821 - val_loss: 0.2863 - val_accuracy: 0.9280
-Epoch 130/1000
-60000/60000 - 6s - loss: 0.0674 - accuracy: 0.9829 - val_loss: 0.2844 - val_accuracy: 0.9297
-Epoch 131/1000
-60000/60000 - 6s - loss: 0.0685 - accuracy: 0.9826 - val_loss: 0.2880 - val_accuracy: 0.9284
-Epoch 132/1000
-60000/60000 - 6s - loss: 0.0680 - accuracy: 0.9831 - val_loss: 0.2932 - val_accuracy: 0.9267
-Epoch 133/1000
-60000/60000 - 6s - loss: 0.0683 - accuracy: 0.9825 - val_loss: 0.2942 - val_accuracy: 0.9268
-Epoch 134/1000
-60000/60000 - 6s - loss: 0.0684 - accuracy: 0.9821 - val_loss: 0.2895 - val_accuracy: 0.9275
-Epoch 135/1000
-60000/60000 - 6s - loss: 0.0676 - accuracy: 0.9829 - val_loss: 0.2927 - val_accuracy: 0.9285
-Epoch 136/1000
-60000/60000 - 6s - loss: 0.0669 - accuracy: 0.9829 - val_loss: 0.2939 - val_accuracy: 0.9266
-Epoch 137/1000
-60000/60000 - 6s - loss: 0.0674 - accuracy: 0.9830 - val_loss: 0.2936 - val_accuracy: 0.9267
-Epoch 138/1000
-60000/60000 - 6s - loss: 0.0662 - accuracy: 0.9834 - val_loss: 0.2942 - val_accuracy: 0.9273
-Epoch 139/1000
-60000/60000 - 6s - loss: 0.0659 - accuracy: 0.9832 - val_loss: 0.2905 - val_accuracy: 0.9277
-Epoch 140/1000
-60000/60000 - 6s - loss: 0.0666 - accuracy: 0.9834 - val_loss: 0.2936 - val_accuracy: 0.9276
-Epoch 141/1000
-60000/60000 - 6s - loss: 0.0693 - accuracy: 0.9821 - val_loss: 0.2961 - val_accuracy: 0.9269
-Epoch 142/1000
-60000/60000 - 6s - loss: 0.0669 - accuracy: 0.9831 - val_loss: 0.2945 - val_accuracy: 0.9281
-Epoch 143/1000
-60000/60000 - 6s - loss: 0.0658 - accuracy: 0.9833 - val_loss: 0.2936 - val_accuracy: 0.9273
-Epoch 144/1000
-60000/60000 - 6s - loss: 0.0655 - accuracy: 0.9833 - val_loss: 0.2970 - val_accuracy: 0.9279
-Epoch 145/1000
-60000/60000 - 6s - loss: 0.0657 - accuracy: 0.9833 - val_loss: 0.2951 - val_accuracy: 0.9271
-Epoch 146/1000
-60000/60000 - 6s - loss: 0.0660 - accuracy: 0.9831 - val_loss: 0.2949 - val_accuracy: 0.9272
-Epoch 147/1000
-60000/60000 - 6s - loss: 0.0662 - accuracy: 0.9835 - val_loss: 0.2964 - val_accuracy: 0.9255
-Epoch 148/1000
-60000/60000 - 6s - loss: 0.0654 - accuracy: 0.9838 - val_loss: 0.2965 - val_accuracy: 0.9272
-Epoch 149/1000
-60000/60000 - 6s - loss: 0.0644 - accuracy: 0.9839 - val_loss: 0.2954 - val_accuracy: 0.9277
-Epoch 150/1000
-60000/60000 - 6s - loss: 0.0645 - accuracy: 0.9837 - val_loss: 0.2975 - val_accuracy: 0.9259
-Epoch 151/1000
-60000/60000 - 6s - loss: 0.0662 - accuracy: 0.9829 - val_loss: 0.2966 - val_accuracy: 0.9280
-Epoch 152/1000
-60000/60000 - 6s - loss: 0.0656 - accuracy: 0.9834 - val_loss: 0.2970 - val_accuracy: 0.9268
-Epoch 153/1000
-60000/60000 - 6s - loss: 0.0640 - accuracy: 0.9841 - val_loss: 0.2983 - val_accuracy: 0.9265
-Epoch 154/1000
-60000/60000 - 6s - loss: 0.0662 - accuracy: 0.9830 - val_loss: 0.2965 - val_accuracy: 0.9274
-Epoch 155/1000
-60000/60000 - 6s - loss: 0.0656 - accuracy: 0.9835 - val_loss: 0.2996 - val_accuracy: 0.9268
-Epoch 156/1000
-60000/60000 - 6s - loss: 0.0657 - accuracy: 0.9834 - val_loss: 0.2998 - val_accuracy: 0.9275
-Epoch 157/1000
-60000/60000 - 6s - loss: 0.0653 - accuracy: 0.9834 - val_loss: 0.3028 - val_accuracy: 0.9265
-Epoch 158/1000
-60000/60000 - 6s - loss: 0.0642 - accuracy: 0.9839 - val_loss: 0.2976 - val_accuracy: 0.9274
-Epoch 159/1000
-60000/60000 - 6s - loss: 0.0631 - accuracy: 0.9843 - val_loss: 0.2982 - val_accuracy: 0.9273
-Epoch 160/1000
-60000/60000 - 6s - loss: 0.0635 - accuracy: 0.9844 - val_loss: 0.2991 - val_accuracy: 0.9262
-Epoch 161/1000
-60000/60000 - 6s - loss: 0.0629 - accuracy: 0.9845 - val_loss: 0.3048 - val_accuracy: 0.9269
-Epoch 162/1000
-60000/60000 - 6s - loss: 0.0625 - accuracy: 0.9844 - val_loss: 0.3024 - val_accuracy: 0.9274
-Epoch 163/1000
-60000/60000 - 6s - loss: 0.0637 - accuracy: 0.9841 - val_loss: 0.3001 - val_accuracy: 0.9276
-Epoch 164/1000
-60000/60000 - 6s - loss: 0.0635 - accuracy: 0.9840 - val_loss: 0.2992 - val_accuracy: 0.9263
-Epoch 165/1000
-60000/60000 - 6s - loss: 0.0638 - accuracy: 0.9840 - val_loss: 0.3004 - val_accuracy: 0.9278
-Epoch 166/1000
-60000/60000 - 6s - loss: 0.0651 - accuracy: 0.9837 - val_loss: 0.3037 - val_accuracy: 0.9273
-Epoch 167/1000
-60000/60000 - 6s - loss: 0.0641 - accuracy: 0.9838 - val_loss: 0.3037 - val_accuracy: 0.9258
-Epoch 168/1000
-60000/60000 - 6s - loss: 0.0642 - accuracy: 0.9841 - val_loss: 0.2974 - val_accuracy: 0.9275
-Epoch 169/1000
-60000/60000 - 6s - loss: 0.0633 - accuracy: 0.9841 - val_loss: 0.3025 - val_accuracy: 0.9257
-Epoch 170/1000
-60000/60000 - 6s - loss: 0.0626 - accuracy: 0.9847 - val_loss: 0.3006 - val_accuracy: 0.9276
-Epoch 171/1000
-60000/60000 - 6s - loss: 0.0637 - accuracy: 0.9843 - val_loss: 0.3007 - val_accuracy: 0.9267
-Epoch 172/1000
-60000/60000 - 6s - loss: 0.0629 - accuracy: 0.9845 - val_loss: 0.3058 - val_accuracy: 0.9252
-Epoch 173/1000
-60000/60000 - 6s - loss: 0.0619 - accuracy: 0.9849 - val_loss: 0.3020 - val_accuracy: 0.9265
-Epoch 174/1000
-60000/60000 - 6s - loss: 0.0621 - accuracy: 0.9847 - val_loss: 0.3067 - val_accuracy: 0.9263
-Epoch 175/1000
-60000/60000 - 6s - loss: 0.0620 - accuracy: 0.9847 - val_loss: 0.3031 - val_accuracy: 0.9263
-Epoch 176/1000
-60000/60000 - 6s - loss: 0.0619 - accuracy: 0.9847 - val_loss: 0.3066 - val_accuracy: 0.9261
-Epoch 177/1000
-60000/60000 - 6s - loss: 0.0626 - accuracy: 0.9843 - val_loss: 0.3066 - val_accuracy: 0.9257
-Epoch 178/1000
-60000/60000 - 6s - loss: 0.0630 - accuracy: 0.9843 - val_loss: 0.3043 - val_accuracy: 0.9266
-Epoch 179/1000
-60000/60000 - 6s - loss: 0.0632 - accuracy: 0.9844 - val_loss: 0.3052 - val_accuracy: 0.9269
-Epoch 180/1000
-60000/60000 - 6s - loss: 0.0627 - accuracy: 0.9845 - val_loss: 0.3039 - val_accuracy: 0.9268
-Epoch 181/1000
-60000/60000 - 6s - loss: 0.0635 - accuracy: 0.9839 - val_loss: 0.3086 - val_accuracy: 0.9255
-Epoch 182/1000
-60000/60000 - 6s - loss: 0.0636 - accuracy: 0.9842 - val_loss: 0.3043 - val_accuracy: 0.9267
-Epoch 183/1000
-60000/60000 - 6s - loss: 0.0634 - accuracy: 0.9842 - val_loss: 0.3033 - val_accuracy: 0.9271
-Epoch 184/1000
-60000/60000 - 6s - loss: 0.0626 - accuracy: 0.9844 - val_loss: 0.3046 - val_accuracy: 0.9259
-Epoch 185/1000
-60000/60000 - 6s - loss: 0.0627 - accuracy: 0.9844 - val_loss: 0.3046 - val_accuracy: 0.9271
-Epoch 186/1000
-60000/60000 - 6s - loss: 0.0614 - accuracy: 0.9849 - val_loss: 0.3061 - val_accuracy: 0.9259
-Epoch 187/1000
-60000/60000 - 6s - loss: 0.0601 - accuracy: 0.9853 - val_loss: 0.3072 - val_accuracy: 0.9261
-Epoch 188/1000
-60000/60000 - 6s - loss: 0.0605 - accuracy: 0.9853 - val_loss: 0.3091 - val_accuracy: 0.9259
-Epoch 189/1000
-60000/60000 - 6s - loss: 0.0617 - accuracy: 0.9850 - val_loss: 0.3081 - val_accuracy: 0.9272
-Epoch 190/1000
-60000/60000 - 6s - loss: 0.0629 - accuracy: 0.9843 - val_loss: 0.3058 - val_accuracy: 0.9266
-Epoch 191/1000
-60000/60000 - 6s - loss: 0.0629 - accuracy: 0.9846 - val_loss: 0.3077 - val_accuracy: 0.9268
-Epoch 192/1000
-60000/60000 - 6s - loss: 0.0623 - accuracy: 0.9845 - val_loss: 0.3066 - val_accuracy: 0.9262
-Epoch 193/1000
-60000/60000 - 6s - loss: 0.0623 - accuracy: 0.9846 - val_loss: 0.3098 - val_accuracy: 0.9255
-Epoch 194/1000
-60000/60000 - 6s - loss: 0.0642 - accuracy: 0.9837 - val_loss: 0.3074 - val_accuracy: 0.9263
-Epoch 195/1000
-60000/60000 - 6s - loss: 0.0658 - accuracy: 0.9834 - val_loss: 0.3087 - val_accuracy: 0.9232
-Epoch 196/1000
-60000/60000 - 6s - loss: 0.0632 - accuracy: 0.9841 - val_loss: 0.3100 - val_accuracy: 0.9260
-Epoch 197/1000
-60000/60000 - 6s - loss: 0.0607 - accuracy: 0.9850 - val_loss: 0.3098 - val_accuracy: 0.9257
-Epoch 198/1000
-60000/60000 - 6s - loss: 0.0618 - accuracy: 0.9848 - val_loss: 0.3102 - val_accuracy: 0.9263
-Epoch 199/1000
-60000/60000 - 6s - loss: 0.0607 - accuracy: 0.9850 - val_loss: 0.3096 - val_accuracy: 0.9247
-Epoch 200/1000
-60000/60000 - 6s - loss: 0.0613 - accuracy: 0.9847 - val_loss: 0.3126 - val_accuracy: 0.9255
-Epoch 201/1000
-60000/60000 - 6s - loss: 0.0621 - accuracy: 0.9845 - val_loss: 0.3082 - val_accuracy: 0.9259
-Epoch 202/1000
-60000/60000 - 6s - loss: 0.0597 - accuracy: 0.9857 - val_loss: 0.3122 - val_accuracy: 0.9252
-Epoch 203/1000
-60000/60000 - 6s - loss: 0.0599 - accuracy: 0.9853 - val_loss: 0.3105 - val_accuracy: 0.9257
-Epoch 204/1000
-60000/60000 - 6s - loss: 0.0604 - accuracy: 0.9854 - val_loss: 0.3107 - val_accuracy: 0.9262
-Epoch 205/1000
-60000/60000 - 6s - loss: 0.0598 - accuracy: 0.9853 - val_loss: 0.3140 - val_accuracy: 0.9244
-Epoch 206/1000
-60000/60000 - 6s - loss: 0.0605 - accuracy: 0.9851 - val_loss: 0.3069 - val_accuracy: 0.9255
-Epoch 207/1000
-60000/60000 - 6s - loss: 0.0605 - accuracy: 0.9851 - val_loss: 0.3125 - val_accuracy: 0.9254
-Epoch 208/1000
-60000/60000 - 6s - loss: 0.0598 - accuracy: 0.9855 - val_loss: 0.3083 - val_accuracy: 0.9267
-Epoch 209/1000
-60000/60000 - 6s - loss: 0.0591 - accuracy: 0.9855 - val_loss: 0.3121 - val_accuracy: 0.9258
-Epoch 210/1000
-60000/60000 - 6s - loss: 0.0588 - accuracy: 0.9857 - val_loss: 0.3107 - val_accuracy: 0.9253
-Epoch 211/1000
-60000/60000 - 6s - loss: 0.0591 - accuracy: 0.9854 - val_loss: 0.3168 - val_accuracy: 0.9258
-Epoch 212/1000
-60000/60000 - 6s - loss: 0.0604 - accuracy: 0.9852 - val_loss: 0.3123 - val_accuracy: 0.9264
-Epoch 213/1000
-60000/60000 - 6s - loss: 0.0607 - accuracy: 0.9849 - val_loss: 0.3122 - val_accuracy: 0.9238
-Epoch 214/1000
-60000/60000 - 6s - loss: 0.0608 - accuracy: 0.9848 - val_loss: 0.3181 - val_accuracy: 0.9239
-Epoch 215/1000
-60000/60000 - 6s - loss: 0.0598 - accuracy: 0.9854 - val_loss: 0.3142 - val_accuracy: 0.9243
-Epoch 216/1000
-60000/60000 - 6s - loss: 0.0589 - accuracy: 0.9857 - val_loss: 0.3125 - val_accuracy: 0.9254
-Epoch 217/1000
-60000/60000 - 6s - loss: 0.0583 - accuracy: 0.9857 - val_loss: 0.3144 - val_accuracy: 0.9256
-Epoch 218/1000
-60000/60000 - 6s - loss: 0.0582 - accuracy: 0.9859 - val_loss: 0.3156 - val_accuracy: 0.9249
-Epoch 219/1000
-60000/60000 - 6s - loss: 0.0594 - accuracy: 0.9852 - val_loss: 0.3129 - val_accuracy: 0.9251
-Epoch 220/1000
-60000/60000 - 6s - loss: 0.0584 - accuracy: 0.9855 - val_loss: 0.3129 - val_accuracy: 0.9251
-Epoch 221/1000
-60000/60000 - 6s - loss: 0.0608 - accuracy: 0.9847 - val_loss: 0.3138 - val_accuracy: 0.9257
-Epoch 222/1000
-60000/60000 - 6s - loss: 0.0595 - accuracy: 0.9852 - val_loss: 0.3150 - val_accuracy: 0.9247
-Epoch 223/1000
-60000/60000 - 6s - loss: 0.0584 - accuracy: 0.9857 - val_loss: 0.3139 - val_accuracy: 0.9256
-Epoch 224/1000
-60000/60000 - 6s - loss: 0.0585 - accuracy: 0.9857 - val_loss: 0.3136 - val_accuracy: 0.9245
-Epoch 225/1000
-60000/60000 - 6s - loss: 0.0586 - accuracy: 0.9855 - val_loss: 0.3174 - val_accuracy: 0.9269
-Epoch 226/1000
-60000/60000 - 6s - loss: 0.0585 - accuracy: 0.9856 - val_loss: 0.3145 - val_accuracy: 0.9253
-Epoch 227/1000
-60000/60000 - 6s - loss: 0.0584 - accuracy: 0.9855 - val_loss: 0.3181 - val_accuracy: 0.9254
-Epoch 228/1000
-60000/60000 - 6s - loss: 0.0595 - accuracy: 0.9852 - val_loss: 0.3168 - val_accuracy: 0.9249
-Epoch 229/1000
-60000/60000 - 6s - loss: 0.0600 - accuracy: 0.9850 - val_loss: 0.3135 - val_accuracy: 0.9245
-Epoch 230/1000
-60000/60000 - 6s - loss: 0.0585 - accuracy: 0.9856 - val_loss: 0.3160 - val_accuracy: 0.9252
-Epoch 231/1000
-60000/60000 - 6s - loss: 0.0584 - accuracy: 0.9855 - val_loss: 0.3157 - val_accuracy: 0.9245
-Epoch 232/1000
-60000/60000 - 6s - loss: 0.0582 - accuracy: 0.9858 - val_loss: 0.3148 - val_accuracy: 0.9255
-Epoch 233/1000
-60000/60000 - 6s - loss: 0.0575 - accuracy: 0.9857 - val_loss: 0.3136 - val_accuracy: 0.9261
-Epoch 234/1000
-60000/60000 - 6s - loss: 0.0578 - accuracy: 0.9858 - val_loss: 0.3157 - val_accuracy: 0.9248
-Epoch 235/1000
-60000/60000 - 6s - loss: 0.0579 - accuracy: 0.9856 - val_loss: 0.3177 - val_accuracy: 0.9243
-Epoch 236/1000
-60000/60000 - 6s - loss: 0.0584 - accuracy: 0.9857 - val_loss: 0.3176 - val_accuracy: 0.9253
-Epoch 237/1000
-60000/60000 - 6s - loss: 0.0579 - accuracy: 0.9859 - val_loss: 0.3165 - val_accuracy: 0.9253
-Epoch 238/1000
-60000/60000 - 6s - loss: 0.0583 - accuracy: 0.9857 - val_loss: 0.3145 - val_accuracy: 0.9248
-Epoch 239/1000
-60000/60000 - 6s - loss: 0.0570 - accuracy: 0.9860 - val_loss: 0.3176 - val_accuracy: 0.9256
-Epoch 240/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9861 - val_loss: 0.3159 - val_accuracy: 0.9253
-Epoch 241/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9862 - val_loss: 0.3189 - val_accuracy: 0.9248
-Epoch 242/1000
-60000/60000 - 6s - loss: 0.0571 - accuracy: 0.9859 - val_loss: 0.3206 - val_accuracy: 0.9251
-Epoch 243/1000
-60000/60000 - 6s - loss: 0.0568 - accuracy: 0.9863 - val_loss: 0.3202 - val_accuracy: 0.9256
-Epoch 244/1000
-60000/60000 - 6s - loss: 0.0573 - accuracy: 0.9858 - val_loss: 0.3196 - val_accuracy: 0.9236
-Epoch 245/1000
-60000/60000 - 6s - loss: 0.0586 - accuracy: 0.9856 - val_loss: 0.3215 - val_accuracy: 0.9248
-Epoch 246/1000
-60000/60000 - 6s - loss: 0.0591 - accuracy: 0.9854 - val_loss: 0.3195 - val_accuracy: 0.9254
-Epoch 247/1000
-60000/60000 - 6s - loss: 0.0594 - accuracy: 0.9851 - val_loss: 0.3231 - val_accuracy: 0.9238
-Epoch 248/1000
-60000/60000 - 6s - loss: 0.0585 - accuracy: 0.9854 - val_loss: 0.3188 - val_accuracy: 0.9251
-Epoch 249/1000
-60000/60000 - 6s - loss: 0.0572 - accuracy: 0.9860 - val_loss: 0.3200 - val_accuracy: 0.9252
-Epoch 250/1000
-60000/60000 - 6s - loss: 0.0571 - accuracy: 0.9861 - val_loss: 0.3196 - val_accuracy: 0.9262
-Epoch 251/1000
-60000/60000 - 6s - loss: 0.0572 - accuracy: 0.9860 - val_loss: 0.3213 - val_accuracy: 0.9256
-Epoch 252/1000
-60000/60000 - 6s - loss: 0.0570 - accuracy: 0.9860 - val_loss: 0.3235 - val_accuracy: 0.9245
-Epoch 253/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9861 - val_loss: 0.3225 - val_accuracy: 0.9242
-Epoch 254/1000
-60000/60000 - 6s - loss: 0.0565 - accuracy: 0.9862 - val_loss: 0.3200 - val_accuracy: 0.9260
-Epoch 255/1000
-60000/60000 - 6s - loss: 0.0565 - accuracy: 0.9862 - val_loss: 0.3185 - val_accuracy: 0.9251
-Epoch 256/1000
-60000/60000 - 6s - loss: 0.0574 - accuracy: 0.9860 - val_loss: 0.3200 - val_accuracy: 0.9252
-Epoch 257/1000
-60000/60000 - 6s - loss: 0.0571 - accuracy: 0.9858 - val_loss: 0.3198 - val_accuracy: 0.9249
-Epoch 258/1000
-60000/60000 - 6s - loss: 0.0567 - accuracy: 0.9861 - val_loss: 0.3219 - val_accuracy: 0.9245
-Epoch 259/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9859 - val_loss: 0.3198 - val_accuracy: 0.9247
-Epoch 260/1000
-60000/60000 - 6s - loss: 0.0576 - accuracy: 0.9857 - val_loss: 0.3171 - val_accuracy: 0.9267
-Epoch 261/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9859 - val_loss: 0.3200 - val_accuracy: 0.9246
-Epoch 262/1000
-60000/60000 - 6s - loss: 0.0564 - accuracy: 0.9862 - val_loss: 0.3210 - val_accuracy: 0.9249
-Epoch 263/1000
-60000/60000 - 6s - loss: 0.0562 - accuracy: 0.9862 - val_loss: 0.3225 - val_accuracy: 0.9251
-Epoch 264/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9860 - val_loss: 0.3227 - val_accuracy: 0.9258
-Epoch 265/1000
-60000/60000 - 6s - loss: 0.0562 - accuracy: 0.9863 - val_loss: 0.3234 - val_accuracy: 0.9249
-Epoch 266/1000
-60000/60000 - 6s - loss: 0.0560 - accuracy: 0.9863 - val_loss: 0.3227 - val_accuracy: 0.9248
-Epoch 267/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9860 - val_loss: 0.3230 - val_accuracy: 0.9259
-Epoch 268/1000
-60000/60000 - 6s - loss: 0.0576 - accuracy: 0.9857 - val_loss: 0.3244 - val_accuracy: 0.9249
-Epoch 269/1000
-60000/60000 - 6s - loss: 0.0576 - accuracy: 0.9858 - val_loss: 0.3228 - val_accuracy: 0.9251
-Epoch 270/1000
-60000/60000 - 6s - loss: 0.0568 - accuracy: 0.9859 - val_loss: 0.3252 - val_accuracy: 0.9242
-Epoch 271/1000
-60000/60000 - 6s - loss: 0.0572 - accuracy: 0.9860 - val_loss: 0.3265 - val_accuracy: 0.9249
-Epoch 272/1000
-60000/60000 - 6s - loss: 0.0562 - accuracy: 0.9862 - val_loss: 0.3261 - val_accuracy: 0.9250
-Epoch 273/1000
-60000/60000 - 6s - loss: 0.0568 - accuracy: 0.9858 - val_loss: 0.3264 - val_accuracy: 0.9248
-Epoch 274/1000
-60000/60000 - 6s - loss: 0.0565 - accuracy: 0.9860 - val_loss: 0.3256 - val_accuracy: 0.9253
-Epoch 275/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9863 - val_loss: 0.3231 - val_accuracy: 0.9247
-Epoch 276/1000
-60000/60000 - 6s - loss: 0.0565 - accuracy: 0.9858 - val_loss: 0.3254 - val_accuracy: 0.9245
-Epoch 277/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9862 - val_loss: 0.3261 - val_accuracy: 0.9251
-Epoch 278/1000
-60000/60000 - 6s - loss: 0.0560 - accuracy: 0.9862 - val_loss: 0.3263 - val_accuracy: 0.9246
-Epoch 279/1000
-60000/60000 - 6s - loss: 0.0556 - accuracy: 0.9863 - val_loss: 0.3246 - val_accuracy: 0.9246
-Epoch 280/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9864 - val_loss: 0.3255 - val_accuracy: 0.9247
-Epoch 281/1000
-60000/60000 - 6s - loss: 0.0555 - accuracy: 0.9863 - val_loss: 0.3264 - val_accuracy: 0.9252
-Epoch 282/1000
-60000/60000 - 6s - loss: 0.0566 - accuracy: 0.9859 - val_loss: 0.3253 - val_accuracy: 0.9244
-Epoch 283/1000
-60000/60000 - 6s - loss: 0.0565 - accuracy: 0.9861 - val_loss: 0.3247 - val_accuracy: 0.9249
-Epoch 284/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9862 - val_loss: 0.3228 - val_accuracy: 0.9251
-Epoch 285/1000
-60000/60000 - 6s - loss: 0.0557 - accuracy: 0.9863 - val_loss: 0.3253 - val_accuracy: 0.9248
-Epoch 286/1000
-60000/60000 - 6s - loss: 0.0558 - accuracy: 0.9863 - val_loss: 0.3271 - val_accuracy: 0.9253
-Epoch 287/1000
-60000/60000 - 6s - loss: 0.0550 - accuracy: 0.9864 - val_loss: 0.3250 - val_accuracy: 0.9249
-Epoch 288/1000
-60000/60000 - 6s - loss: 0.0555 - accuracy: 0.9865 - val_loss: 0.3314 - val_accuracy: 0.9247
-Epoch 289/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9861 - val_loss: 0.3243 - val_accuracy: 0.9252
-Epoch 290/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9862 - val_loss: 0.3262 - val_accuracy: 0.9243
-Epoch 291/1000
-60000/60000 - 6s - loss: 0.0556 - accuracy: 0.9863 - val_loss: 0.3280 - val_accuracy: 0.9244
-Epoch 292/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9861 - val_loss: 0.3276 - val_accuracy: 0.9255
-Epoch 293/1000
-60000/60000 - 6s - loss: 0.0574 - accuracy: 0.9855 - val_loss: 0.3280 - val_accuracy: 0.9245
-Epoch 294/1000
-60000/60000 - 6s - loss: 0.0568 - accuracy: 0.9861 - val_loss: 0.3277 - val_accuracy: 0.9255
-Epoch 295/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9863 - val_loss: 0.3254 - val_accuracy: 0.9253
-Epoch 296/1000
-60000/60000 - 6s - loss: 0.0549 - accuracy: 0.9865 - val_loss: 0.3270 - val_accuracy: 0.9257
-Epoch 297/1000
-60000/60000 - 6s - loss: 0.0555 - accuracy: 0.9864 - val_loss: 0.3242 - val_accuracy: 0.9247
-Epoch 298/1000
-60000/60000 - 6s - loss: 0.0564 - accuracy: 0.9862 - val_loss: 0.3279 - val_accuracy: 0.9253
-Epoch 299/1000
-60000/60000 - 6s - loss: 0.0562 - accuracy: 0.9861 - val_loss: 0.3262 - val_accuracy: 0.9256
-Epoch 300/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9861 - val_loss: 0.3265 - val_accuracy: 0.9235
-Epoch 301/1000
-60000/60000 - 6s - loss: 0.0561 - accuracy: 0.9863 - val_loss: 0.3274 - val_accuracy: 0.9256
-Epoch 302/1000
-60000/60000 - 6s - loss: 0.0557 - accuracy: 0.9863 - val_loss: 0.3270 - val_accuracy: 0.9255
-Epoch 303/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9859 - val_loss: 0.3301 - val_accuracy: 0.9235
-Epoch 304/1000
-60000/60000 - 6s - loss: 0.0558 - accuracy: 0.9863 - val_loss: 0.3257 - val_accuracy: 0.9256
-Epoch 305/1000
-60000/60000 - 6s - loss: 0.0550 - accuracy: 0.9865 - val_loss: 0.3266 - val_accuracy: 0.9256
-Epoch 306/1000
-60000/60000 - 6s - loss: 0.0550 - accuracy: 0.9864 - val_loss: 0.3289 - val_accuracy: 0.9262
-Epoch 307/1000
-60000/60000 - 6s - loss: 0.0553 - accuracy: 0.9863 - val_loss: 0.3283 - val_accuracy: 0.9252
-Epoch 308/1000
-60000/60000 - 6s - loss: 0.0548 - accuracy: 0.9866 - val_loss: 0.3268 - val_accuracy: 0.9256
-Epoch 309/1000
-60000/60000 - 6s - loss: 0.0555 - accuracy: 0.9862 - val_loss: 0.3294 - val_accuracy: 0.9242
-Epoch 310/1000
-60000/60000 - 6s - loss: 0.0599 - accuracy: 0.9847 - val_loss: 0.3250 - val_accuracy: 0.9247
-Epoch 311/1000
-60000/60000 - 6s - loss: 0.0588 - accuracy: 0.9851 - val_loss: 0.3250 - val_accuracy: 0.9256
-Epoch 312/1000
-60000/60000 - 6s - loss: 0.0592 - accuracy: 0.9850 - val_loss: 0.3266 - val_accuracy: 0.9247
-Epoch 313/1000
-60000/60000 - 6s - loss: 0.0610 - accuracy: 0.9842 - val_loss: 0.3300 - val_accuracy: 0.9237
-Epoch 314/1000
-60000/60000 - 6s - loss: 0.0605 - accuracy: 0.9847 - val_loss: 0.3267 - val_accuracy: 0.9251
-Epoch 315/1000
-60000/60000 - 6s - loss: 0.0571 - accuracy: 0.9858 - val_loss: 0.3278 - val_accuracy: 0.9250
-Epoch 316/1000
-60000/60000 - 6s - loss: 0.0571 - accuracy: 0.9858 - val_loss: 0.3336 - val_accuracy: 0.9247
-Epoch 317/1000
-60000/60000 - 6s - loss: 0.0575 - accuracy: 0.9857 - val_loss: 0.3312 - val_accuracy: 0.9246
-Epoch 318/1000
-60000/60000 - 6s - loss: 0.0568 - accuracy: 0.9859 - val_loss: 0.3296 - val_accuracy: 0.9247
-Epoch 319/1000
-60000/60000 - 6s - loss: 0.0572 - accuracy: 0.9855 - val_loss: 0.3296 - val_accuracy: 0.9249
-Epoch 320/1000
-60000/60000 - 6s - loss: 0.0560 - accuracy: 0.9863 - val_loss: 0.3285 - val_accuracy: 0.9252
-Epoch 321/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9863 - val_loss: 0.3331 - val_accuracy: 0.9243
-Epoch 322/1000
-60000/60000 - 6s - loss: 0.0557 - accuracy: 0.9863 - val_loss: 0.3317 - val_accuracy: 0.9238
-Epoch 323/1000
-60000/60000 - 6s - loss: 0.0547 - accuracy: 0.9868 - val_loss: 0.3371 - val_accuracy: 0.9235
-Epoch 324/1000
-60000/60000 - 6s - loss: 0.0554 - accuracy: 0.9863 - val_loss: 0.3325 - val_accuracy: 0.9244
-Epoch 325/1000
-60000/60000 - 6s - loss: 0.0545 - accuracy: 0.9866 - val_loss: 0.3315 - val_accuracy: 0.9251
-Epoch 326/1000
-60000/60000 - 6s - loss: 0.0552 - accuracy: 0.9863 - val_loss: 0.3324 - val_accuracy: 0.9248
-Epoch 327/1000
-60000/60000 - 6s - loss: 0.0548 - accuracy: 0.9866 - val_loss: 0.3296 - val_accuracy: 0.9254
-Epoch 328/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9861 - val_loss: 0.3355 - val_accuracy: 0.9229
-Epoch 329/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9858 - val_loss: 0.3302 - val_accuracy: 0.9243
-Epoch 330/1000
-60000/60000 - 6s - loss: 0.0551 - accuracy: 0.9866 - val_loss: 0.3332 - val_accuracy: 0.9246
-Epoch 331/1000
-60000/60000 - 6s - loss: 0.0546 - accuracy: 0.9866 - val_loss: 0.3318 - val_accuracy: 0.9252
-Epoch 332/1000
-60000/60000 - 6s - loss: 0.0545 - accuracy: 0.9864 - val_loss: 0.3340 - val_accuracy: 0.9241
-Epoch 333/1000
-60000/60000 - 6s - loss: 0.0545 - accuracy: 0.9866 - val_loss: 0.3352 - val_accuracy: 0.9246
-Epoch 334/1000
-60000/60000 - 6s - loss: 0.0576 - accuracy: 0.9856 - val_loss: 0.3350 - val_accuracy: 0.9244
-Epoch 335/1000
-60000/60000 - 6s - loss: 0.0567 - accuracy: 0.9859 - val_loss: 0.3326 - val_accuracy: 0.9243
-Epoch 336/1000
-60000/60000 - 6s - loss: 0.0564 - accuracy: 0.9860 - val_loss: 0.3320 - val_accuracy: 0.9247
-Epoch 337/1000
-60000/60000 - 6s - loss: 0.0554 - accuracy: 0.9865 - val_loss: 0.3330 - val_accuracy: 0.9244
-Epoch 338/1000
-60000/60000 - 6s - loss: 0.0547 - accuracy: 0.9866 - val_loss: 0.3327 - val_accuracy: 0.9240
-Epoch 339/1000
-60000/60000 - 6s - loss: 0.0541 - accuracy: 0.9867 - val_loss: 0.3324 - val_accuracy: 0.9253
-Epoch 340/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9865 - val_loss: 0.3344 - val_accuracy: 0.9233
-Epoch 341/1000
-60000/60000 - 6s - loss: 0.0545 - accuracy: 0.9866 - val_loss: 0.3344 - val_accuracy: 0.9247
-Epoch 342/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9868 - val_loss: 0.3348 - val_accuracy: 0.9236
-Epoch 343/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9867 - val_loss: 0.3358 - val_accuracy: 0.9251
-Epoch 344/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9866 - val_loss: 0.3322 - val_accuracy: 0.9246
-Epoch 345/1000
-60000/60000 - 6s - loss: 0.0543 - accuracy: 0.9865 - val_loss: 0.3340 - val_accuracy: 0.9242
-Epoch 346/1000
-60000/60000 - 6s - loss: 0.0539 - accuracy: 0.9867 - val_loss: 0.3320 - val_accuracy: 0.9248
-Epoch 347/1000
-60000/60000 - 6s - loss: 0.0542 - accuracy: 0.9866 - val_loss: 0.3342 - val_accuracy: 0.9245
-Epoch 348/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9868 - val_loss: 0.3333 - val_accuracy: 0.9250
-Epoch 349/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9869 - val_loss: 0.3329 - val_accuracy: 0.9247
-Epoch 350/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9870 - val_loss: 0.3331 - val_accuracy: 0.9251
-Epoch 351/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9865 - val_loss: 0.3363 - val_accuracy: 0.9262
-Epoch 352/1000
-60000/60000 - 6s - loss: 0.0550 - accuracy: 0.9862 - val_loss: 0.3365 - val_accuracy: 0.9253
-Epoch 353/1000
-60000/60000 - 6s - loss: 0.0548 - accuracy: 0.9864 - val_loss: 0.3340 - val_accuracy: 0.9245
-Epoch 354/1000
-60000/60000 - 6s - loss: 0.0541 - accuracy: 0.9867 - val_loss: 0.3340 - val_accuracy: 0.9246
-Epoch 355/1000
-60000/60000 - 6s - loss: 0.0541 - accuracy: 0.9867 - val_loss: 0.3349 - val_accuracy: 0.9244
-Epoch 356/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9864 - val_loss: 0.3352 - val_accuracy: 0.9252
-Epoch 357/1000
-60000/60000 - 6s - loss: 0.0537 - accuracy: 0.9868 - val_loss: 0.3380 - val_accuracy: 0.9240
-Epoch 358/1000
-60000/60000 - 6s - loss: 0.0541 - accuracy: 0.9865 - val_loss: 0.3359 - val_accuracy: 0.9246
-Epoch 359/1000
-60000/60000 - 6s - loss: 0.0535 - accuracy: 0.9869 - val_loss: 0.3340 - val_accuracy: 0.9257
-Epoch 360/1000
-60000/60000 - 6s - loss: 0.0530 - accuracy: 0.9869 - val_loss: 0.3349 - val_accuracy: 0.9251
-Epoch 361/1000
-60000/60000 - 6s - loss: 0.0529 - accuracy: 0.9868 - val_loss: 0.3350 - val_accuracy: 0.9250
-Epoch 362/1000
-60000/60000 - 6s - loss: 0.0539 - accuracy: 0.9865 - val_loss: 0.3344 - val_accuracy: 0.9252
-Epoch 363/1000
-60000/60000 - 6s - loss: 0.0542 - accuracy: 0.9866 - val_loss: 0.3362 - val_accuracy: 0.9240
-Epoch 364/1000
-60000/60000 - 6s - loss: 0.0545 - accuracy: 0.9866 - val_loss: 0.3342 - val_accuracy: 0.9248
-Epoch 365/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9864 - val_loss: 0.3347 - val_accuracy: 0.9246
-Epoch 366/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9869 - val_loss: 0.3393 - val_accuracy: 0.9237
-Epoch 367/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9869 - val_loss: 0.3350 - val_accuracy: 0.9244
-Epoch 368/1000
-60000/60000 - 6s - loss: 0.0530 - accuracy: 0.9870 - val_loss: 0.3362 - val_accuracy: 0.9253
-Epoch 369/1000
-60000/60000 - 6s - loss: 0.0527 - accuracy: 0.9870 - val_loss: 0.3372 - val_accuracy: 0.9246
-Epoch 370/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9867 - val_loss: 0.3359 - val_accuracy: 0.9250
-Epoch 371/1000
-60000/60000 - 6s - loss: 0.0537 - accuracy: 0.9867 - val_loss: 0.3392 - val_accuracy: 0.9247
-Epoch 372/1000
-60000/60000 - 6s - loss: 0.0529 - accuracy: 0.9869 - val_loss: 0.3356 - val_accuracy: 0.9235
-Epoch 373/1000
-60000/60000 - 6s - loss: 0.0532 - accuracy: 0.9869 - val_loss: 0.3388 - val_accuracy: 0.9241
-Epoch 374/1000
-60000/60000 - 6s - loss: 0.0530 - accuracy: 0.9870 - val_loss: 0.3396 - val_accuracy: 0.9234
-Epoch 375/1000
-60000/60000 - 6s - loss: 0.0542 - accuracy: 0.9862 - val_loss: 0.3385 - val_accuracy: 0.9238
-Epoch 376/1000
-60000/60000 - 6s - loss: 0.0559 - accuracy: 0.9858 - val_loss: 0.3359 - val_accuracy: 0.9241
-Epoch 377/1000
-60000/60000 - 6s - loss: 0.0562 - accuracy: 0.9858 - val_loss: 0.3378 - val_accuracy: 0.9244
-Epoch 378/1000
-60000/60000 - 6s - loss: 0.0573 - accuracy: 0.9856 - val_loss: 0.3377 - val_accuracy: 0.9248
-Epoch 379/1000
-60000/60000 - 6s - loss: 0.0566 - accuracy: 0.9858 - val_loss: 0.3371 - val_accuracy: 0.9238
-Epoch 380/1000
-60000/60000 - 6s - loss: 0.0552 - accuracy: 0.9862 - val_loss: 0.3357 - val_accuracy: 0.9250
-Epoch 381/1000
-60000/60000 - 6s - loss: 0.0564 - accuracy: 0.9858 - val_loss: 0.3359 - val_accuracy: 0.9262
-Epoch 382/1000
-60000/60000 - 6s - loss: 0.0554 - accuracy: 0.9863 - val_loss: 0.3360 - val_accuracy: 0.9236
-Epoch 383/1000
-60000/60000 - 6s - loss: 0.0546 - accuracy: 0.9865 - val_loss: 0.3351 - val_accuracy: 0.9249
-Epoch 384/1000
-60000/60000 - 6s - loss: 0.0543 - accuracy: 0.9867 - val_loss: 0.3375 - val_accuracy: 0.9253
-Epoch 385/1000
-60000/60000 - 6s - loss: 0.0534 - accuracy: 0.9868 - val_loss: 0.3374 - val_accuracy: 0.9247
-Epoch 386/1000
-60000/60000 - 6s - loss: 0.0529 - accuracy: 0.9870 - val_loss: 0.3376 - val_accuracy: 0.9255
-Epoch 387/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9869 - val_loss: 0.3383 - val_accuracy: 0.9249
-Epoch 388/1000
-60000/60000 - 6s - loss: 0.0537 - accuracy: 0.9867 - val_loss: 0.3391 - val_accuracy: 0.9246
-Epoch 389/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9866 - val_loss: 0.3361 - val_accuracy: 0.9250
-Epoch 390/1000
-60000/60000 - 6s - loss: 0.0539 - accuracy: 0.9866 - val_loss: 0.3396 - val_accuracy: 0.9248
-Epoch 391/1000
-60000/60000 - 6s - loss: 0.0529 - accuracy: 0.9870 - val_loss: 0.3405 - val_accuracy: 0.9249
-Epoch 392/1000
-60000/60000 - 6s - loss: 0.0532 - accuracy: 0.9870 - val_loss: 0.3386 - val_accuracy: 0.9247
-Epoch 393/1000
-60000/60000 - 6s - loss: 0.0527 - accuracy: 0.9871 - val_loss: 0.3389 - val_accuracy: 0.9240
-Epoch 394/1000
-60000/60000 - 6s - loss: 0.0525 - accuracy: 0.9872 - val_loss: 0.3380 - val_accuracy: 0.9249
-Epoch 395/1000
-60000/60000 - 6s - loss: 0.0525 - accuracy: 0.9871 - val_loss: 0.3439 - val_accuracy: 0.9246
-Epoch 396/1000
-60000/60000 - 6s - loss: 0.0534 - accuracy: 0.9867 - val_loss: 0.3405 - val_accuracy: 0.9249
-Epoch 397/1000
-60000/60000 - 6s - loss: 0.0528 - accuracy: 0.9870 - val_loss: 0.3391 - val_accuracy: 0.9249
-Epoch 398/1000
-60000/60000 - 6s - loss: 0.0525 - accuracy: 0.9871 - val_loss: 0.3398 - val_accuracy: 0.9241
-Epoch 399/1000
-60000/60000 - 6s - loss: 0.0532 - accuracy: 0.9868 - val_loss: 0.3408 - val_accuracy: 0.9241
-Epoch 400/1000
-60000/60000 - 6s - loss: 0.0547 - accuracy: 0.9865 - val_loss: 0.3399 - val_accuracy: 0.9249
-Epoch 401/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9866 - val_loss: 0.3400 - val_accuracy: 0.9240
-Epoch 402/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9869 - val_loss: 0.3427 - val_accuracy: 0.9229
-Epoch 403/1000
-60000/60000 - 6s - loss: 0.0530 - accuracy: 0.9870 - val_loss: 0.3412 - val_accuracy: 0.9238
-Epoch 404/1000
-60000/60000 - 6s - loss: 0.0529 - accuracy: 0.9870 - val_loss: 0.3415 - val_accuracy: 0.9242
-Epoch 405/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9869 - val_loss: 0.3398 - val_accuracy: 0.9245
-Epoch 406/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9865 - val_loss: 0.3401 - val_accuracy: 0.9240
-Epoch 407/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9869 - val_loss: 0.3415 - val_accuracy: 0.9248
-Epoch 408/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9870 - val_loss: 0.3420 - val_accuracy: 0.9231
-Epoch 409/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9870 - val_loss: 0.3393 - val_accuracy: 0.9242
-Epoch 410/1000
-60000/60000 - 6s - loss: 0.0539 - accuracy: 0.9866 - val_loss: 0.3405 - val_accuracy: 0.9243
-Epoch 411/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9872 - val_loss: 0.3390 - val_accuracy: 0.9246
-Epoch 412/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9872 - val_loss: 0.3382 - val_accuracy: 0.9247
-Epoch 413/1000
-60000/60000 - 6s - loss: 0.0525 - accuracy: 0.9869 - val_loss: 0.3407 - val_accuracy: 0.9238
-Epoch 414/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9872 - val_loss: 0.3414 - val_accuracy: 0.9237
-Epoch 415/1000
-60000/60000 - 6s - loss: 0.0523 - accuracy: 0.9872 - val_loss: 0.3410 - val_accuracy: 0.9245
-Epoch 416/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9868 - val_loss: 0.3409 - val_accuracy: 0.9250
-Epoch 417/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9867 - val_loss: 0.3382 - val_accuracy: 0.9239
-Epoch 418/1000
-60000/60000 - 6s - loss: 0.0537 - accuracy: 0.9867 - val_loss: 0.3415 - val_accuracy: 0.9231
-Epoch 419/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9868 - val_loss: 0.3411 - val_accuracy: 0.9244
-Epoch 420/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9867 - val_loss: 0.3413 - val_accuracy: 0.9231
-Epoch 421/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9867 - val_loss: 0.3422 - val_accuracy: 0.9235
-Epoch 422/1000
-60000/60000 - 6s - loss: 0.0527 - accuracy: 0.9871 - val_loss: 0.3413 - val_accuracy: 0.9241
-Epoch 423/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9870 - val_loss: 0.3403 - val_accuracy: 0.9233
-Epoch 424/1000
-60000/60000 - 6s - loss: 0.0525 - accuracy: 0.9869 - val_loss: 0.3391 - val_accuracy: 0.9246
-Epoch 425/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9870 - val_loss: 0.3408 - val_accuracy: 0.9243
-Epoch 426/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9871 - val_loss: 0.3417 - val_accuracy: 0.9251
-Epoch 427/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9865 - val_loss: 0.3431 - val_accuracy: 0.9237
-Epoch 428/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9870 - val_loss: 0.3412 - val_accuracy: 0.9246
-Epoch 429/1000
-60000/60000 - 6s - loss: 0.0526 - accuracy: 0.9869 - val_loss: 0.3420 - val_accuracy: 0.9242
-Epoch 430/1000
-60000/60000 - 6s - loss: 0.0528 - accuracy: 0.9871 - val_loss: 0.3434 - val_accuracy: 0.9227
-Epoch 431/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9869 - val_loss: 0.3430 - val_accuracy: 0.9244
-Epoch 432/1000
-60000/60000 - 6s - loss: 0.0544 - accuracy: 0.9863 - val_loss: 0.3432 - val_accuracy: 0.9230
-Epoch 433/1000
-60000/60000 - 6s - loss: 0.0543 - accuracy: 0.9863 - val_loss: 0.3422 - val_accuracy: 0.9244
-Epoch 434/1000
-60000/60000 - 6s - loss: 0.0527 - accuracy: 0.9872 - val_loss: 0.3394 - val_accuracy: 0.9236
-Epoch 435/1000
-60000/60000 - 6s - loss: 0.0526 - accuracy: 0.9870 - val_loss: 0.3426 - val_accuracy: 0.9245
-Epoch 436/1000
-60000/60000 - 6s - loss: 0.0548 - accuracy: 0.9863 - val_loss: 0.3410 - val_accuracy: 0.9245
-Epoch 437/1000
-60000/60000 - 6s - loss: 0.0546 - accuracy: 0.9865 - val_loss: 0.3423 - val_accuracy: 0.9243
-Epoch 438/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9866 - val_loss: 0.3421 - val_accuracy: 0.9231
-Epoch 439/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9872 - val_loss: 0.3425 - val_accuracy: 0.9231
-Epoch 440/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9874 - val_loss: 0.3426 - val_accuracy: 0.9238
-Epoch 441/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9872 - val_loss: 0.3417 - val_accuracy: 0.9233
-Epoch 442/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9871 - val_loss: 0.3428 - val_accuracy: 0.9241
-Epoch 443/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9872 - val_loss: 0.3415 - val_accuracy: 0.9247
-Epoch 444/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9871 - val_loss: 0.3447 - val_accuracy: 0.9236
-Epoch 445/1000
-60000/60000 - 6s - loss: 0.0528 - accuracy: 0.9871 - val_loss: 0.3442 - val_accuracy: 0.9241
-Epoch 446/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9873 - val_loss: 0.3443 - val_accuracy: 0.9236
-Epoch 447/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9872 - val_loss: 0.3425 - val_accuracy: 0.9229
-Epoch 448/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9873 - val_loss: 0.3450 - val_accuracy: 0.9230
-Epoch 449/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9873 - val_loss: 0.3446 - val_accuracy: 0.9237
-Epoch 450/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9873 - val_loss: 0.3411 - val_accuracy: 0.9236
-Epoch 451/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9871 - val_loss: 0.3418 - val_accuracy: 0.9235
-Epoch 452/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9870 - val_loss: 0.3397 - val_accuracy: 0.9238
-Epoch 453/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9870 - val_loss: 0.3411 - val_accuracy: 0.9236
-Epoch 454/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9866 - val_loss: 0.3389 - val_accuracy: 0.9236
-Epoch 455/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9870 - val_loss: 0.3419 - val_accuracy: 0.9243
-Epoch 456/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9872 - val_loss: 0.3401 - val_accuracy: 0.9237
-Epoch 457/1000
-60000/60000 - 6s - loss: 0.0548 - accuracy: 0.9862 - val_loss: 0.3403 - val_accuracy: 0.9242
-Epoch 458/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9865 - val_loss: 0.3420 - val_accuracy: 0.9247
-Epoch 459/1000
-60000/60000 - 6s - loss: 0.0528 - accuracy: 0.9869 - val_loss: 0.3411 - val_accuracy: 0.9234
-Epoch 460/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9872 - val_loss: 0.3414 - val_accuracy: 0.9241
-Epoch 461/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9873 - val_loss: 0.3426 - val_accuracy: 0.9242
-Epoch 462/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9873 - val_loss: 0.3429 - val_accuracy: 0.9243
-Epoch 463/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9873 - val_loss: 0.3425 - val_accuracy: 0.9245
-Epoch 464/1000
-60000/60000 - 6s - loss: 0.0523 - accuracy: 0.9872 - val_loss: 0.3416 - val_accuracy: 0.9240
-Epoch 465/1000
-60000/60000 - 6s - loss: 0.0527 - accuracy: 0.9869 - val_loss: 0.3426 - val_accuracy: 0.9239
-Epoch 466/1000
-60000/60000 - 6s - loss: 0.0526 - accuracy: 0.9871 - val_loss: 0.3424 - val_accuracy: 0.9234
-Epoch 467/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9873 - val_loss: 0.3426 - val_accuracy: 0.9229
-Epoch 468/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9873 - val_loss: 0.3435 - val_accuracy: 0.9233
-Epoch 469/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9873 - val_loss: 0.3436 - val_accuracy: 0.9233
-Epoch 470/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9872 - val_loss: 0.3448 - val_accuracy: 0.9242
-Epoch 471/1000
-60000/60000 - 6s - loss: 0.0515 - accuracy: 0.9873 - val_loss: 0.3414 - val_accuracy: 0.9236
-Epoch 472/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9874 - val_loss: 0.3447 - val_accuracy: 0.9241
-Epoch 473/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9866 - val_loss: 0.3395 - val_accuracy: 0.9245
-Epoch 474/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9870 - val_loss: 0.3406 - val_accuracy: 0.9242
-Epoch 475/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9874 - val_loss: 0.3437 - val_accuracy: 0.9242
-Epoch 476/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9874 - val_loss: 0.3447 - val_accuracy: 0.9249
-Epoch 477/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9873 - val_loss: 0.3444 - val_accuracy: 0.9240
-Epoch 478/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9869 - val_loss: 0.3440 - val_accuracy: 0.9234
-Epoch 479/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9874 - val_loss: 0.3462 - val_accuracy: 0.9232
-Epoch 480/1000
-60000/60000 - 6s - loss: 0.0511 - accuracy: 0.9874 - val_loss: 0.3437 - val_accuracy: 0.9246
-Epoch 481/1000
-60000/60000 - 6s - loss: 0.0512 - accuracy: 0.9875 - val_loss: 0.3472 - val_accuracy: 0.9236
-Epoch 482/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9870 - val_loss: 0.3437 - val_accuracy: 0.9236
-Epoch 483/1000
-60000/60000 - 6s - loss: 0.0515 - accuracy: 0.9872 - val_loss: 0.3456 - val_accuracy: 0.9235
-Epoch 484/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9873 - val_loss: 0.3455 - val_accuracy: 0.9233
-Epoch 485/1000
-60000/60000 - 6s - loss: 0.0509 - accuracy: 0.9874 - val_loss: 0.3447 - val_accuracy: 0.9231
-Epoch 486/1000
-60000/60000 - 6s - loss: 0.0512 - accuracy: 0.9872 - val_loss: 0.3447 - val_accuracy: 0.9235
-Epoch 487/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9873 - val_loss: 0.3461 - val_accuracy: 0.9227
-Epoch 488/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9872 - val_loss: 0.3453 - val_accuracy: 0.9247
-Epoch 489/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9875 - val_loss: 0.3459 - val_accuracy: 0.9240
-Epoch 490/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9873 - val_loss: 0.3453 - val_accuracy: 0.9254
-Epoch 491/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9871 - val_loss: 0.3473 - val_accuracy: 0.9246
-Epoch 492/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9872 - val_loss: 0.3445 - val_accuracy: 0.9248
-Epoch 493/1000
-60000/60000 - 6s - loss: 0.0515 - accuracy: 0.9874 - val_loss: 0.3458 - val_accuracy: 0.9242
-Epoch 494/1000
-60000/60000 - 6s - loss: 0.0519 - accuracy: 0.9871 - val_loss: 0.3435 - val_accuracy: 0.9249
-Epoch 495/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9872 - val_loss: 0.3432 - val_accuracy: 0.9240
-Epoch 496/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9872 - val_loss: 0.3446 - val_accuracy: 0.9244
-Epoch 497/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9874 - val_loss: 0.3452 - val_accuracy: 0.9239
-Epoch 498/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9873 - val_loss: 0.3463 - val_accuracy: 0.9245
-Epoch 499/1000
-60000/60000 - 6s - loss: 0.0512 - accuracy: 0.9873 - val_loss: 0.3443 - val_accuracy: 0.9240
-Epoch 500/1000
-60000/60000 - 6s - loss: 0.0511 - accuracy: 0.9874 - val_loss: 0.3439 - val_accuracy: 0.9233
-Epoch 501/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9868 - val_loss: 0.3447 - val_accuracy: 0.9243
-Epoch 502/1000
-60000/60000 - 6s - loss: 0.0526 - accuracy: 0.9869 - val_loss: 0.3447 - val_accuracy: 0.9245
-Epoch 503/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3462 - val_accuracy: 0.9236
-Epoch 504/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3452 - val_accuracy: 0.9242
-Epoch 505/1000
-60000/60000 - 6s - loss: 0.0530 - accuracy: 0.9866 - val_loss: 0.3445 - val_accuracy: 0.9244
-Epoch 506/1000
-60000/60000 - 6s - loss: 0.0535 - accuracy: 0.9867 - val_loss: 0.3442 - val_accuracy: 0.9258
-Epoch 507/1000
-60000/60000 - 6s - loss: 0.0546 - accuracy: 0.9863 - val_loss: 0.3456 - val_accuracy: 0.9237
-Epoch 508/1000
-60000/60000 - 6s - loss: 0.0535 - accuracy: 0.9866 - val_loss: 0.3446 - val_accuracy: 0.9236
-Epoch 509/1000
-60000/60000 - 6s - loss: 0.0538 - accuracy: 0.9865 - val_loss: 0.3441 - val_accuracy: 0.9226
-Epoch 510/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9870 - val_loss: 0.3475 - val_accuracy: 0.9239
-Epoch 511/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9873 - val_loss: 0.3446 - val_accuracy: 0.9235
-Epoch 512/1000
-60000/60000 - 6s - loss: 0.0524 - accuracy: 0.9871 - val_loss: 0.3491 - val_accuracy: 0.9235
-Epoch 513/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9872 - val_loss: 0.3466 - val_accuracy: 0.9228
-Epoch 514/1000
-60000/60000 - 6s - loss: 0.0509 - accuracy: 0.9874 - val_loss: 0.3457 - val_accuracy: 0.9235
-Epoch 515/1000
-60000/60000 - 6s - loss: 0.0511 - accuracy: 0.9872 - val_loss: 0.3455 - val_accuracy: 0.9235
-Epoch 516/1000
-60000/60000 - 6s - loss: 0.0509 - accuracy: 0.9874 - val_loss: 0.3457 - val_accuracy: 0.9241
-Epoch 517/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9873 - val_loss: 0.3438 - val_accuracy: 0.9238
-Epoch 518/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3457 - val_accuracy: 0.9233
-Epoch 519/1000
-60000/60000 - 6s - loss: 0.0532 - accuracy: 0.9865 - val_loss: 0.3450 - val_accuracy: 0.9229
-Epoch 520/1000
-60000/60000 - 6s - loss: 0.0528 - accuracy: 0.9869 - val_loss: 0.3460 - val_accuracy: 0.9226
-Epoch 521/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9872 - val_loss: 0.3444 - val_accuracy: 0.9229
-Epoch 522/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9872 - val_loss: 0.3475 - val_accuracy: 0.9223
-Epoch 523/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3459 - val_accuracy: 0.9231
-Epoch 524/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9874 - val_loss: 0.3472 - val_accuracy: 0.9232
-Epoch 525/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9877 - val_loss: 0.3464 - val_accuracy: 0.9237
-Epoch 526/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9876 - val_loss: 0.3453 - val_accuracy: 0.9239
-Epoch 527/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9876 - val_loss: 0.3465 - val_accuracy: 0.9233
-Epoch 528/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9874 - val_loss: 0.3497 - val_accuracy: 0.9225
-Epoch 529/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9875 - val_loss: 0.3475 - val_accuracy: 0.9240
-Epoch 530/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9874 - val_loss: 0.3500 - val_accuracy: 0.9237
-Epoch 531/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9873 - val_loss: 0.3485 - val_accuracy: 0.9239
-Epoch 532/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9873 - val_loss: 0.3496 - val_accuracy: 0.9224
-Epoch 533/1000
-60000/60000 - 6s - loss: 0.0507 - accuracy: 0.9875 - val_loss: 0.3458 - val_accuracy: 0.9235
-Epoch 534/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9874 - val_loss: 0.3464 - val_accuracy: 0.9231
-Epoch 535/1000
-60000/60000 - 6s - loss: 0.0519 - accuracy: 0.9871 - val_loss: 0.3486 - val_accuracy: 0.9225
-Epoch 536/1000
-60000/60000 - 6s - loss: 0.0525 - accuracy: 0.9867 - val_loss: 0.3474 - val_accuracy: 0.9235
-Epoch 537/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9870 - val_loss: 0.3485 - val_accuracy: 0.9244
-Epoch 538/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9873 - val_loss: 0.3476 - val_accuracy: 0.9223
-Epoch 539/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9872 - val_loss: 0.3467 - val_accuracy: 0.9232
-Epoch 540/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9871 - val_loss: 0.3465 - val_accuracy: 0.9234
-Epoch 541/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3458 - val_accuracy: 0.9246
-Epoch 542/1000
-60000/60000 - 6s - loss: 0.0512 - accuracy: 0.9874 - val_loss: 0.3470 - val_accuracy: 0.9234
-Epoch 543/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9876 - val_loss: 0.3456 - val_accuracy: 0.9238
-Epoch 544/1000
-60000/60000 - 6s - loss: 0.0508 - accuracy: 0.9874 - val_loss: 0.3449 - val_accuracy: 0.9243
-Epoch 545/1000
-60000/60000 - 6s - loss: 0.0508 - accuracy: 0.9875 - val_loss: 0.3460 - val_accuracy: 0.9237
-Epoch 546/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9875 - val_loss: 0.3471 - val_accuracy: 0.9235
-Epoch 547/1000
-60000/60000 - 6s - loss: 0.0519 - accuracy: 0.9871 - val_loss: 0.3454 - val_accuracy: 0.9238
-Epoch 548/1000
-60000/60000 - 6s - loss: 0.0511 - accuracy: 0.9872 - val_loss: 0.3479 - val_accuracy: 0.9237
-Epoch 549/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9874 - val_loss: 0.3479 - val_accuracy: 0.9233
-Epoch 550/1000
-60000/60000 - 6s - loss: 0.0507 - accuracy: 0.9873 - val_loss: 0.3484 - val_accuracy: 0.9229
-Epoch 551/1000
-60000/60000 - 6s - loss: 0.0508 - accuracy: 0.9875 - val_loss: 0.3484 - val_accuracy: 0.9236
-Epoch 552/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9872 - val_loss: 0.3487 - val_accuracy: 0.9228
-Epoch 553/1000
-60000/60000 - 6s - loss: 0.0533 - accuracy: 0.9866 - val_loss: 0.3469 - val_accuracy: 0.9227
-Epoch 554/1000
-60000/60000 - 6s - loss: 0.0519 - accuracy: 0.9871 - val_loss: 0.3482 - val_accuracy: 0.9229
-Epoch 555/1000
-60000/60000 - 6s - loss: 0.0522 - accuracy: 0.9869 - val_loss: 0.3502 - val_accuracy: 0.9229
-Epoch 556/1000
-60000/60000 - 6s - loss: 0.0515 - accuracy: 0.9873 - val_loss: 0.3507 - val_accuracy: 0.9231
-Epoch 557/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9876 - val_loss: 0.3480 - val_accuracy: 0.9240
-Epoch 558/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9876 - val_loss: 0.3475 - val_accuracy: 0.9235
-Epoch 559/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9871 - val_loss: 0.3490 - val_accuracy: 0.9236
-Epoch 560/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9873 - val_loss: 0.3463 - val_accuracy: 0.9243
-Epoch 561/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9876 - val_loss: 0.3490 - val_accuracy: 0.9232
-Epoch 562/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9875 - val_loss: 0.3497 - val_accuracy: 0.9243
-Epoch 563/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9874 - val_loss: 0.3482 - val_accuracy: 0.9241
-Epoch 564/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9876 - val_loss: 0.3476 - val_accuracy: 0.9243
-Epoch 565/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9876 - val_loss: 0.3483 - val_accuracy: 0.9240
-Epoch 566/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9876 - val_loss: 0.3506 - val_accuracy: 0.9230
-Epoch 567/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9873 - val_loss: 0.3500 - val_accuracy: 0.9228
-Epoch 568/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9872 - val_loss: 0.3506 - val_accuracy: 0.9229
-Epoch 569/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9876 - val_loss: 0.3490 - val_accuracy: 0.9236
-Epoch 570/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9876 - val_loss: 0.3494 - val_accuracy: 0.9228
-Epoch 571/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9874 - val_loss: 0.3479 - val_accuracy: 0.9233
-Epoch 572/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9875 - val_loss: 0.3487 - val_accuracy: 0.9241
-Epoch 573/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9877 - val_loss: 0.3499 - val_accuracy: 0.9242
-Epoch 574/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9876 - val_loss: 0.3489 - val_accuracy: 0.9243
-Epoch 575/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9876 - val_loss: 0.3487 - val_accuracy: 0.9234
-Epoch 576/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9875 - val_loss: 0.3482 - val_accuracy: 0.9236
-Epoch 577/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9878 - val_loss: 0.3482 - val_accuracy: 0.9241
-Epoch 578/1000
-60000/60000 - 6s - loss: 0.0507 - accuracy: 0.9874 - val_loss: 0.3503 - val_accuracy: 0.9238
-Epoch 579/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9876 - val_loss: 0.3488 - val_accuracy: 0.9233
-Epoch 580/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9875 - val_loss: 0.3469 - val_accuracy: 0.9237
-Epoch 581/1000
-60000/60000 - 6s - loss: 0.0507 - accuracy: 0.9872 - val_loss: 0.3498 - val_accuracy: 0.9230
-Epoch 582/1000
-60000/60000 - 6s - loss: 0.0507 - accuracy: 0.9873 - val_loss: 0.3485 - val_accuracy: 0.9227
-Epoch 583/1000
-60000/60000 - 6s - loss: 0.0523 - accuracy: 0.9869 - val_loss: 0.3470 - val_accuracy: 0.9237
-Epoch 584/1000
-60000/60000 - 6s - loss: 0.0523 - accuracy: 0.9869 - val_loss: 0.3492 - val_accuracy: 0.9233
-Epoch 585/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9866 - val_loss: 0.3564 - val_accuracy: 0.9231
-Epoch 586/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9862 - val_loss: 0.3501 - val_accuracy: 0.9236
-Epoch 587/1000
-60000/60000 - 6s - loss: 0.0539 - accuracy: 0.9863 - val_loss: 0.3563 - val_accuracy: 0.9214
-Epoch 588/1000
-60000/60000 - 6s - loss: 0.0543 - accuracy: 0.9862 - val_loss: 0.3497 - val_accuracy: 0.9230
-Epoch 589/1000
-60000/60000 - 6s - loss: 0.0538 - accuracy: 0.9863 - val_loss: 0.3488 - val_accuracy: 0.9231
-Epoch 590/1000
-60000/60000 - 6s - loss: 0.0518 - accuracy: 0.9871 - val_loss: 0.3476 - val_accuracy: 0.9242
-Epoch 591/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9869 - val_loss: 0.3482 - val_accuracy: 0.9236
-Epoch 592/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3530 - val_accuracy: 0.9235
-Epoch 593/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9870 - val_loss: 0.3529 - val_accuracy: 0.9237
-Epoch 594/1000
-60000/60000 - 6s - loss: 0.0512 - accuracy: 0.9873 - val_loss: 0.3492 - val_accuracy: 0.9238
-Epoch 595/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9872 - val_loss: 0.3538 - val_accuracy: 0.9226
-Epoch 596/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9875 - val_loss: 0.3523 - val_accuracy: 0.9240
-Epoch 597/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9874 - val_loss: 0.3527 - val_accuracy: 0.9223
-Epoch 598/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9876 - val_loss: 0.3506 - val_accuracy: 0.9245
-Epoch 599/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9872 - val_loss: 0.3542 - val_accuracy: 0.9235
-Epoch 600/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9874 - val_loss: 0.3526 - val_accuracy: 0.9229
-Epoch 601/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9875 - val_loss: 0.3516 - val_accuracy: 0.9235
-Epoch 602/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9877 - val_loss: 0.3520 - val_accuracy: 0.9230
-Epoch 603/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9876 - val_loss: 0.3504 - val_accuracy: 0.9228
-Epoch 604/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9875 - val_loss: 0.3489 - val_accuracy: 0.9239
-Epoch 605/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9874 - val_loss: 0.3509 - val_accuracy: 0.9240
-Epoch 606/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9876 - val_loss: 0.3501 - val_accuracy: 0.9239
-Epoch 607/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9876 - val_loss: 0.3499 - val_accuracy: 0.9229
-Epoch 608/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9875 - val_loss: 0.3519 - val_accuracy: 0.9224
-Epoch 609/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9877 - val_loss: 0.3519 - val_accuracy: 0.9227
-Epoch 610/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9874 - val_loss: 0.3497 - val_accuracy: 0.9242
-Epoch 611/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9877 - val_loss: 0.3507 - val_accuracy: 0.9242
-Epoch 612/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9873 - val_loss: 0.3516 - val_accuracy: 0.9228
-Epoch 613/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9873 - val_loss: 0.3504 - val_accuracy: 0.9235
-Epoch 614/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9874 - val_loss: 0.3494 - val_accuracy: 0.9246
-Epoch 615/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9876 - val_loss: 0.3508 - val_accuracy: 0.9231
-Epoch 616/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9877 - val_loss: 0.3511 - val_accuracy: 0.9237
-Epoch 617/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9876 - val_loss: 0.3518 - val_accuracy: 0.9237
-Epoch 618/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9877 - val_loss: 0.3513 - val_accuracy: 0.9236
-Epoch 619/1000
-60000/60000 - 6s - loss: 0.0493 - accuracy: 0.9877 - val_loss: 0.3533 - val_accuracy: 0.9236
-Epoch 620/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9871 - val_loss: 0.3507 - val_accuracy: 0.9227
-Epoch 621/1000
-60000/60000 - 6s - loss: 0.0535 - accuracy: 0.9865 - val_loss: 0.3498 - val_accuracy: 0.9245
-Epoch 622/1000
-60000/60000 - 6s - loss: 0.0517 - accuracy: 0.9869 - val_loss: 0.3514 - val_accuracy: 0.9232
-Epoch 623/1000
-60000/60000 - 6s - loss: 0.0511 - accuracy: 0.9873 - val_loss: 0.3508 - val_accuracy: 0.9241
-Epoch 624/1000
-60000/60000 - 6s - loss: 0.0515 - accuracy: 0.9870 - val_loss: 0.3480 - val_accuracy: 0.9240
-Epoch 625/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9875 - val_loss: 0.3482 - val_accuracy: 0.9245
-Epoch 626/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9876 - val_loss: 0.3486 - val_accuracy: 0.9236
-Epoch 627/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9875 - val_loss: 0.3491 - val_accuracy: 0.9233
-Epoch 628/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9876 - val_loss: 0.3465 - val_accuracy: 0.9241
-Epoch 629/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9874 - val_loss: 0.3486 - val_accuracy: 0.9232
-Epoch 630/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9876 - val_loss: 0.3495 - val_accuracy: 0.9239
-Epoch 631/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9876 - val_loss: 0.3492 - val_accuracy: 0.9227
-Epoch 632/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9877 - val_loss: 0.3476 - val_accuracy: 0.9244
-Epoch 633/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9876 - val_loss: 0.3501 - val_accuracy: 0.9238
-Epoch 634/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9878 - val_loss: 0.3491 - val_accuracy: 0.9246
-Epoch 635/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9877 - val_loss: 0.3488 - val_accuracy: 0.9233
-Epoch 636/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9875 - val_loss: 0.3477 - val_accuracy: 0.9250
-Epoch 637/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9875 - val_loss: 0.3494 - val_accuracy: 0.9243
-Epoch 638/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9872 - val_loss: 0.3474 - val_accuracy: 0.9241
-Epoch 639/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9875 - val_loss: 0.3472 - val_accuracy: 0.9247
-Epoch 640/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9877 - val_loss: 0.3487 - val_accuracy: 0.9232
-Epoch 641/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9875 - val_loss: 0.3496 - val_accuracy: 0.9233
-Epoch 642/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9875 - val_loss: 0.3510 - val_accuracy: 0.9236
-Epoch 643/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9875 - val_loss: 0.3507 - val_accuracy: 0.9241
-Epoch 644/1000
-60000/60000 - 6s - loss: 0.0493 - accuracy: 0.9878 - val_loss: 0.3514 - val_accuracy: 0.9241
-Epoch 645/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9876 - val_loss: 0.3499 - val_accuracy: 0.9242
-Epoch 646/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9874 - val_loss: 0.3500 - val_accuracy: 0.9242
-Epoch 647/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9872 - val_loss: 0.3494 - val_accuracy: 0.9243
-Epoch 648/1000
-60000/60000 - 6s - loss: 0.0508 - accuracy: 0.9871 - val_loss: 0.3512 - val_accuracy: 0.9241
-Epoch 649/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9874 - val_loss: 0.3516 - val_accuracy: 0.9241
-Epoch 650/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9874 - val_loss: 0.3492 - val_accuracy: 0.9236
-Epoch 651/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9876 - val_loss: 0.3501 - val_accuracy: 0.9239
-Epoch 652/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9872 - val_loss: 0.3518 - val_accuracy: 0.9236
-Epoch 653/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9877 - val_loss: 0.3506 - val_accuracy: 0.9235
-Epoch 654/1000
-60000/60000 - 6s - loss: 0.0505 - accuracy: 0.9873 - val_loss: 0.3519 - val_accuracy: 0.9225
-Epoch 655/1000
-60000/60000 - 6s - loss: 0.0509 - accuracy: 0.9869 - val_loss: 0.3521 - val_accuracy: 0.9227
-Epoch 656/1000
-60000/60000 - 6s - loss: 0.0509 - accuracy: 0.9872 - val_loss: 0.3510 - val_accuracy: 0.9241
-Epoch 657/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9875 - val_loss: 0.3518 - val_accuracy: 0.9231
-Epoch 658/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9875 - val_loss: 0.3501 - val_accuracy: 0.9247
-Epoch 659/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9876 - val_loss: 0.3513 - val_accuracy: 0.9245
-Epoch 660/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9876 - val_loss: 0.3509 - val_accuracy: 0.9240
-Epoch 661/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9878 - val_loss: 0.3518 - val_accuracy: 0.9238
-Epoch 662/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9876 - val_loss: 0.3515 - val_accuracy: 0.9236
-Epoch 663/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9877 - val_loss: 0.3487 - val_accuracy: 0.9243
-Epoch 664/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9877 - val_loss: 0.3515 - val_accuracy: 0.9241
-Epoch 665/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9877 - val_loss: 0.3503 - val_accuracy: 0.9239
-Epoch 666/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9877 - val_loss: 0.3503 - val_accuracy: 0.9242
-Epoch 667/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9877 - val_loss: 0.3522 - val_accuracy: 0.9237
-Epoch 668/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9873 - val_loss: 0.3495 - val_accuracy: 0.9240
-Epoch 669/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9876 - val_loss: 0.3513 - val_accuracy: 0.9234
-Epoch 670/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9877 - val_loss: 0.3506 - val_accuracy: 0.9242
-Epoch 671/1000
-60000/60000 - 6s - loss: 0.0509 - accuracy: 0.9872 - val_loss: 0.3537 - val_accuracy: 0.9237
-Epoch 672/1000
-60000/60000 - 6s - loss: 0.0504 - accuracy: 0.9874 - val_loss: 0.3522 - val_accuracy: 0.9238
-Epoch 673/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9877 - val_loss: 0.3525 - val_accuracy: 0.9239
-Epoch 674/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9875 - val_loss: 0.3552 - val_accuracy: 0.9234
-Epoch 675/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9876 - val_loss: 0.3534 - val_accuracy: 0.9240
-Epoch 676/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9876 - val_loss: 0.3540 - val_accuracy: 0.9239
-Epoch 677/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9870 - val_loss: 0.3576 - val_accuracy: 0.9227
-Epoch 678/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9873 - val_loss: 0.3513 - val_accuracy: 0.9243
-Epoch 679/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9877 - val_loss: 0.3523 - val_accuracy: 0.9236
-Epoch 680/1000
-60000/60000 - 6s - loss: 0.0493 - accuracy: 0.9876 - val_loss: 0.3524 - val_accuracy: 0.9234
-Epoch 681/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9878 - val_loss: 0.3511 - val_accuracy: 0.9232
-Epoch 682/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9878 - val_loss: 0.3511 - val_accuracy: 0.9240
-Epoch 683/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9875 - val_loss: 0.3521 - val_accuracy: 0.9225
-Epoch 684/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9879 - val_loss: 0.3503 - val_accuracy: 0.9232
-Epoch 685/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9877 - val_loss: 0.3514 - val_accuracy: 0.9232
-Epoch 686/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9864 - val_loss: 0.3504 - val_accuracy: 0.9234
-Epoch 687/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9872 - val_loss: 0.3512 - val_accuracy: 0.9230
-Epoch 688/1000
-60000/60000 - 6s - loss: 0.0514 - accuracy: 0.9871 - val_loss: 0.3502 - val_accuracy: 0.9239
-Epoch 689/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9875 - val_loss: 0.3528 - val_accuracy: 0.9230
-Epoch 690/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9876 - val_loss: 0.3539 - val_accuracy: 0.9235
-Epoch 691/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9878 - val_loss: 0.3524 - val_accuracy: 0.9234
-Epoch 692/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9878 - val_loss: 0.3510 - val_accuracy: 0.9230
-Epoch 693/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9879 - val_loss: 0.3532 - val_accuracy: 0.9234
-Epoch 694/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3551 - val_accuracy: 0.9229
-Epoch 695/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9877 - val_loss: 0.3522 - val_accuracy: 0.9229
-Epoch 696/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3537 - val_accuracy: 0.9232
-Epoch 697/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9877 - val_loss: 0.3516 - val_accuracy: 0.9237
-Epoch 698/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9877 - val_loss: 0.3522 - val_accuracy: 0.9226
-Epoch 699/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9879 - val_loss: 0.3537 - val_accuracy: 0.9220
-Epoch 700/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9875 - val_loss: 0.3495 - val_accuracy: 0.9238
-Epoch 701/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9878 - val_loss: 0.3517 - val_accuracy: 0.9233
-Epoch 702/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3526 - val_accuracy: 0.9224
-Epoch 703/1000
-60000/60000 - 6s - loss: 0.0493 - accuracy: 0.9876 - val_loss: 0.3536 - val_accuracy: 0.9228
-Epoch 704/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9874 - val_loss: 0.3540 - val_accuracy: 0.9223
-Epoch 705/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9876 - val_loss: 0.3544 - val_accuracy: 0.9230
-Epoch 706/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9877 - val_loss: 0.3514 - val_accuracy: 0.9226
-Epoch 707/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3526 - val_accuracy: 0.9233
-Epoch 708/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9876 - val_loss: 0.3551 - val_accuracy: 0.9211
-Epoch 709/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9876 - val_loss: 0.3531 - val_accuracy: 0.9231
-Epoch 710/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9879 - val_loss: 0.3539 - val_accuracy: 0.9227
-Epoch 711/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9879 - val_loss: 0.3545 - val_accuracy: 0.9221
-Epoch 712/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9877 - val_loss: 0.3519 - val_accuracy: 0.9228
-Epoch 713/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9878 - val_loss: 0.3526 - val_accuracy: 0.9236
-Epoch 714/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9879 - val_loss: 0.3539 - val_accuracy: 0.9234
-Epoch 715/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3514 - val_accuracy: 0.9230
-Epoch 716/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3509 - val_accuracy: 0.9238
-Epoch 717/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9878 - val_loss: 0.3528 - val_accuracy: 0.9236
-Epoch 718/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3521 - val_accuracy: 0.9230
-Epoch 719/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9880 - val_loss: 0.3512 - val_accuracy: 0.9231
-Epoch 720/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3526 - val_accuracy: 0.9229
-Epoch 721/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9877 - val_loss: 0.3558 - val_accuracy: 0.9228
-Epoch 722/1000
-60000/60000 - 6s - loss: 0.0493 - accuracy: 0.9878 - val_loss: 0.3540 - val_accuracy: 0.9233
-Epoch 723/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9879 - val_loss: 0.3538 - val_accuracy: 0.9234
-Epoch 724/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9879 - val_loss: 0.3535 - val_accuracy: 0.9227
-Epoch 725/1000
-60000/60000 - 6s - loss: 0.0497 - accuracy: 0.9875 - val_loss: 0.3508 - val_accuracy: 0.9241
-Epoch 726/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9877 - val_loss: 0.3535 - val_accuracy: 0.9231
-Epoch 727/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9873 - val_loss: 0.3535 - val_accuracy: 0.9247
-Epoch 728/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9868 - val_loss: 0.3582 - val_accuracy: 0.9231
-Epoch 729/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9872 - val_loss: 0.3550 - val_accuracy: 0.9235
-Epoch 730/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9876 - val_loss: 0.3529 - val_accuracy: 0.9237
-Epoch 731/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9879 - val_loss: 0.3538 - val_accuracy: 0.9233
-Epoch 732/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9874 - val_loss: 0.3555 - val_accuracy: 0.9230
-Epoch 733/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9877 - val_loss: 0.3560 - val_accuracy: 0.9227
-Epoch 734/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9876 - val_loss: 0.3557 - val_accuracy: 0.9229
-Epoch 735/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3548 - val_accuracy: 0.9227
-Epoch 736/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3533 - val_accuracy: 0.9235
-Epoch 737/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3524 - val_accuracy: 0.9231
-Epoch 738/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9878 - val_loss: 0.3565 - val_accuracy: 0.9233
-Epoch 739/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3555 - val_accuracy: 0.9229
-Epoch 740/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9878 - val_loss: 0.3548 - val_accuracy: 0.9232
-Epoch 741/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9880 - val_loss: 0.3560 - val_accuracy: 0.9220
-Epoch 742/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9878 - val_loss: 0.3575 - val_accuracy: 0.9232
-Epoch 743/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9877 - val_loss: 0.3560 - val_accuracy: 0.9223
-Epoch 744/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9877 - val_loss: 0.3579 - val_accuracy: 0.9222
-Epoch 745/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9878 - val_loss: 0.3564 - val_accuracy: 0.9230
-Epoch 746/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3558 - val_accuracy: 0.9230
-Epoch 747/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3567 - val_accuracy: 0.9232
-Epoch 748/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3555 - val_accuracy: 0.9225
-Epoch 749/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3542 - val_accuracy: 0.9230
-Epoch 750/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3561 - val_accuracy: 0.9224
-Epoch 751/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3569 - val_accuracy: 0.9230
-Epoch 752/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3554 - val_accuracy: 0.9229
-Epoch 753/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9872 - val_loss: 0.3572 - val_accuracy: 0.9213
-Epoch 754/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9877 - val_loss: 0.3560 - val_accuracy: 0.9228
-Epoch 755/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3576 - val_accuracy: 0.9225
-Epoch 756/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9880 - val_loss: 0.3567 - val_accuracy: 0.9232
-Epoch 757/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3547 - val_accuracy: 0.9224
-Epoch 758/1000
-60000/60000 - 6s - loss: 0.0500 - accuracy: 0.9873 - val_loss: 0.3540 - val_accuracy: 0.9210
-Epoch 759/1000
-60000/60000 - 6s - loss: 0.0499 - accuracy: 0.9874 - val_loss: 0.3569 - val_accuracy: 0.9222
-Epoch 760/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9877 - val_loss: 0.3578 - val_accuracy: 0.9218
-Epoch 761/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9876 - val_loss: 0.3578 - val_accuracy: 0.9228
-Epoch 762/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9877 - val_loss: 0.3574 - val_accuracy: 0.9224
-Epoch 763/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3558 - val_accuracy: 0.9231
-Epoch 764/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3565 - val_accuracy: 0.9228
-Epoch 765/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9878 - val_loss: 0.3564 - val_accuracy: 0.9226
-Epoch 766/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9877 - val_loss: 0.3564 - val_accuracy: 0.9221
-Epoch 767/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3575 - val_accuracy: 0.9223
-Epoch 768/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3565 - val_accuracy: 0.9229
-Epoch 769/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9876 - val_loss: 0.3544 - val_accuracy: 0.9229
-Epoch 770/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3559 - val_accuracy: 0.9233
-Epoch 771/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9877 - val_loss: 0.3541 - val_accuracy: 0.9228
-Epoch 772/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9875 - val_loss: 0.3570 - val_accuracy: 0.9226
-Epoch 773/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9877 - val_loss: 0.3549 - val_accuracy: 0.9233
-Epoch 774/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9877 - val_loss: 0.3600 - val_accuracy: 0.9224
-Epoch 775/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9877 - val_loss: 0.3582 - val_accuracy: 0.9222
-Epoch 776/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9879 - val_loss: 0.3587 - val_accuracy: 0.9227
-Epoch 777/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9877 - val_loss: 0.3586 - val_accuracy: 0.9222
-Epoch 778/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3597 - val_accuracy: 0.9217
-Epoch 779/1000
-60000/60000 - 6s - loss: 0.0507 - accuracy: 0.9871 - val_loss: 0.3596 - val_accuracy: 0.9211
-Epoch 780/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9877 - val_loss: 0.3577 - val_accuracy: 0.9216
-Epoch 781/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9875 - val_loss: 0.3569 - val_accuracy: 0.9224
-Epoch 782/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9876 - val_loss: 0.3570 - val_accuracy: 0.9232
-Epoch 783/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9877 - val_loss: 0.3577 - val_accuracy: 0.9228
-Epoch 784/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3580 - val_accuracy: 0.9224
-Epoch 785/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3572 - val_accuracy: 0.9223
-Epoch 786/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9876 - val_loss: 0.3575 - val_accuracy: 0.9234
-Epoch 787/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9878 - val_loss: 0.3574 - val_accuracy: 0.9228
-Epoch 788/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9878 - val_loss: 0.3556 - val_accuracy: 0.9233
-Epoch 789/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9878 - val_loss: 0.3570 - val_accuracy: 0.9228
-Epoch 790/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3570 - val_accuracy: 0.9232
-Epoch 791/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9877 - val_loss: 0.3584 - val_accuracy: 0.9224
-Epoch 792/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9878 - val_loss: 0.3575 - val_accuracy: 0.9227
-Epoch 793/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3574 - val_accuracy: 0.9226
-Epoch 794/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3581 - val_accuracy: 0.9219
-Epoch 795/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3582 - val_accuracy: 0.9219
-Epoch 796/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9879 - val_loss: 0.3576 - val_accuracy: 0.9224
-Epoch 797/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3572 - val_accuracy: 0.9235
-Epoch 798/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3578 - val_accuracy: 0.9226
-Epoch 799/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9882 - val_loss: 0.3584 - val_accuracy: 0.9223
-Epoch 800/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3595 - val_accuracy: 0.9222
-Epoch 801/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9881 - val_loss: 0.3581 - val_accuracy: 0.9217
-Epoch 802/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9879 - val_loss: 0.3597 - val_accuracy: 0.9233
-Epoch 803/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9877 - val_loss: 0.3590 - val_accuracy: 0.9226
-Epoch 804/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9879 - val_loss: 0.3609 - val_accuracy: 0.9221
-Epoch 805/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9877 - val_loss: 0.3583 - val_accuracy: 0.9221
-Epoch 806/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3585 - val_accuracy: 0.9210
-Epoch 807/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9877 - val_loss: 0.3591 - val_accuracy: 0.9224
-Epoch 808/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3579 - val_accuracy: 0.9224
-Epoch 809/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3580 - val_accuracy: 0.9219
-Epoch 810/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3590 - val_accuracy: 0.9227
-Epoch 811/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9881 - val_loss: 0.3603 - val_accuracy: 0.9218
-Epoch 812/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3594 - val_accuracy: 0.9230
-Epoch 813/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3589 - val_accuracy: 0.9221
-Epoch 814/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3590 - val_accuracy: 0.9213
-Epoch 815/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9879 - val_loss: 0.3583 - val_accuracy: 0.9218
-Epoch 816/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3612 - val_accuracy: 0.9218
-Epoch 817/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9879 - val_loss: 0.3597 - val_accuracy: 0.9217
-Epoch 818/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3575 - val_accuracy: 0.9228
-Epoch 819/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3585 - val_accuracy: 0.9218
-Epoch 820/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9879 - val_loss: 0.3596 - val_accuracy: 0.9220
-Epoch 821/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3588 - val_accuracy: 0.9215
-Epoch 822/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9871 - val_loss: 0.3606 - val_accuracy: 0.9225
-Epoch 823/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9873 - val_loss: 0.3574 - val_accuracy: 0.9227
-Epoch 824/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9878 - val_loss: 0.3574 - val_accuracy: 0.9216
-Epoch 825/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3584 - val_accuracy: 0.9223
-Epoch 826/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9880 - val_loss: 0.3587 - val_accuracy: 0.9223
-Epoch 827/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3587 - val_accuracy: 0.9217
-Epoch 828/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3586 - val_accuracy: 0.9227
-Epoch 829/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3580 - val_accuracy: 0.9225
-Epoch 830/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3591 - val_accuracy: 0.9225
-Epoch 831/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9874 - val_loss: 0.3617 - val_accuracy: 0.9226
-Epoch 832/1000
-60000/60000 - 6s - loss: 0.0502 - accuracy: 0.9873 - val_loss: 0.3625 - val_accuracy: 0.9220
-Epoch 833/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9879 - val_loss: 0.3601 - val_accuracy: 0.9217
-Epoch 834/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9877 - val_loss: 0.3620 - val_accuracy: 0.9218
-Epoch 835/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9879 - val_loss: 0.3591 - val_accuracy: 0.9227
-Epoch 836/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9882 - val_loss: 0.3582 - val_accuracy: 0.9221
-Epoch 837/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3584 - val_accuracy: 0.9228
-Epoch 838/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3582 - val_accuracy: 0.9227
-Epoch 839/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3580 - val_accuracy: 0.9228
-Epoch 840/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3598 - val_accuracy: 0.9231
-Epoch 841/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3603 - val_accuracy: 0.9218
-Epoch 842/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9881 - val_loss: 0.3595 - val_accuracy: 0.9221
-Epoch 843/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3590 - val_accuracy: 0.9224
-Epoch 844/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9878 - val_loss: 0.3617 - val_accuracy: 0.9233
-Epoch 845/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3621 - val_accuracy: 0.9221
-Epoch 846/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3605 - val_accuracy: 0.9219
-Epoch 847/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9881 - val_loss: 0.3607 - val_accuracy: 0.9221
-Epoch 848/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3599 - val_accuracy: 0.9227
-Epoch 849/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3596 - val_accuracy: 0.9221
-Epoch 850/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3605 - val_accuracy: 0.9233
-Epoch 851/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3597 - val_accuracy: 0.9225
-Epoch 852/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9878 - val_loss: 0.3602 - val_accuracy: 0.9223
-Epoch 853/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9877 - val_loss: 0.3592 - val_accuracy: 0.9213
-Epoch 854/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9877 - val_loss: 0.3591 - val_accuracy: 0.9243
-Epoch 855/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3604 - val_accuracy: 0.9221
-Epoch 856/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9881 - val_loss: 0.3579 - val_accuracy: 0.9227
-Epoch 857/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3589 - val_accuracy: 0.9220
-Epoch 858/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3607 - val_accuracy: 0.9227
-Epoch 859/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3605 - val_accuracy: 0.9223
-Epoch 860/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9881 - val_loss: 0.3605 - val_accuracy: 0.9231
-Epoch 861/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9880 - val_loss: 0.3608 - val_accuracy: 0.9220
-Epoch 862/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9881 - val_loss: 0.3624 - val_accuracy: 0.9226
-Epoch 863/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3622 - val_accuracy: 0.9220
-Epoch 864/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3607 - val_accuracy: 0.9213
-Epoch 865/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3605 - val_accuracy: 0.9220
-Epoch 866/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9882 - val_loss: 0.3602 - val_accuracy: 0.9217
-Epoch 867/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9880 - val_loss: 0.3611 - val_accuracy: 0.9223
-Epoch 868/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9880 - val_loss: 0.3598 - val_accuracy: 0.9222
-Epoch 869/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3610 - val_accuracy: 0.9213
-Epoch 870/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3596 - val_accuracy: 0.9222
-Epoch 871/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3607 - val_accuracy: 0.9219
-Epoch 872/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3591 - val_accuracy: 0.9227
-Epoch 873/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9879 - val_loss: 0.3609 - val_accuracy: 0.9215
-Epoch 874/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9881 - val_loss: 0.3609 - val_accuracy: 0.9219
-Epoch 875/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3608 - val_accuracy: 0.9212
-Epoch 876/1000
-60000/60000 - 6s - loss: 0.0494 - accuracy: 0.9875 - val_loss: 0.3579 - val_accuracy: 0.9228
-Epoch 877/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9876 - val_loss: 0.3603 - val_accuracy: 0.9226
-Epoch 878/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9878 - val_loss: 0.3620 - val_accuracy: 0.9215
-Epoch 879/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9880 - val_loss: 0.3623 - val_accuracy: 0.9220
-Epoch 880/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9880 - val_loss: 0.3591 - val_accuracy: 0.9229
-Epoch 881/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3631 - val_accuracy: 0.9217
-Epoch 882/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3585 - val_accuracy: 0.9218
-Epoch 883/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3600 - val_accuracy: 0.9218
-Epoch 884/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9881 - val_loss: 0.3605 - val_accuracy: 0.9220
-Epoch 885/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9881 - val_loss: 0.3609 - val_accuracy: 0.9218
-Epoch 886/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3608 - val_accuracy: 0.9208
-Epoch 887/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9881 - val_loss: 0.3593 - val_accuracy: 0.9222
-Epoch 888/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3593 - val_accuracy: 0.9225
-Epoch 889/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3596 - val_accuracy: 0.9221
-Epoch 890/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9880 - val_loss: 0.3629 - val_accuracy: 0.9225
-Epoch 891/1000
-60000/60000 - 6s - loss: 0.0490 - accuracy: 0.9876 - val_loss: 0.3597 - val_accuracy: 0.9221
-Epoch 892/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9879 - val_loss: 0.3615 - val_accuracy: 0.9223
-Epoch 893/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3610 - val_accuracy: 0.9218
-Epoch 894/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3621 - val_accuracy: 0.9216
-Epoch 895/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3618 - val_accuracy: 0.9226
-Epoch 896/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3632 - val_accuracy: 0.9221
-Epoch 897/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9882 - val_loss: 0.3621 - val_accuracy: 0.9218
-Epoch 898/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3605 - val_accuracy: 0.9221
-Epoch 899/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3630 - val_accuracy: 0.9217
-Epoch 900/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3606 - val_accuracy: 0.9225
-Epoch 901/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3597 - val_accuracy: 0.9224
-Epoch 902/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3633 - val_accuracy: 0.9213
-Epoch 903/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3627 - val_accuracy: 0.9214
-Epoch 904/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3624 - val_accuracy: 0.9212
-Epoch 905/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3597 - val_accuracy: 0.9224
-Epoch 906/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3612 - val_accuracy: 0.9219
-Epoch 907/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9882 - val_loss: 0.3623 - val_accuracy: 0.9221
-Epoch 908/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3616 - val_accuracy: 0.9231
-Epoch 909/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3604 - val_accuracy: 0.9229
-Epoch 910/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3606 - val_accuracy: 0.9223
-Epoch 911/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9882 - val_loss: 0.3608 - val_accuracy: 0.9229
-Epoch 912/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9881 - val_loss: 0.3614 - val_accuracy: 0.9225
-Epoch 913/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3599 - val_accuracy: 0.9218
-Epoch 914/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3611 - val_accuracy: 0.9223
-Epoch 915/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3598 - val_accuracy: 0.9222
-Epoch 916/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9878 - val_loss: 0.3612 - val_accuracy: 0.9211
-Epoch 917/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9876 - val_loss: 0.3620 - val_accuracy: 0.9208
-Epoch 918/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9876 - val_loss: 0.3646 - val_accuracy: 0.9208
-Epoch 919/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9880 - val_loss: 0.3589 - val_accuracy: 0.9228
-Epoch 920/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3585 - val_accuracy: 0.9222
-Epoch 921/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3594 - val_accuracy: 0.9235
-Epoch 922/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9881 - val_loss: 0.3603 - val_accuracy: 0.9226
-Epoch 923/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3617 - val_accuracy: 0.9221
-Epoch 924/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9879 - val_loss: 0.3623 - val_accuracy: 0.9217
-Epoch 925/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9879 - val_loss: 0.3582 - val_accuracy: 0.9222
-Epoch 926/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9880 - val_loss: 0.3598 - val_accuracy: 0.9224
-Epoch 927/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9880 - val_loss: 0.3568 - val_accuracy: 0.9223
-Epoch 928/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3612 - val_accuracy: 0.9218
-Epoch 929/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9881 - val_loss: 0.3616 - val_accuracy: 0.9221
-Epoch 930/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9880 - val_loss: 0.3608 - val_accuracy: 0.9223
-Epoch 931/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3627 - val_accuracy: 0.9216
-Epoch 932/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3607 - val_accuracy: 0.9223
-Epoch 933/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3612 - val_accuracy: 0.9225
-Epoch 934/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9881 - val_loss: 0.3617 - val_accuracy: 0.9219
-Epoch 935/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3611 - val_accuracy: 0.9219
-Epoch 936/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9882 - val_loss: 0.3622 - val_accuracy: 0.9215
-Epoch 937/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9879 - val_loss: 0.3635 - val_accuracy: 0.9218
-Epoch 938/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3612 - val_accuracy: 0.9222
-Epoch 939/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3619 - val_accuracy: 0.9214
-Epoch 940/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3614 - val_accuracy: 0.9218
-Epoch 941/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9881 - val_loss: 0.3621 - val_accuracy: 0.9220
-Epoch 942/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9883 - val_loss: 0.3639 - val_accuracy: 0.9212
-Epoch 943/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9882 - val_loss: 0.3616 - val_accuracy: 0.9231
-Epoch 944/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9880 - val_loss: 0.3627 - val_accuracy: 0.9206
-Epoch 945/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3649 - val_accuracy: 0.9219
-Epoch 946/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9882 - val_loss: 0.3647 - val_accuracy: 0.9217
-Epoch 947/1000
-60000/60000 - 6s - loss: 0.0471 - accuracy: 0.9882 - val_loss: 0.3641 - val_accuracy: 0.9220
-Epoch 948/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9882 - val_loss: 0.3639 - val_accuracy: 0.9209
-Epoch 949/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9876 - val_loss: 0.3631 - val_accuracy: 0.9213
-Epoch 950/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3633 - val_accuracy: 0.9220
-Epoch 951/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9877 - val_loss: 0.3623 - val_accuracy: 0.9221
-Epoch 952/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9878 - val_loss: 0.3646 - val_accuracy: 0.9208
-Epoch 953/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3649 - val_accuracy: 0.9222
-Epoch 954/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9881 - val_loss: 0.3624 - val_accuracy: 0.9219
-Epoch 955/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3625 - val_accuracy: 0.9212
-Epoch 956/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3643 - val_accuracy: 0.9211
-Epoch 957/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9878 - val_loss: 0.3648 - val_accuracy: 0.9222
-Epoch 958/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3630 - val_accuracy: 0.9213
-Epoch 959/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9882 - val_loss: 0.3627 - val_accuracy: 0.9216
-Epoch 960/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9880 - val_loss: 0.3632 - val_accuracy: 0.9214
-Epoch 961/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9882 - val_loss: 0.3621 - val_accuracy: 0.9210
-Epoch 962/1000
-60000/60000 - 6s - loss: 0.0488 - accuracy: 0.9877 - val_loss: 0.3611 - val_accuracy: 0.9227
-Epoch 963/1000
-60000/60000 - 6s - loss: 0.0492 - accuracy: 0.9876 - val_loss: 0.3618 - val_accuracy: 0.9215
-Epoch 964/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9879 - val_loss: 0.3629 - val_accuracy: 0.9231
-Epoch 965/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9880 - val_loss: 0.3602 - val_accuracy: 0.9217
-Epoch 966/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9879 - val_loss: 0.3643 - val_accuracy: 0.9221
-Epoch 967/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9882 - val_loss: 0.3625 - val_accuracy: 0.9224
-Epoch 968/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9882 - val_loss: 0.3633 - val_accuracy: 0.9220
-Epoch 969/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9882 - val_loss: 0.3633 - val_accuracy: 0.9218
-Epoch 970/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3632 - val_accuracy: 0.9224
-Epoch 971/1000
-60000/60000 - 6s - loss: 0.0485 - accuracy: 0.9878 - val_loss: 0.3629 - val_accuracy: 0.9222
-Epoch 972/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3645 - val_accuracy: 0.9208
-Epoch 973/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9878 - val_loss: 0.3620 - val_accuracy: 0.9213
-Epoch 974/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3639 - val_accuracy: 0.9217
-Epoch 975/1000
-60000/60000 - 6s - loss: 0.0479 - accuracy: 0.9880 - val_loss: 0.3632 - val_accuracy: 0.9219
-Epoch 976/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3640 - val_accuracy: 0.9227
-Epoch 977/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9880 - val_loss: 0.3622 - val_accuracy: 0.9211
-Epoch 978/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9880 - val_loss: 0.3652 - val_accuracy: 0.9221
-Epoch 979/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9879 - val_loss: 0.3631 - val_accuracy: 0.9222
-Epoch 980/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9880 - val_loss: 0.3640 - val_accuracy: 0.9224
-Epoch 981/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9882 - val_loss: 0.3666 - val_accuracy: 0.9208
-Epoch 982/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9876 - val_loss: 0.3658 - val_accuracy: 0.9216
-Epoch 983/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9883 - val_loss: 0.3642 - val_accuracy: 0.9219
-Epoch 984/1000
-60000/60000 - 6s - loss: 0.0472 - accuracy: 0.9882 - val_loss: 0.3649 - val_accuracy: 0.9218
-Epoch 985/1000
-60000/60000 - 6s - loss: 0.0471 - accuracy: 0.9882 - val_loss: 0.3639 - val_accuracy: 0.9215
-Epoch 986/1000
-60000/60000 - 6s - loss: 0.0470 - accuracy: 0.9883 - val_loss: 0.3657 - val_accuracy: 0.9221
-Epoch 987/1000
-60000/60000 - 6s - loss: 0.0470 - accuracy: 0.9883 - val_loss: 0.3638 - val_accuracy: 0.9222
-Epoch 988/1000
-60000/60000 - 6s - loss: 0.0472 - accuracy: 0.9882 - val_loss: 0.3640 - val_accuracy: 0.9214
-Epoch 989/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9879 - val_loss: 0.3640 - val_accuracy: 0.9225
-Epoch 990/1000
-60000/60000 - 6s - loss: 0.0481 - accuracy: 0.9878 - val_loss: 0.3649 - val_accuracy: 0.9224
-Epoch 991/1000
-60000/60000 - 6s - loss: 0.0476 - accuracy: 0.9881 - val_loss: 0.3628 - val_accuracy: 0.9227
-Epoch 992/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9882 - val_loss: 0.3646 - val_accuracy: 0.9228
-Epoch 993/1000
-60000/60000 - 6s - loss: 0.0496 - accuracy: 0.9875 - val_loss: 0.3657 - val_accuracy: 0.9223
-Epoch 994/1000
-60000/60000 - 6s - loss: 0.0484 - accuracy: 0.9879 - val_loss: 0.3606 - val_accuracy: 0.9233
-Epoch 995/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9881 - val_loss: 0.3625 - val_accuracy: 0.9228
-Epoch 996/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9882 - val_loss: 0.3632 - val_accuracy: 0.9222
-Epoch 997/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9881 - val_loss: 0.3634 - val_accuracy: 0.9225
-Epoch 998/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9880 - val_loss: 0.3660 - val_accuracy: 0.9220
-Epoch 999/1000
-60000/60000 - 6s - loss: 0.0470 - accuracy: 0.9882 - val_loss: 0.3632 - val_accuracy: 0.9219
-Epoch 1000/1000
-60000/60000 - 6s - loss: 0.0473 - accuracy: 0.9881 - val_loss: 0.3640 - val_accuracy: 0.9218
-Test loss was 0.2384, test accuracy was 0.9324
diff --git a/MNIST/nonlinear_overriden.txt b/MNIST/nonlinear_overriden.txt
deleted file mode 100644
index cb685cf..0000000
--- a/MNIST/nonlinear_overriden.txt
+++ /dev/null
@@ -1,2005 +0,0 @@
-X_train shape: (60000, 784)
-60000 train samples
-10000 test samples
-Train on 60000 samples, validate on 10000 samples
-Epoch 1/1000
-60000/60000 - 7s - loss: 1.6067 - accuracy: 0.4892 - val_loss: 1.0555 - val_accuracy: 0.6518
-Epoch 2/1000
-60000/60000 - 5s - loss: 0.9542 - accuracy: 0.6926 - val_loss: 0.8209 - val_accuracy: 0.7388
-Epoch 3/1000
-60000/60000 - 5s - loss: 0.7931 - accuracy: 0.7519 - val_loss: 0.7211 - val_accuracy: 0.7768
-Epoch 4/1000
-60000/60000 - 5s - loss: 0.7119 - accuracy: 0.7799 - val_loss: 0.6585 - val_accuracy: 0.7962
-Epoch 5/1000
-60000/60000 - 5s - loss: 0.6562 - accuracy: 0.7993 - val_loss: 0.6167 - val_accuracy: 0.8100
-Epoch 6/1000
-60000/60000 - 5s - loss: 0.6164 - accuracy: 0.8131 - val_loss: 0.5879 - val_accuracy: 0.8192
-Epoch 7/1000
-60000/60000 - 5s - loss: 0.5865 - accuracy: 0.8229 - val_loss: 0.5626 - val_accuracy: 0.8295
-Epoch 8/1000
-60000/60000 - 5s - loss: 0.5613 - accuracy: 0.8310 - val_loss: 0.5442 - val_accuracy: 0.8345
-Epoch 9/1000
-60000/60000 - 5s - loss: 0.5401 - accuracy: 0.8369 - val_loss: 0.5268 - val_accuracy: 0.8410
-Epoch 10/1000
-60000/60000 - 5s - loss: 0.5233 - accuracy: 0.8430 - val_loss: 0.5132 - val_accuracy: 0.8465
-Epoch 11/1000
-60000/60000 - 5s - loss: 0.5088 - accuracy: 0.8470 - val_loss: 0.5012 - val_accuracy: 0.8513
-Epoch 12/1000
-60000/60000 - 5s - loss: 0.4961 - accuracy: 0.8516 - val_loss: 0.4918 - val_accuracy: 0.8537
-Epoch 13/1000
-60000/60000 - 5s - loss: 0.4834 - accuracy: 0.8553 - val_loss: 0.4824 - val_accuracy: 0.8583
-Epoch 14/1000
-60000/60000 - 5s - loss: 0.4725 - accuracy: 0.8595 - val_loss: 0.4759 - val_accuracy: 0.8608
-Epoch 15/1000
-60000/60000 - 5s - loss: 0.4628 - accuracy: 0.8630 - val_loss: 0.4689 - val_accuracy: 0.8620
-Epoch 16/1000
-60000/60000 - 5s - loss: 0.4538 - accuracy: 0.8656 - val_loss: 0.4608 - val_accuracy: 0.8653
-Epoch 17/1000
-60000/60000 - 5s - loss: 0.4449 - accuracy: 0.8685 - val_loss: 0.4554 - val_accuracy: 0.8664
-Epoch 18/1000
-60000/60000 - 5s - loss: 0.4370 - accuracy: 0.8709 - val_loss: 0.4507 - val_accuracy: 0.8684
-Epoch 19/1000
-60000/60000 - 5s - loss: 0.4301 - accuracy: 0.8733 - val_loss: 0.4446 - val_accuracy: 0.8707
-Epoch 20/1000
-60000/60000 - 5s - loss: 0.4227 - accuracy: 0.8755 - val_loss: 0.4411 - val_accuracy: 0.8708
-Epoch 21/1000
-60000/60000 - 5s - loss: 0.4174 - accuracy: 0.8772 - val_loss: 0.4357 - val_accuracy: 0.8719
-Epoch 22/1000
-60000/60000 - 5s - loss: 0.4105 - accuracy: 0.8791 - val_loss: 0.4326 - val_accuracy: 0.8729
-Epoch 23/1000
-60000/60000 - 5s - loss: 0.4048 - accuracy: 0.8807 - val_loss: 0.4288 - val_accuracy: 0.8753
-Epoch 24/1000
-60000/60000 - 5s - loss: 0.3997 - accuracy: 0.8831 - val_loss: 0.4252 - val_accuracy: 0.8759
-Epoch 25/1000
-60000/60000 - 5s - loss: 0.3945 - accuracy: 0.8845 - val_loss: 0.4199 - val_accuracy: 0.8783
-Epoch 26/1000
-60000/60000 - 5s - loss: 0.3886 - accuracy: 0.8863 - val_loss: 0.4161 - val_accuracy: 0.8801
-Epoch 27/1000
-60000/60000 - 5s - loss: 0.3838 - accuracy: 0.8875 - val_loss: 0.4134 - val_accuracy: 0.8806
-Epoch 28/1000
-60000/60000 - 5s - loss: 0.3792 - accuracy: 0.8889 - val_loss: 0.4093 - val_accuracy: 0.8800
-Epoch 29/1000
-60000/60000 - 5s - loss: 0.3749 - accuracy: 0.8904 - val_loss: 0.4050 - val_accuracy: 0.8824
-Epoch 30/1000
-60000/60000 - 5s - loss: 0.3702 - accuracy: 0.8917 - val_loss: 0.4023 - val_accuracy: 0.8830
-Epoch 31/1000
-60000/60000 - 5s - loss: 0.3662 - accuracy: 0.8929 - val_loss: 0.4001 - val_accuracy: 0.8850
-Epoch 32/1000
-60000/60000 - 5s - loss: 0.3623 - accuracy: 0.8939 - val_loss: 0.3966 - val_accuracy: 0.8855
-Epoch 33/1000
-60000/60000 - 5s - loss: 0.3586 - accuracy: 0.8949 - val_loss: 0.3926 - val_accuracy: 0.8872
-Epoch 34/1000
-60000/60000 - 5s - loss: 0.3546 - accuracy: 0.8962 - val_loss: 0.3888 - val_accuracy: 0.8876
-Epoch 35/1000
-60000/60000 - 5s - loss: 0.3503 - accuracy: 0.8978 - val_loss: 0.3873 - val_accuracy: 0.8880
-Epoch 36/1000
-60000/60000 - 5s - loss: 0.3475 - accuracy: 0.8984 - val_loss: 0.3845 - val_accuracy: 0.8887
-Epoch 37/1000
-60000/60000 - 5s - loss: 0.3439 - accuracy: 0.8998 - val_loss: 0.3818 - val_accuracy: 0.8896
-Epoch 38/1000
-60000/60000 - 5s - loss: 0.3404 - accuracy: 0.9010 - val_loss: 0.3803 - val_accuracy: 0.8899
-Epoch 39/1000
-60000/60000 - 5s - loss: 0.3377 - accuracy: 0.9016 - val_loss: 0.3770 - val_accuracy: 0.8906
-Epoch 40/1000
-60000/60000 - 5s - loss: 0.3344 - accuracy: 0.9026 - val_loss: 0.3743 - val_accuracy: 0.8910
-Epoch 41/1000
-60000/60000 - 5s - loss: 0.3318 - accuracy: 0.9031 - val_loss: 0.3722 - val_accuracy: 0.8937
-Epoch 42/1000
-60000/60000 - 5s - loss: 0.3290 - accuracy: 0.9044 - val_loss: 0.3710 - val_accuracy: 0.8942
-Epoch 43/1000
-60000/60000 - 5s - loss: 0.3266 - accuracy: 0.9051 - val_loss: 0.3687 - val_accuracy: 0.8935
-Epoch 44/1000
-60000/60000 - 5s - loss: 0.3240 - accuracy: 0.9060 - val_loss: 0.3691 - val_accuracy: 0.8945
-Epoch 45/1000
-60000/60000 - 5s - loss: 0.3217 - accuracy: 0.9066 - val_loss: 0.3652 - val_accuracy: 0.8949
-Epoch 46/1000
-60000/60000 - 5s - loss: 0.3193 - accuracy: 0.9075 - val_loss: 0.3645 - val_accuracy: 0.8950
-Epoch 47/1000
-60000/60000 - 5s - loss: 0.3171 - accuracy: 0.9081 - val_loss: 0.3622 - val_accuracy: 0.8958
-Epoch 48/1000
-60000/60000 - 5s - loss: 0.3146 - accuracy: 0.9090 - val_loss: 0.3601 - val_accuracy: 0.8962
-Epoch 49/1000
-60000/60000 - 5s - loss: 0.3125 - accuracy: 0.9098 - val_loss: 0.3578 - val_accuracy: 0.8967
-Epoch 50/1000
-60000/60000 - 5s - loss: 0.3098 - accuracy: 0.9110 - val_loss: 0.3576 - val_accuracy: 0.8980
-Epoch 51/1000
-60000/60000 - 5s - loss: 0.3076 - accuracy: 0.9113 - val_loss: 0.3563 - val_accuracy: 0.8980
-Epoch 52/1000
-60000/60000 - 5s - loss: 0.3059 - accuracy: 0.9120 - val_loss: 0.3542 - val_accuracy: 0.8997
-Epoch 53/1000
-60000/60000 - 5s - loss: 0.3036 - accuracy: 0.9131 - val_loss: 0.3514 - val_accuracy: 0.8985
-Epoch 54/1000
-60000/60000 - 5s - loss: 0.3010 - accuracy: 0.9136 - val_loss: 0.3511 - val_accuracy: 0.8987
-Epoch 55/1000
-60000/60000 - 5s - loss: 0.2994 - accuracy: 0.9138 - val_loss: 0.3483 - val_accuracy: 0.9006
-Epoch 56/1000
-60000/60000 - 5s - loss: 0.2974 - accuracy: 0.9149 - val_loss: 0.3471 - val_accuracy: 0.9009
-Epoch 57/1000
-60000/60000 - 5s - loss: 0.2952 - accuracy: 0.9152 - val_loss: 0.3463 - val_accuracy: 0.9006
-Epoch 58/1000
-60000/60000 - 5s - loss: 0.2937 - accuracy: 0.9158 - val_loss: 0.3437 - val_accuracy: 0.9024
-Epoch 59/1000
-60000/60000 - 5s - loss: 0.2918 - accuracy: 0.9164 - val_loss: 0.3428 - val_accuracy: 0.9013
-Epoch 60/1000
-60000/60000 - 5s - loss: 0.2901 - accuracy: 0.9169 - val_loss: 0.3405 - val_accuracy: 0.9018
-Epoch 61/1000
-60000/60000 - 5s - loss: 0.2881 - accuracy: 0.9176 - val_loss: 0.3394 - val_accuracy: 0.9023
-Epoch 62/1000
-60000/60000 - 5s - loss: 0.2863 - accuracy: 0.9178 - val_loss: 0.3392 - val_accuracy: 0.9022
-Epoch 63/1000
-60000/60000 - 5s - loss: 0.2852 - accuracy: 0.9182 - val_loss: 0.3372 - val_accuracy: 0.9032
-Epoch 64/1000
-60000/60000 - 5s - loss: 0.2832 - accuracy: 0.9190 - val_loss: 0.3357 - val_accuracy: 0.9036
-Epoch 65/1000
-60000/60000 - 5s - loss: 0.2814 - accuracy: 0.9190 - val_loss: 0.3333 - val_accuracy: 0.9051
-Epoch 66/1000
-60000/60000 - 5s - loss: 0.2797 - accuracy: 0.9195 - val_loss: 0.3324 - val_accuracy: 0.9037
-Epoch 67/1000
-60000/60000 - 5s - loss: 0.2781 - accuracy: 0.9198 - val_loss: 0.3322 - val_accuracy: 0.9049
-Epoch 68/1000
-60000/60000 - 5s - loss: 0.2768 - accuracy: 0.9203 - val_loss: 0.3293 - val_accuracy: 0.9064
-Epoch 69/1000
-60000/60000 - 5s - loss: 0.2749 - accuracy: 0.9209 - val_loss: 0.3291 - val_accuracy: 0.9068
-Epoch 70/1000
-60000/60000 - 5s - loss: 0.2736 - accuracy: 0.9220 - val_loss: 0.3267 - val_accuracy: 0.9076
-Epoch 71/1000
-60000/60000 - 5s - loss: 0.2717 - accuracy: 0.9213 - val_loss: 0.3258 - val_accuracy: 0.9070
-Epoch 72/1000
-60000/60000 - 5s - loss: 0.2704 - accuracy: 0.9217 - val_loss: 0.3261 - val_accuracy: 0.9068
-Epoch 73/1000
-60000/60000 - 5s - loss: 0.2692 - accuracy: 0.9220 - val_loss: 0.3239 - val_accuracy: 0.9081
-Epoch 74/1000
-60000/60000 - 5s - loss: 0.2678 - accuracy: 0.9227 - val_loss: 0.3220 - val_accuracy: 0.9083
-Epoch 75/1000
-60000/60000 - 5s - loss: 0.2664 - accuracy: 0.9228 - val_loss: 0.3214 - val_accuracy: 0.9086
-Epoch 76/1000
-60000/60000 - 5s - loss: 0.2647 - accuracy: 0.9235 - val_loss: 0.3202 - val_accuracy: 0.9091
-Epoch 77/1000
-60000/60000 - 5s - loss: 0.2636 - accuracy: 0.9235 - val_loss: 0.3205 - val_accuracy: 0.9081
-Epoch 78/1000
-60000/60000 - 5s - loss: 0.2621 - accuracy: 0.9241 - val_loss: 0.3187 - val_accuracy: 0.9082
-Epoch 79/1000
-60000/60000 - 5s - loss: 0.2606 - accuracy: 0.9247 - val_loss: 0.3176 - val_accuracy: 0.9097
-Epoch 80/1000
-60000/60000 - 5s - loss: 0.2597 - accuracy: 0.9249 - val_loss: 0.3169 - val_accuracy: 0.9091
-Epoch 81/1000
-60000/60000 - 5s - loss: 0.2582 - accuracy: 0.9254 - val_loss: 0.3161 - val_accuracy: 0.9092
-Epoch 82/1000
-60000/60000 - 5s - loss: 0.2572 - accuracy: 0.9255 - val_loss: 0.3160 - val_accuracy: 0.9097
-Epoch 83/1000
-60000/60000 - 5s - loss: 0.2558 - accuracy: 0.9258 - val_loss: 0.3154 - val_accuracy: 0.9098
-Epoch 84/1000
-60000/60000 - 5s - loss: 0.2550 - accuracy: 0.9264 - val_loss: 0.3142 - val_accuracy: 0.9097
-Epoch 85/1000
-60000/60000 - 5s - loss: 0.2534 - accuracy: 0.9270 - val_loss: 0.3141 - val_accuracy: 0.9100
-Epoch 86/1000
-60000/60000 - 5s - loss: 0.2522 - accuracy: 0.9269 - val_loss: 0.3127 - val_accuracy: 0.9104
-Epoch 87/1000
-60000/60000 - 5s - loss: 0.2514 - accuracy: 0.9274 - val_loss: 0.3120 - val_accuracy: 0.9103
-Epoch 88/1000
-60000/60000 - 5s - loss: 0.2501 - accuracy: 0.9281 - val_loss: 0.3118 - val_accuracy: 0.9107
-Epoch 89/1000
-60000/60000 - 5s - loss: 0.2490 - accuracy: 0.9284 - val_loss: 0.3107 - val_accuracy: 0.9108
-Epoch 90/1000
-60000/60000 - 5s - loss: 0.2477 - accuracy: 0.9284 - val_loss: 0.3103 - val_accuracy: 0.9112
-Epoch 91/1000
-60000/60000 - 5s - loss: 0.2468 - accuracy: 0.9290 - val_loss: 0.3101 - val_accuracy: 0.9117
-Epoch 92/1000
-60000/60000 - 5s - loss: 0.2459 - accuracy: 0.9295 - val_loss: 0.3089 - val_accuracy: 0.9124
-Epoch 93/1000
-60000/60000 - 5s - loss: 0.2445 - accuracy: 0.9299 - val_loss: 0.3065 - val_accuracy: 0.9132
-Epoch 94/1000
-60000/60000 - 5s - loss: 0.2436 - accuracy: 0.9301 - val_loss: 0.3068 - val_accuracy: 0.9120
-Epoch 95/1000
-60000/60000 - 5s - loss: 0.2425 - accuracy: 0.9304 - val_loss: 0.3071 - val_accuracy: 0.9122
-Epoch 96/1000
-60000/60000 - 5s - loss: 0.2416 - accuracy: 0.9309 - val_loss: 0.3060 - val_accuracy: 0.9129
-Epoch 97/1000
-60000/60000 - 5s - loss: 0.2407 - accuracy: 0.9313 - val_loss: 0.3061 - val_accuracy: 0.9131
-Epoch 98/1000
-60000/60000 - 5s - loss: 0.2397 - accuracy: 0.9314 - val_loss: 0.3054 - val_accuracy: 0.9127
-Epoch 99/1000
-60000/60000 - 5s - loss: 0.2387 - accuracy: 0.9323 - val_loss: 0.3040 - val_accuracy: 0.9134
-Epoch 100/1000
-60000/60000 - 5s - loss: 0.2377 - accuracy: 0.9326 - val_loss: 0.3049 - val_accuracy: 0.9131
-Epoch 101/1000
-60000/60000 - 5s - loss: 0.2369 - accuracy: 0.9328 - val_loss: 0.3041 - val_accuracy: 0.9145
-Epoch 102/1000
-60000/60000 - 5s - loss: 0.2359 - accuracy: 0.9332 - val_loss: 0.3037 - val_accuracy: 0.9142
-Epoch 103/1000
-60000/60000 - 5s - loss: 0.2350 - accuracy: 0.9335 - val_loss: 0.3030 - val_accuracy: 0.9151
-Epoch 104/1000
-60000/60000 - 5s - loss: 0.2339 - accuracy: 0.9338 - val_loss: 0.3027 - val_accuracy: 0.9149
-Epoch 105/1000
-60000/60000 - 5s - loss: 0.2331 - accuracy: 0.9340 - val_loss: 0.3025 - val_accuracy: 0.9150
-Epoch 106/1000
-60000/60000 - 5s - loss: 0.2321 - accuracy: 0.9343 - val_loss: 0.3025 - val_accuracy: 0.9147
-Epoch 107/1000
-60000/60000 - 5s - loss: 0.2312 - accuracy: 0.9346 - val_loss: 0.3023 - val_accuracy: 0.9154
-Epoch 108/1000
-60000/60000 - 5s - loss: 0.2303 - accuracy: 0.9347 - val_loss: 0.3016 - val_accuracy: 0.9161
-Epoch 109/1000
-60000/60000 - 5s - loss: 0.2295 - accuracy: 0.9352 - val_loss: 0.3012 - val_accuracy: 0.9150
-Epoch 110/1000
-60000/60000 - 5s - loss: 0.2288 - accuracy: 0.9355 - val_loss: 0.3021 - val_accuracy: 0.9147
-Epoch 111/1000
-60000/60000 - 5s - loss: 0.2282 - accuracy: 0.9354 - val_loss: 0.3007 - val_accuracy: 0.9156
-Epoch 112/1000
-60000/60000 - 5s - loss: 0.2271 - accuracy: 0.9359 - val_loss: 0.3009 - val_accuracy: 0.9162
-Epoch 113/1000
-60000/60000 - 5s - loss: 0.2264 - accuracy: 0.9360 - val_loss: 0.3003 - val_accuracy: 0.9164
-Epoch 114/1000
-60000/60000 - 5s - loss: 0.2257 - accuracy: 0.9361 - val_loss: 0.2998 - val_accuracy: 0.9174
-Epoch 115/1000
-60000/60000 - 5s - loss: 0.2249 - accuracy: 0.9363 - val_loss: 0.2983 - val_accuracy: 0.9168
-Epoch 116/1000
-60000/60000 - 5s - loss: 0.2242 - accuracy: 0.9366 - val_loss: 0.2995 - val_accuracy: 0.9159
-Epoch 117/1000
-60000/60000 - 5s - loss: 0.2232 - accuracy: 0.9371 - val_loss: 0.2988 - val_accuracy: 0.9162
-Epoch 118/1000
-60000/60000 - 5s - loss: 0.2226 - accuracy: 0.9376 - val_loss: 0.2988 - val_accuracy: 0.9163
-Epoch 119/1000
-60000/60000 - 5s - loss: 0.2214 - accuracy: 0.9374 - val_loss: 0.2988 - val_accuracy: 0.9163
-Epoch 120/1000
-60000/60000 - 5s - loss: 0.2211 - accuracy: 0.9377 - val_loss: 0.2990 - val_accuracy: 0.9167
-Epoch 121/1000
-60000/60000 - 5s - loss: 0.2202 - accuracy: 0.9377 - val_loss: 0.2982 - val_accuracy: 0.9161
-Epoch 122/1000
-60000/60000 - 5s - loss: 0.2197 - accuracy: 0.9384 - val_loss: 0.2977 - val_accuracy: 0.9170
-Epoch 123/1000
-60000/60000 - 5s - loss: 0.2188 - accuracy: 0.9385 - val_loss: 0.2973 - val_accuracy: 0.9169
-Epoch 124/1000
-60000/60000 - 5s - loss: 0.2180 - accuracy: 0.9387 - val_loss: 0.2969 - val_accuracy: 0.9171
-Epoch 125/1000
-60000/60000 - 5s - loss: 0.2176 - accuracy: 0.9391 - val_loss: 0.2967 - val_accuracy: 0.9168
-Epoch 126/1000
-60000/60000 - 5s - loss: 0.2166 - accuracy: 0.9396 - val_loss: 0.2961 - val_accuracy: 0.9168
-Epoch 127/1000
-60000/60000 - 5s - loss: 0.2161 - accuracy: 0.9397 - val_loss: 0.2962 - val_accuracy: 0.9167
-Epoch 128/1000
-60000/60000 - 5s - loss: 0.2153 - accuracy: 0.9397 - val_loss: 0.2955 - val_accuracy: 0.9171
-Epoch 129/1000
-60000/60000 - 5s - loss: 0.2146 - accuracy: 0.9401 - val_loss: 0.2955 - val_accuracy: 0.9175
-Epoch 130/1000
-60000/60000 - 5s - loss: 0.2140 - accuracy: 0.9403 - val_loss: 0.2946 - val_accuracy: 0.9164
-Epoch 131/1000
-60000/60000 - 5s - loss: 0.2132 - accuracy: 0.9404 - val_loss: 0.2950 - val_accuracy: 0.9175
-Epoch 132/1000
-60000/60000 - 5s - loss: 0.2126 - accuracy: 0.9410 - val_loss: 0.2950 - val_accuracy: 0.9172
-Epoch 133/1000
-60000/60000 - 5s - loss: 0.2120 - accuracy: 0.9410 - val_loss: 0.2956 - val_accuracy: 0.9180
-Epoch 134/1000
-60000/60000 - 5s - loss: 0.2113 - accuracy: 0.9415 - val_loss: 0.2950 - val_accuracy: 0.9180
-Epoch 135/1000
-60000/60000 - 5s - loss: 0.2106 - accuracy: 0.9417 - val_loss: 0.2955 - val_accuracy: 0.9169
-Epoch 136/1000
-60000/60000 - 5s - loss: 0.2100 - accuracy: 0.9420 - val_loss: 0.2957 - val_accuracy: 0.9181
-Epoch 137/1000
-60000/60000 - 5s - loss: 0.2093 - accuracy: 0.9425 - val_loss: 0.2948 - val_accuracy: 0.9189
-Epoch 138/1000
-60000/60000 - 5s - loss: 0.2086 - accuracy: 0.9422 - val_loss: 0.2946 - val_accuracy: 0.9174
-Epoch 139/1000
-60000/60000 - 5s - loss: 0.2080 - accuracy: 0.9425 - val_loss: 0.2950 - val_accuracy: 0.9171
-Epoch 140/1000
-60000/60000 - 5s - loss: 0.2075 - accuracy: 0.9427 - val_loss: 0.2948 - val_accuracy: 0.9173
-Epoch 141/1000
-60000/60000 - 5s - loss: 0.2068 - accuracy: 0.9433 - val_loss: 0.2955 - val_accuracy: 0.9178
-Epoch 142/1000
-60000/60000 - 5s - loss: 0.2064 - accuracy: 0.9431 - val_loss: 0.2953 - val_accuracy: 0.9175
-Epoch 143/1000
-60000/60000 - 5s - loss: 0.2058 - accuracy: 0.9433 - val_loss: 0.2952 - val_accuracy: 0.9175
-Epoch 144/1000
-60000/60000 - 5s - loss: 0.2052 - accuracy: 0.9437 - val_loss: 0.2963 - val_accuracy: 0.9172
-Epoch 145/1000
-60000/60000 - 5s - loss: 0.2047 - accuracy: 0.9437 - val_loss: 0.2952 - val_accuracy: 0.9178
-Epoch 146/1000
-60000/60000 - 5s - loss: 0.2039 - accuracy: 0.9436 - val_loss: 0.2951 - val_accuracy: 0.9177
-Epoch 147/1000
-60000/60000 - 5s - loss: 0.2033 - accuracy: 0.9439 - val_loss: 0.2952 - val_accuracy: 0.9181
-Epoch 148/1000
-60000/60000 - 5s - loss: 0.2029 - accuracy: 0.9441 - val_loss: 0.2944 - val_accuracy: 0.9180
-Epoch 149/1000
-60000/60000 - 5s - loss: 0.2024 - accuracy: 0.9444 - val_loss: 0.2953 - val_accuracy: 0.9178
-Epoch 150/1000
-60000/60000 - 5s - loss: 0.2020 - accuracy: 0.9443 - val_loss: 0.2943 - val_accuracy: 0.9182
-Epoch 151/1000
-60000/60000 - 5s - loss: 0.2013 - accuracy: 0.9445 - val_loss: 0.2952 - val_accuracy: 0.9186
-Epoch 152/1000
-60000/60000 - 5s - loss: 0.2006 - accuracy: 0.9449 - val_loss: 0.2958 - val_accuracy: 0.9181
-Epoch 153/1000
-60000/60000 - 5s - loss: 0.2001 - accuracy: 0.9448 - val_loss: 0.2956 - val_accuracy: 0.9179
-Epoch 154/1000
-60000/60000 - 5s - loss: 0.1997 - accuracy: 0.9450 - val_loss: 0.2950 - val_accuracy: 0.9177
-Epoch 155/1000
-60000/60000 - 5s - loss: 0.1991 - accuracy: 0.9451 - val_loss: 0.2951 - val_accuracy: 0.9180
-Epoch 156/1000
-60000/60000 - 5s - loss: 0.1986 - accuracy: 0.9455 - val_loss: 0.2956 - val_accuracy: 0.9174
-Epoch 157/1000
-60000/60000 - 5s - loss: 0.1983 - accuracy: 0.9455 - val_loss: 0.2949 - val_accuracy: 0.9187
-Epoch 158/1000
-60000/60000 - 5s - loss: 0.1975 - accuracy: 0.9456 - val_loss: 0.2963 - val_accuracy: 0.9176
-Epoch 159/1000
-60000/60000 - 5s - loss: 0.1970 - accuracy: 0.9459 - val_loss: 0.2952 - val_accuracy: 0.9190
-Epoch 160/1000
-60000/60000 - 5s - loss: 0.1967 - accuracy: 0.9459 - val_loss: 0.2947 - val_accuracy: 0.9182
-Epoch 161/1000
-60000/60000 - 5s - loss: 0.1961 - accuracy: 0.9461 - val_loss: 0.2955 - val_accuracy: 0.9186
-Epoch 162/1000
-60000/60000 - 5s - loss: 0.1957 - accuracy: 0.9462 - val_loss: 0.2959 - val_accuracy: 0.9174
-Epoch 163/1000
-60000/60000 - 5s - loss: 0.1953 - accuracy: 0.9462 - val_loss: 0.2953 - val_accuracy: 0.9188
-Epoch 164/1000
-60000/60000 - 5s - loss: 0.1947 - accuracy: 0.9466 - val_loss: 0.2952 - val_accuracy: 0.9192
-Epoch 165/1000
-60000/60000 - 5s - loss: 0.1941 - accuracy: 0.9470 - val_loss: 0.2955 - val_accuracy: 0.9177
-Epoch 166/1000
-60000/60000 - 5s - loss: 0.1937 - accuracy: 0.9468 - val_loss: 0.2959 - val_accuracy: 0.9181
-Epoch 167/1000
-60000/60000 - 5s - loss: 0.1933 - accuracy: 0.9468 - val_loss: 0.2954 - val_accuracy: 0.9179
-Epoch 168/1000
-60000/60000 - 5s - loss: 0.1930 - accuracy: 0.9468 - val_loss: 0.2961 - val_accuracy: 0.9180
-Epoch 169/1000
-60000/60000 - 5s - loss: 0.1925 - accuracy: 0.9472 - val_loss: 0.2950 - val_accuracy: 0.9194
-Epoch 170/1000
-60000/60000 - 5s - loss: 0.1919 - accuracy: 0.9475 - val_loss: 0.2953 - val_accuracy: 0.9183
-Epoch 171/1000
-60000/60000 - 5s - loss: 0.1915 - accuracy: 0.9475 - val_loss: 0.2959 - val_accuracy: 0.9179
-Epoch 172/1000
-60000/60000 - 5s - loss: 0.1914 - accuracy: 0.9475 - val_loss: 0.2961 - val_accuracy: 0.9184
-Epoch 173/1000
-60000/60000 - 5s - loss: 0.1909 - accuracy: 0.9478 - val_loss: 0.2955 - val_accuracy: 0.9180
-Epoch 174/1000
-60000/60000 - 5s - loss: 0.1903 - accuracy: 0.9479 - val_loss: 0.2955 - val_accuracy: 0.9191
-Epoch 175/1000
-60000/60000 - 5s - loss: 0.1899 - accuracy: 0.9483 - val_loss: 0.2957 - val_accuracy: 0.9187
-Epoch 176/1000
-60000/60000 - 5s - loss: 0.1894 - accuracy: 0.9483 - val_loss: 0.2960 - val_accuracy: 0.9189
-Epoch 177/1000
-60000/60000 - 5s - loss: 0.1890 - accuracy: 0.9482 - val_loss: 0.2959 - val_accuracy: 0.9190
-Epoch 178/1000
-60000/60000 - 5s - loss: 0.1885 - accuracy: 0.9486 - val_loss: 0.2966 - val_accuracy: 0.9178
-Epoch 179/1000
-60000/60000 - 5s - loss: 0.1883 - accuracy: 0.9485 - val_loss: 0.2962 - val_accuracy: 0.9183
-Epoch 180/1000
-60000/60000 - 5s - loss: 0.1878 - accuracy: 0.9485 - val_loss: 0.2953 - val_accuracy: 0.9195
-Epoch 181/1000
-60000/60000 - 5s - loss: 0.1876 - accuracy: 0.9487 - val_loss: 0.2948 - val_accuracy: 0.9187
-Epoch 182/1000
-60000/60000 - 5s - loss: 0.1870 - accuracy: 0.9491 - val_loss: 0.2954 - val_accuracy: 0.9192
-Epoch 183/1000
-60000/60000 - 5s - loss: 0.1866 - accuracy: 0.9488 - val_loss: 0.2961 - val_accuracy: 0.9192
-Epoch 184/1000
-60000/60000 - 5s - loss: 0.1863 - accuracy: 0.9488 - val_loss: 0.2950 - val_accuracy: 0.9196
-Epoch 185/1000
-60000/60000 - 5s - loss: 0.1858 - accuracy: 0.9493 - val_loss: 0.2957 - val_accuracy: 0.9188
-Epoch 186/1000
-60000/60000 - 5s - loss: 0.1856 - accuracy: 0.9491 - val_loss: 0.2952 - val_accuracy: 0.9191
-Epoch 187/1000
-60000/60000 - 5s - loss: 0.1851 - accuracy: 0.9494 - val_loss: 0.2946 - val_accuracy: 0.9192
-Epoch 188/1000
-60000/60000 - 5s - loss: 0.1847 - accuracy: 0.9493 - val_loss: 0.2951 - val_accuracy: 0.9197
-Epoch 189/1000
-60000/60000 - 5s - loss: 0.1843 - accuracy: 0.9497 - val_loss: 0.2943 - val_accuracy: 0.9192
-Epoch 190/1000
-60000/60000 - 5s - loss: 0.1838 - accuracy: 0.9499 - val_loss: 0.2957 - val_accuracy: 0.9181
-Epoch 191/1000
-60000/60000 - 5s - loss: 0.1835 - accuracy: 0.9500 - val_loss: 0.2946 - val_accuracy: 0.9185
-Epoch 192/1000
-60000/60000 - 5s - loss: 0.1831 - accuracy: 0.9499 - val_loss: 0.2947 - val_accuracy: 0.9188
-Epoch 193/1000
-60000/60000 - 5s - loss: 0.1826 - accuracy: 0.9503 - val_loss: 0.2951 - val_accuracy: 0.9187
-Epoch 194/1000
-60000/60000 - 5s - loss: 0.1823 - accuracy: 0.9500 - val_loss: 0.2955 - val_accuracy: 0.9187
-Epoch 195/1000
-60000/60000 - 5s - loss: 0.1821 - accuracy: 0.9505 - val_loss: 0.2949 - val_accuracy: 0.9183
-Epoch 196/1000
-60000/60000 - 5s - loss: 0.1816 - accuracy: 0.9506 - val_loss: 0.2943 - val_accuracy: 0.9197
-Epoch 197/1000
-60000/60000 - 5s - loss: 0.1812 - accuracy: 0.9501 - val_loss: 0.2951 - val_accuracy: 0.9194
-Epoch 198/1000
-60000/60000 - 5s - loss: 0.1808 - accuracy: 0.9505 - val_loss: 0.2951 - val_accuracy: 0.9186
-Epoch 199/1000
-60000/60000 - 5s - loss: 0.1805 - accuracy: 0.9512 - val_loss: 0.2943 - val_accuracy: 0.9186
-Epoch 200/1000
-60000/60000 - 5s - loss: 0.1801 - accuracy: 0.9510 - val_loss: 0.2960 - val_accuracy: 0.9182
-Epoch 201/1000
-60000/60000 - 5s - loss: 0.1798 - accuracy: 0.9506 - val_loss: 0.2951 - val_accuracy: 0.9187
-Epoch 202/1000
-60000/60000 - 5s - loss: 0.1795 - accuracy: 0.9509 - val_loss: 0.2949 - val_accuracy: 0.9186
-Epoch 203/1000
-60000/60000 - 5s - loss: 0.1791 - accuracy: 0.9511 - val_loss: 0.2949 - val_accuracy: 0.9189
-Epoch 204/1000
-60000/60000 - 5s - loss: 0.1787 - accuracy: 0.9513 - val_loss: 0.2947 - val_accuracy: 0.9185
-Epoch 205/1000
-60000/60000 - 5s - loss: 0.1784 - accuracy: 0.9512 - val_loss: 0.2949 - val_accuracy: 0.9187
-Epoch 206/1000
-60000/60000 - 5s - loss: 0.1780 - accuracy: 0.9515 - val_loss: 0.2951 - val_accuracy: 0.9183
-Epoch 207/1000
-60000/60000 - 5s - loss: 0.1775 - accuracy: 0.9516 - val_loss: 0.2943 - val_accuracy: 0.9190
-Epoch 208/1000
-60000/60000 - 5s - loss: 0.1774 - accuracy: 0.9513 - val_loss: 0.2945 - val_accuracy: 0.9187
-Epoch 209/1000
-60000/60000 - 5s - loss: 0.1770 - accuracy: 0.9513 - val_loss: 0.2959 - val_accuracy: 0.9186
-Epoch 210/1000
-60000/60000 - 5s - loss: 0.1767 - accuracy: 0.9515 - val_loss: 0.2959 - val_accuracy: 0.9187
-Epoch 211/1000
-60000/60000 - 5s - loss: 0.1764 - accuracy: 0.9517 - val_loss: 0.2955 - val_accuracy: 0.9187
-Epoch 212/1000
-60000/60000 - 5s - loss: 0.1759 - accuracy: 0.9518 - val_loss: 0.2945 - val_accuracy: 0.9191
-Epoch 213/1000
-60000/60000 - 5s - loss: 0.1758 - accuracy: 0.9519 - val_loss: 0.2955 - val_accuracy: 0.9180
-Epoch 214/1000
-60000/60000 - 5s - loss: 0.1752 - accuracy: 0.9519 - val_loss: 0.2953 - val_accuracy: 0.9186
-Epoch 215/1000
-60000/60000 - 5s - loss: 0.1749 - accuracy: 0.9518 - val_loss: 0.2958 - val_accuracy: 0.9187
-Epoch 216/1000
-60000/60000 - 5s - loss: 0.1746 - accuracy: 0.9523 - val_loss: 0.2947 - val_accuracy: 0.9192
-Epoch 217/1000
-60000/60000 - 5s - loss: 0.1742 - accuracy: 0.9520 - val_loss: 0.2953 - val_accuracy: 0.9195
-Epoch 218/1000
-60000/60000 - 5s - loss: 0.1740 - accuracy: 0.9521 - val_loss: 0.2951 - val_accuracy: 0.9193
-Epoch 219/1000
-60000/60000 - 5s - loss: 0.1735 - accuracy: 0.9525 - val_loss: 0.2952 - val_accuracy: 0.9189
-Epoch 220/1000
-60000/60000 - 5s - loss: 0.1733 - accuracy: 0.9524 - val_loss: 0.2955 - val_accuracy: 0.9191
-Epoch 221/1000
-60000/60000 - 5s - loss: 0.1729 - accuracy: 0.9527 - val_loss: 0.2949 - val_accuracy: 0.9194
-Epoch 222/1000
-60000/60000 - 5s - loss: 0.1725 - accuracy: 0.9528 - val_loss: 0.2949 - val_accuracy: 0.9188
-Epoch 223/1000
-60000/60000 - 5s - loss: 0.1722 - accuracy: 0.9531 - val_loss: 0.2944 - val_accuracy: 0.9186
-Epoch 224/1000
-60000/60000 - 5s - loss: 0.1721 - accuracy: 0.9529 - val_loss: 0.2959 - val_accuracy: 0.9183
-Epoch 225/1000
-60000/60000 - 5s - loss: 0.1716 - accuracy: 0.9529 - val_loss: 0.2955 - val_accuracy: 0.9182
-Epoch 226/1000
-60000/60000 - 5s - loss: 0.1713 - accuracy: 0.9529 - val_loss: 0.2950 - val_accuracy: 0.9184
-Epoch 227/1000
-60000/60000 - 5s - loss: 0.1711 - accuracy: 0.9532 - val_loss: 0.2943 - val_accuracy: 0.9187
-Epoch 228/1000
-60000/60000 - 5s - loss: 0.1705 - accuracy: 0.9531 - val_loss: 0.2961 - val_accuracy: 0.9186
-Epoch 229/1000
-60000/60000 - 5s - loss: 0.1703 - accuracy: 0.9530 - val_loss: 0.2951 - val_accuracy: 0.9186
-Epoch 230/1000
-60000/60000 - 5s - loss: 0.1699 - accuracy: 0.9535 - val_loss: 0.2941 - val_accuracy: 0.9186
-Epoch 231/1000
-60000/60000 - 5s - loss: 0.1699 - accuracy: 0.9535 - val_loss: 0.2945 - val_accuracy: 0.9189
-Epoch 232/1000
-60000/60000 - 5s - loss: 0.1696 - accuracy: 0.9535 - val_loss: 0.2949 - val_accuracy: 0.9186
-Epoch 233/1000
-60000/60000 - 5s - loss: 0.1692 - accuracy: 0.9539 - val_loss: 0.2946 - val_accuracy: 0.9189
-Epoch 234/1000
-60000/60000 - 5s - loss: 0.1688 - accuracy: 0.9538 - val_loss: 0.2941 - val_accuracy: 0.9187
-Epoch 235/1000
-60000/60000 - 5s - loss: 0.1685 - accuracy: 0.9539 - val_loss: 0.2947 - val_accuracy: 0.9198
-Epoch 236/1000
-60000/60000 - 5s - loss: 0.1683 - accuracy: 0.9540 - val_loss: 0.2948 - val_accuracy: 0.9187
-Epoch 237/1000
-60000/60000 - 5s - loss: 0.1679 - accuracy: 0.9538 - val_loss: 0.2955 - val_accuracy: 0.9192
-Epoch 238/1000
-60000/60000 - 5s - loss: 0.1676 - accuracy: 0.9540 - val_loss: 0.2943 - val_accuracy: 0.9194
-Epoch 239/1000
-60000/60000 - 5s - loss: 0.1675 - accuracy: 0.9543 - val_loss: 0.2949 - val_accuracy: 0.9189
-Epoch 240/1000
-60000/60000 - 5s - loss: 0.1671 - accuracy: 0.9542 - val_loss: 0.2940 - val_accuracy: 0.9201
-Epoch 241/1000
-60000/60000 - 5s - loss: 0.1668 - accuracy: 0.9541 - val_loss: 0.2942 - val_accuracy: 0.9200
-Epoch 242/1000
-60000/60000 - 5s - loss: 0.1665 - accuracy: 0.9541 - val_loss: 0.2940 - val_accuracy: 0.9201
-Epoch 243/1000
-60000/60000 - 4s - loss: 0.1663 - accuracy: 0.9542 - val_loss: 0.2950 - val_accuracy: 0.9201
-Epoch 244/1000
-60000/60000 - 5s - loss: 0.1661 - accuracy: 0.9546 - val_loss: 0.2946 - val_accuracy: 0.9197
-Epoch 245/1000
-60000/60000 - 5s - loss: 0.1657 - accuracy: 0.9546 - val_loss: 0.2942 - val_accuracy: 0.9206
-Epoch 246/1000
-60000/60000 - 5s - loss: 0.1654 - accuracy: 0.9546 - val_loss: 0.2953 - val_accuracy: 0.9197
-Epoch 247/1000
-60000/60000 - 5s - loss: 0.1651 - accuracy: 0.9548 - val_loss: 0.2944 - val_accuracy: 0.9207
-Epoch 248/1000
-60000/60000 - 5s - loss: 0.1648 - accuracy: 0.9546 - val_loss: 0.2947 - val_accuracy: 0.9206
-Epoch 249/1000
-60000/60000 - 5s - loss: 0.1647 - accuracy: 0.9546 - val_loss: 0.2948 - val_accuracy: 0.9205
-Epoch 250/1000
-60000/60000 - 5s - loss: 0.1643 - accuracy: 0.9551 - val_loss: 0.2943 - val_accuracy: 0.9210
-Epoch 251/1000
-60000/60000 - 5s - loss: 0.1640 - accuracy: 0.9549 - val_loss: 0.2942 - val_accuracy: 0.9209
-Epoch 252/1000
-60000/60000 - 5s - loss: 0.1638 - accuracy: 0.9550 - val_loss: 0.2946 - val_accuracy: 0.9206
-Epoch 253/1000
-60000/60000 - 5s - loss: 0.1635 - accuracy: 0.9556 - val_loss: 0.2955 - val_accuracy: 0.9210
-Epoch 254/1000
-60000/60000 - 5s - loss: 0.1632 - accuracy: 0.9554 - val_loss: 0.2941 - val_accuracy: 0.9216
-Epoch 255/1000
-60000/60000 - 5s - loss: 0.1631 - accuracy: 0.9554 - val_loss: 0.2952 - val_accuracy: 0.9205
-Epoch 256/1000
-60000/60000 - 5s - loss: 0.1628 - accuracy: 0.9556 - val_loss: 0.2959 - val_accuracy: 0.9207
-Epoch 257/1000
-60000/60000 - 5s - loss: 0.1626 - accuracy: 0.9553 - val_loss: 0.2943 - val_accuracy: 0.9208
-Epoch 258/1000
-60000/60000 - 5s - loss: 0.1622 - accuracy: 0.9555 - val_loss: 0.2941 - val_accuracy: 0.9222
-Epoch 259/1000
-60000/60000 - 5s - loss: 0.1619 - accuracy: 0.9559 - val_loss: 0.2938 - val_accuracy: 0.9219
-Epoch 260/1000
-60000/60000 - 5s - loss: 0.1617 - accuracy: 0.9559 - val_loss: 0.2939 - val_accuracy: 0.9220
-Epoch 261/1000
-60000/60000 - 5s - loss: 0.1615 - accuracy: 0.9558 - val_loss: 0.2941 - val_accuracy: 0.9220
-Epoch 262/1000
-60000/60000 - 5s - loss: 0.1614 - accuracy: 0.9559 - val_loss: 0.2937 - val_accuracy: 0.9227
-Epoch 263/1000
-60000/60000 - 5s - loss: 0.1611 - accuracy: 0.9560 - val_loss: 0.2941 - val_accuracy: 0.9222
-Epoch 264/1000
-60000/60000 - 5s - loss: 0.1608 - accuracy: 0.9564 - val_loss: 0.2936 - val_accuracy: 0.9223
-Epoch 265/1000
-60000/60000 - 5s - loss: 0.1604 - accuracy: 0.9564 - val_loss: 0.2949 - val_accuracy: 0.9219
-Epoch 266/1000
-60000/60000 - 5s - loss: 0.1604 - accuracy: 0.9563 - val_loss: 0.2929 - val_accuracy: 0.9230
-Epoch 267/1000
-60000/60000 - 5s - loss: 0.1603 - accuracy: 0.9560 - val_loss: 0.2939 - val_accuracy: 0.9218
-Epoch 268/1000
-60000/60000 - 5s - loss: 0.1600 - accuracy: 0.9562 - val_loss: 0.2936 - val_accuracy: 0.9226
-Epoch 269/1000
-60000/60000 - 5s - loss: 0.1596 - accuracy: 0.9563 - val_loss: 0.2937 - val_accuracy: 0.9224
-Epoch 270/1000
-60000/60000 - 5s - loss: 0.1593 - accuracy: 0.9566 - val_loss: 0.2937 - val_accuracy: 0.9223
-Epoch 271/1000
-60000/60000 - 5s - loss: 0.1593 - accuracy: 0.9567 - val_loss: 0.2938 - val_accuracy: 0.9223
-Epoch 272/1000
-60000/60000 - 5s - loss: 0.1590 - accuracy: 0.9567 - val_loss: 0.2940 - val_accuracy: 0.9212
-Epoch 273/1000
-60000/60000 - 5s - loss: 0.1586 - accuracy: 0.9568 - val_loss: 0.2939 - val_accuracy: 0.9215
-Epoch 274/1000
-60000/60000 - 5s - loss: 0.1586 - accuracy: 0.9567 - val_loss: 0.2934 - val_accuracy: 0.9222
-Epoch 275/1000
-60000/60000 - 5s - loss: 0.1584 - accuracy: 0.9571 - val_loss: 0.2934 - val_accuracy: 0.9220
-Epoch 276/1000
-60000/60000 - 5s - loss: 0.1581 - accuracy: 0.9570 - val_loss: 0.2941 - val_accuracy: 0.9219
-Epoch 277/1000
-60000/60000 - 5s - loss: 0.1579 - accuracy: 0.9571 - val_loss: 0.2942 - val_accuracy: 0.9223
-Epoch 278/1000
-60000/60000 - 5s - loss: 0.1577 - accuracy: 0.9570 - val_loss: 0.2934 - val_accuracy: 0.9219
-Epoch 279/1000
-60000/60000 - 5s - loss: 0.1575 - accuracy: 0.9573 - val_loss: 0.2938 - val_accuracy: 0.9220
-Epoch 280/1000
-60000/60000 - 5s - loss: 0.1572 - accuracy: 0.9574 - val_loss: 0.2934 - val_accuracy: 0.9217
-Epoch 281/1000
-60000/60000 - 5s - loss: 0.1570 - accuracy: 0.9574 - val_loss: 0.2941 - val_accuracy: 0.9216
-Epoch 282/1000
-60000/60000 - 5s - loss: 0.1568 - accuracy: 0.9574 - val_loss: 0.2937 - val_accuracy: 0.9222
-Epoch 283/1000
-60000/60000 - 5s - loss: 0.1566 - accuracy: 0.9574 - val_loss: 0.2939 - val_accuracy: 0.9218
-Epoch 284/1000
-60000/60000 - 5s - loss: 0.1565 - accuracy: 0.9574 - val_loss: 0.2947 - val_accuracy: 0.9218
-Epoch 285/1000
-60000/60000 - 5s - loss: 0.1563 - accuracy: 0.9579 - val_loss: 0.2943 - val_accuracy: 0.9216
-Epoch 286/1000
-60000/60000 - 5s - loss: 0.1562 - accuracy: 0.9578 - val_loss: 0.2949 - val_accuracy: 0.9218
-Epoch 287/1000
-60000/60000 - 5s - loss: 0.1559 - accuracy: 0.9577 - val_loss: 0.2943 - val_accuracy: 0.9216
-Epoch 288/1000
-60000/60000 - 5s - loss: 0.1556 - accuracy: 0.9578 - val_loss: 0.2937 - val_accuracy: 0.9217
-Epoch 289/1000
-60000/60000 - 5s - loss: 0.1556 - accuracy: 0.9580 - val_loss: 0.2951 - val_accuracy: 0.9219
-Epoch 290/1000
-60000/60000 - 5s - loss: 0.1553 - accuracy: 0.9578 - val_loss: 0.2943 - val_accuracy: 0.9224
-Epoch 291/1000
-60000/60000 - 5s - loss: 0.1551 - accuracy: 0.9580 - val_loss: 0.2949 - val_accuracy: 0.9217
-Epoch 292/1000
-60000/60000 - 5s - loss: 0.1549 - accuracy: 0.9582 - val_loss: 0.2948 - val_accuracy: 0.9219
-Epoch 293/1000
-60000/60000 - 5s - loss: 0.1548 - accuracy: 0.9579 - val_loss: 0.2944 - val_accuracy: 0.9216
-Epoch 294/1000
-60000/60000 - 5s - loss: 0.1545 - accuracy: 0.9578 - val_loss: 0.2946 - val_accuracy: 0.9220
-Epoch 295/1000
-60000/60000 - 5s - loss: 0.1543 - accuracy: 0.9583 - val_loss: 0.2949 - val_accuracy: 0.9217
-Epoch 296/1000
-60000/60000 - 5s - loss: 0.1542 - accuracy: 0.9581 - val_loss: 0.2946 - val_accuracy: 0.9216
-Epoch 297/1000
-60000/60000 - 5s - loss: 0.1539 - accuracy: 0.9584 - val_loss: 0.2951 - val_accuracy: 0.9223
-Epoch 298/1000
-60000/60000 - 5s - loss: 0.1538 - accuracy: 0.9584 - val_loss: 0.2953 - val_accuracy: 0.9220
-Epoch 299/1000
-60000/60000 - 5s - loss: 0.1536 - accuracy: 0.9584 - val_loss: 0.2954 - val_accuracy: 0.9220
-Epoch 300/1000
-60000/60000 - 5s - loss: 0.1532 - accuracy: 0.9585 - val_loss: 0.2958 - val_accuracy: 0.9223
-Epoch 301/1000
-60000/60000 - 5s - loss: 0.1532 - accuracy: 0.9581 - val_loss: 0.2952 - val_accuracy: 0.9219
-Epoch 302/1000
-60000/60000 - 5s - loss: 0.1528 - accuracy: 0.9586 - val_loss: 0.2961 - val_accuracy: 0.9218
-Epoch 303/1000
-60000/60000 - 5s - loss: 0.1529 - accuracy: 0.9586 - val_loss: 0.2960 - val_accuracy: 0.9220
-Epoch 304/1000
-60000/60000 - 5s - loss: 0.1526 - accuracy: 0.9588 - val_loss: 0.2974 - val_accuracy: 0.9213
-Epoch 305/1000
-60000/60000 - 5s - loss: 0.1524 - accuracy: 0.9586 - val_loss: 0.2964 - val_accuracy: 0.9222
-Epoch 306/1000
-60000/60000 - 5s - loss: 0.1523 - accuracy: 0.9587 - val_loss: 0.2965 - val_accuracy: 0.9223
-Epoch 307/1000
-60000/60000 - 5s - loss: 0.1521 - accuracy: 0.9589 - val_loss: 0.2964 - val_accuracy: 0.9215
-Epoch 308/1000
-60000/60000 - 5s - loss: 0.1517 - accuracy: 0.9590 - val_loss: 0.2963 - val_accuracy: 0.9215
-Epoch 309/1000
-60000/60000 - 5s - loss: 0.1517 - accuracy: 0.9589 - val_loss: 0.2963 - val_accuracy: 0.9222
-Epoch 310/1000
-60000/60000 - 5s - loss: 0.1516 - accuracy: 0.9590 - val_loss: 0.2964 - val_accuracy: 0.9222
-Epoch 311/1000
-60000/60000 - 5s - loss: 0.1514 - accuracy: 0.9588 - val_loss: 0.2964 - val_accuracy: 0.9220
-Epoch 312/1000
-60000/60000 - 5s - loss: 0.1512 - accuracy: 0.9590 - val_loss: 0.2978 - val_accuracy: 0.9224
-Epoch 313/1000
-60000/60000 - 5s - loss: 0.1509 - accuracy: 0.9593 - val_loss: 0.2968 - val_accuracy: 0.9226
-Epoch 314/1000
-60000/60000 - 5s - loss: 0.1506 - accuracy: 0.9594 - val_loss: 0.2966 - val_accuracy: 0.9227
-Epoch 315/1000
-60000/60000 - 5s - loss: 0.1507 - accuracy: 0.9592 - val_loss: 0.2974 - val_accuracy: 0.9225
-Epoch 316/1000
-60000/60000 - 5s - loss: 0.1505 - accuracy: 0.9590 - val_loss: 0.2969 - val_accuracy: 0.9226
-Epoch 317/1000
-60000/60000 - 5s - loss: 0.1503 - accuracy: 0.9592 - val_loss: 0.2964 - val_accuracy: 0.9230
-Epoch 318/1000
-60000/60000 - 5s - loss: 0.1500 - accuracy: 0.9594 - val_loss: 0.2974 - val_accuracy: 0.9235
-Epoch 319/1000
-60000/60000 - 5s - loss: 0.1498 - accuracy: 0.9593 - val_loss: 0.2968 - val_accuracy: 0.9228
-Epoch 320/1000
-60000/60000 - 5s - loss: 0.1496 - accuracy: 0.9593 - val_loss: 0.2967 - val_accuracy: 0.9226
-Epoch 321/1000
-60000/60000 - 5s - loss: 0.1496 - accuracy: 0.9594 - val_loss: 0.2973 - val_accuracy: 0.9233
-Epoch 322/1000
-60000/60000 - 5s - loss: 0.1494 - accuracy: 0.9595 - val_loss: 0.2971 - val_accuracy: 0.9232
-Epoch 323/1000
-60000/60000 - 5s - loss: 0.1492 - accuracy: 0.9597 - val_loss: 0.2968 - val_accuracy: 0.9232
-Epoch 324/1000
-60000/60000 - 5s - loss: 0.1492 - accuracy: 0.9595 - val_loss: 0.2978 - val_accuracy: 0.9227
-Epoch 325/1000
-60000/60000 - 5s - loss: 0.1489 - accuracy: 0.9596 - val_loss: 0.2982 - val_accuracy: 0.9231
-Epoch 326/1000
-60000/60000 - 5s - loss: 0.1487 - accuracy: 0.9598 - val_loss: 0.2977 - val_accuracy: 0.9229
-Epoch 327/1000
-60000/60000 - 5s - loss: 0.1484 - accuracy: 0.9597 - val_loss: 0.2977 - val_accuracy: 0.9231
-Epoch 328/1000
-60000/60000 - 5s - loss: 0.1482 - accuracy: 0.9599 - val_loss: 0.2972 - val_accuracy: 0.9239
-Epoch 329/1000
-60000/60000 - 5s - loss: 0.1481 - accuracy: 0.9598 - val_loss: 0.2977 - val_accuracy: 0.9231
-Epoch 330/1000
-60000/60000 - 5s - loss: 0.1480 - accuracy: 0.9601 - val_loss: 0.2974 - val_accuracy: 0.9233
-Epoch 331/1000
-60000/60000 - 5s - loss: 0.1477 - accuracy: 0.9601 - val_loss: 0.2982 - val_accuracy: 0.9221
-Epoch 332/1000
-60000/60000 - 5s - loss: 0.1475 - accuracy: 0.9603 - val_loss: 0.2984 - val_accuracy: 0.9228
-Epoch 333/1000
-60000/60000 - 5s - loss: 0.1474 - accuracy: 0.9598 - val_loss: 0.2987 - val_accuracy: 0.9227
-Epoch 334/1000
-60000/60000 - 5s - loss: 0.1472 - accuracy: 0.9600 - val_loss: 0.2978 - val_accuracy: 0.9221
-Epoch 335/1000
-60000/60000 - 5s - loss: 0.1471 - accuracy: 0.9602 - val_loss: 0.2973 - val_accuracy: 0.9232
-Epoch 336/1000
-60000/60000 - 5s - loss: 0.1469 - accuracy: 0.9605 - val_loss: 0.2989 - val_accuracy: 0.9229
-Epoch 337/1000
-60000/60000 - 5s - loss: 0.1468 - accuracy: 0.9602 - val_loss: 0.2985 - val_accuracy: 0.9233
-Epoch 338/1000
-60000/60000 - 5s - loss: 0.1466 - accuracy: 0.9605 - val_loss: 0.2984 - val_accuracy: 0.9237
-Epoch 339/1000
-60000/60000 - 5s - loss: 0.1464 - accuracy: 0.9602 - val_loss: 0.2980 - val_accuracy: 0.9230
-Epoch 340/1000
-60000/60000 - 5s - loss: 0.1462 - accuracy: 0.9605 - val_loss: 0.2984 - val_accuracy: 0.9223
-Epoch 341/1000
-60000/60000 - 5s - loss: 0.1461 - accuracy: 0.9604 - val_loss: 0.2987 - val_accuracy: 0.9220
-Epoch 342/1000
-60000/60000 - 5s - loss: 0.1459 - accuracy: 0.9602 - val_loss: 0.2990 - val_accuracy: 0.9227
-Epoch 343/1000
-60000/60000 - 5s - loss: 0.1458 - accuracy: 0.9607 - val_loss: 0.2983 - val_accuracy: 0.9226
-Epoch 344/1000
-60000/60000 - 5s - loss: 0.1457 - accuracy: 0.9604 - val_loss: 0.2991 - val_accuracy: 0.9227
-Epoch 345/1000
-60000/60000 - 5s - loss: 0.1454 - accuracy: 0.9607 - val_loss: 0.3001 - val_accuracy: 0.9222
-Epoch 346/1000
-60000/60000 - 5s - loss: 0.1454 - accuracy: 0.9607 - val_loss: 0.2988 - val_accuracy: 0.9221
-Epoch 347/1000
-60000/60000 - 5s - loss: 0.1451 - accuracy: 0.9607 - val_loss: 0.2994 - val_accuracy: 0.9225
-Epoch 348/1000
-60000/60000 - 5s - loss: 0.1450 - accuracy: 0.9607 - val_loss: 0.2989 - val_accuracy: 0.9227
-Epoch 349/1000
-60000/60000 - 5s - loss: 0.1449 - accuracy: 0.9607 - val_loss: 0.2994 - val_accuracy: 0.9233
-Epoch 350/1000
-60000/60000 - 5s - loss: 0.1447 - accuracy: 0.9609 - val_loss: 0.2994 - val_accuracy: 0.9227
-Epoch 351/1000
-60000/60000 - 5s - loss: 0.1443 - accuracy: 0.9613 - val_loss: 0.2992 - val_accuracy: 0.9231
-Epoch 352/1000
-60000/60000 - 5s - loss: 0.1444 - accuracy: 0.9614 - val_loss: 0.2997 - val_accuracy: 0.9226
-Epoch 353/1000
-60000/60000 - 5s - loss: 0.1443 - accuracy: 0.9615 - val_loss: 0.2995 - val_accuracy: 0.9219
-Epoch 354/1000
-60000/60000 - 5s - loss: 0.1441 - accuracy: 0.9615 - val_loss: 0.2992 - val_accuracy: 0.9233
-Epoch 355/1000
-60000/60000 - 5s - loss: 0.1440 - accuracy: 0.9614 - val_loss: 0.2999 - val_accuracy: 0.9235
-Epoch 356/1000
-60000/60000 - 5s - loss: 0.1438 - accuracy: 0.9611 - val_loss: 0.2993 - val_accuracy: 0.9228
-Epoch 357/1000
-60000/60000 - 5s - loss: 0.1437 - accuracy: 0.9613 - val_loss: 0.3001 - val_accuracy: 0.9225
-Epoch 358/1000
-60000/60000 - 5s - loss: 0.1434 - accuracy: 0.9616 - val_loss: 0.2997 - val_accuracy: 0.9230
-Epoch 359/1000
-60000/60000 - 5s - loss: 0.1434 - accuracy: 0.9617 - val_loss: 0.2999 - val_accuracy: 0.9231
-Epoch 360/1000
-60000/60000 - 5s - loss: 0.1432 - accuracy: 0.9615 - val_loss: 0.2994 - val_accuracy: 0.9226
-Epoch 361/1000
-60000/60000 - 5s - loss: 0.1431 - accuracy: 0.9614 - val_loss: 0.3002 - val_accuracy: 0.9230
-Epoch 362/1000
-60000/60000 - 5s - loss: 0.1430 - accuracy: 0.9615 - val_loss: 0.2999 - val_accuracy: 0.9232
-Epoch 363/1000
-60000/60000 - 5s - loss: 0.1429 - accuracy: 0.9615 - val_loss: 0.3009 - val_accuracy: 0.9233
-Epoch 364/1000
-60000/60000 - 5s - loss: 0.1427 - accuracy: 0.9619 - val_loss: 0.3003 - val_accuracy: 0.9231
-Epoch 365/1000
-60000/60000 - 5s - loss: 0.1426 - accuracy: 0.9620 - val_loss: 0.2998 - val_accuracy: 0.9228
-Epoch 366/1000
-60000/60000 - 5s - loss: 0.1424 - accuracy: 0.9620 - val_loss: 0.3003 - val_accuracy: 0.9233
-Epoch 367/1000
-60000/60000 - 5s - loss: 0.1423 - accuracy: 0.9623 - val_loss: 0.3010 - val_accuracy: 0.9230
-Epoch 368/1000
-60000/60000 - 5s - loss: 0.1422 - accuracy: 0.9622 - val_loss: 0.3015 - val_accuracy: 0.9225
-Epoch 369/1000
-60000/60000 - 5s - loss: 0.1420 - accuracy: 0.9623 - val_loss: 0.3007 - val_accuracy: 0.9232
-Epoch 370/1000
-60000/60000 - 5s - loss: 0.1417 - accuracy: 0.9622 - val_loss: 0.3005 - val_accuracy: 0.9231
-Epoch 371/1000
-60000/60000 - 5s - loss: 0.1418 - accuracy: 0.9620 - val_loss: 0.3009 - val_accuracy: 0.9232
-Epoch 372/1000
-60000/60000 - 5s - loss: 0.1416 - accuracy: 0.9627 - val_loss: 0.3011 - val_accuracy: 0.9229
-Epoch 373/1000
-60000/60000 - 5s - loss: 0.1414 - accuracy: 0.9626 - val_loss: 0.3017 - val_accuracy: 0.9228
-Epoch 374/1000
-60000/60000 - 5s - loss: 0.1413 - accuracy: 0.9624 - val_loss: 0.3015 - val_accuracy: 0.9226
-Epoch 375/1000
-60000/60000 - 5s - loss: 0.1410 - accuracy: 0.9625 - val_loss: 0.3019 - val_accuracy: 0.9227
-Epoch 376/1000
-60000/60000 - 5s - loss: 0.1410 - accuracy: 0.9623 - val_loss: 0.3015 - val_accuracy: 0.9231
-Epoch 377/1000
-60000/60000 - 5s - loss: 0.1409 - accuracy: 0.9624 - val_loss: 0.3020 - val_accuracy: 0.9224
-Epoch 378/1000
-60000/60000 - 5s - loss: 0.1407 - accuracy: 0.9626 - val_loss: 0.3022 - val_accuracy: 0.9230
-Epoch 379/1000
-60000/60000 - 5s - loss: 0.1406 - accuracy: 0.9624 - val_loss: 0.3019 - val_accuracy: 0.9226
-Epoch 380/1000
-60000/60000 - 5s - loss: 0.1403 - accuracy: 0.9627 - val_loss: 0.3017 - val_accuracy: 0.9230
-Epoch 381/1000
-60000/60000 - 5s - loss: 0.1403 - accuracy: 0.9628 - val_loss: 0.3025 - val_accuracy: 0.9214
-Epoch 382/1000
-60000/60000 - 5s - loss: 0.1402 - accuracy: 0.9632 - val_loss: 0.3026 - val_accuracy: 0.9224
-Epoch 383/1000
-60000/60000 - 5s - loss: 0.1401 - accuracy: 0.9628 - val_loss: 0.3017 - val_accuracy: 0.9228
-Epoch 384/1000
-60000/60000 - 5s - loss: 0.1398 - accuracy: 0.9631 - val_loss: 0.3030 - val_accuracy: 0.9229
-Epoch 385/1000
-60000/60000 - 5s - loss: 0.1399 - accuracy: 0.9629 - val_loss: 0.3031 - val_accuracy: 0.9220
-Epoch 386/1000
-60000/60000 - 5s - loss: 0.1396 - accuracy: 0.9630 - val_loss: 0.3030 - val_accuracy: 0.9224
-Epoch 387/1000
-60000/60000 - 5s - loss: 0.1396 - accuracy: 0.9631 - val_loss: 0.3021 - val_accuracy: 0.9230
-Epoch 388/1000
-60000/60000 - 5s - loss: 0.1394 - accuracy: 0.9633 - val_loss: 0.3025 - val_accuracy: 0.9227
-Epoch 389/1000
-60000/60000 - 5s - loss: 0.1393 - accuracy: 0.9631 - val_loss: 0.3028 - val_accuracy: 0.9225
-Epoch 390/1000
-60000/60000 - 5s - loss: 0.1392 - accuracy: 0.9631 - val_loss: 0.3030 - val_accuracy: 0.9225
-Epoch 391/1000
-60000/60000 - 5s - loss: 0.1391 - accuracy: 0.9631 - val_loss: 0.3037 - val_accuracy: 0.9221
-Epoch 392/1000
-60000/60000 - 5s - loss: 0.1389 - accuracy: 0.9631 - val_loss: 0.3034 - val_accuracy: 0.9221
-Epoch 393/1000
-60000/60000 - 5s - loss: 0.1387 - accuracy: 0.9629 - val_loss: 0.3026 - val_accuracy: 0.9223
-Epoch 394/1000
-60000/60000 - 5s - loss: 0.1387 - accuracy: 0.9630 - val_loss: 0.3037 - val_accuracy: 0.9226
-Epoch 395/1000
-60000/60000 - 5s - loss: 0.1386 - accuracy: 0.9633 - val_loss: 0.3040 - val_accuracy: 0.9225
-Epoch 396/1000
-60000/60000 - 5s - loss: 0.1384 - accuracy: 0.9635 - val_loss: 0.3038 - val_accuracy: 0.9227
-Epoch 397/1000
-60000/60000 - 5s - loss: 0.1383 - accuracy: 0.9636 - val_loss: 0.3035 - val_accuracy: 0.9225
-Epoch 398/1000
-60000/60000 - 5s - loss: 0.1381 - accuracy: 0.9634 - val_loss: 0.3035 - val_accuracy: 0.9231
-Epoch 399/1000
-60000/60000 - 5s - loss: 0.1381 - accuracy: 0.9636 - val_loss: 0.3040 - val_accuracy: 0.9229
-Epoch 400/1000
-60000/60000 - 5s - loss: 0.1379 - accuracy: 0.9636 - val_loss: 0.3046 - val_accuracy: 0.9226
-Epoch 401/1000
-60000/60000 - 5s - loss: 0.1378 - accuracy: 0.9635 - val_loss: 0.3043 - val_accuracy: 0.9229
-Epoch 402/1000
-60000/60000 - 5s - loss: 0.1376 - accuracy: 0.9639 - val_loss: 0.3042 - val_accuracy: 0.9229
-Epoch 403/1000
-60000/60000 - 5s - loss: 0.1376 - accuracy: 0.9636 - val_loss: 0.3044 - val_accuracy: 0.9226
-Epoch 404/1000
-60000/60000 - 5s - loss: 0.1375 - accuracy: 0.9638 - val_loss: 0.3044 - val_accuracy: 0.9234
-Epoch 405/1000
-60000/60000 - 5s - loss: 0.1374 - accuracy: 0.9637 - val_loss: 0.3046 - val_accuracy: 0.9223
-Epoch 406/1000
-60000/60000 - 5s - loss: 0.1373 - accuracy: 0.9639 - val_loss: 0.3046 - val_accuracy: 0.9232
-Epoch 407/1000
-60000/60000 - 5s - loss: 0.1371 - accuracy: 0.9640 - val_loss: 0.3051 - val_accuracy: 0.9228
-Epoch 408/1000
-60000/60000 - 5s - loss: 0.1369 - accuracy: 0.9641 - val_loss: 0.3039 - val_accuracy: 0.9233
-Epoch 409/1000
-60000/60000 - 5s - loss: 0.1368 - accuracy: 0.9640 - val_loss: 0.3044 - val_accuracy: 0.9228
-Epoch 410/1000
-60000/60000 - 5s - loss: 0.1368 - accuracy: 0.9642 - val_loss: 0.3047 - val_accuracy: 0.9230
-Epoch 411/1000
-60000/60000 - 5s - loss: 0.1367 - accuracy: 0.9641 - val_loss: 0.3040 - val_accuracy: 0.9234
-Epoch 412/1000
-60000/60000 - 5s - loss: 0.1364 - accuracy: 0.9640 - val_loss: 0.3054 - val_accuracy: 0.9224
-Epoch 413/1000
-60000/60000 - 5s - loss: 0.1364 - accuracy: 0.9642 - val_loss: 0.3047 - val_accuracy: 0.9235
-Epoch 414/1000
-60000/60000 - 5s - loss: 0.1362 - accuracy: 0.9643 - val_loss: 0.3048 - val_accuracy: 0.9212
-Epoch 415/1000
-60000/60000 - 5s - loss: 0.1362 - accuracy: 0.9645 - val_loss: 0.3051 - val_accuracy: 0.9222
-Epoch 416/1000
-60000/60000 - 5s - loss: 0.1361 - accuracy: 0.9642 - val_loss: 0.3052 - val_accuracy: 0.9229
-Epoch 417/1000
-60000/60000 - 5s - loss: 0.1360 - accuracy: 0.9647 - val_loss: 0.3054 - val_accuracy: 0.9229
-Epoch 418/1000
-60000/60000 - 5s - loss: 0.1358 - accuracy: 0.9645 - val_loss: 0.3050 - val_accuracy: 0.9234
-Epoch 419/1000
-60000/60000 - 5s - loss: 0.1357 - accuracy: 0.9646 - val_loss: 0.3059 - val_accuracy: 0.9224
-Epoch 420/1000
-60000/60000 - 5s - loss: 0.1356 - accuracy: 0.9647 - val_loss: 0.3060 - val_accuracy: 0.9228
-Epoch 421/1000
-60000/60000 - 5s - loss: 0.1355 - accuracy: 0.9646 - val_loss: 0.3054 - val_accuracy: 0.9231
-Epoch 422/1000
-60000/60000 - 5s - loss: 0.1352 - accuracy: 0.9645 - val_loss: 0.3059 - val_accuracy: 0.9221
-Epoch 423/1000
-60000/60000 - 5s - loss: 0.1352 - accuracy: 0.9649 - val_loss: 0.3055 - val_accuracy: 0.9227
-Epoch 424/1000
-60000/60000 - 5s - loss: 0.1351 - accuracy: 0.9648 - val_loss: 0.3061 - val_accuracy: 0.9225
-Epoch 425/1000
-60000/60000 - 5s - loss: 0.1350 - accuracy: 0.9645 - val_loss: 0.3057 - val_accuracy: 0.9231
-Epoch 426/1000
-60000/60000 - 5s - loss: 0.1350 - accuracy: 0.9646 - val_loss: 0.3058 - val_accuracy: 0.9235
-Epoch 427/1000
-60000/60000 - 5s - loss: 0.1348 - accuracy: 0.9651 - val_loss: 0.3060 - val_accuracy: 0.9230
-Epoch 428/1000
-60000/60000 - 5s - loss: 0.1347 - accuracy: 0.9649 - val_loss: 0.3065 - val_accuracy: 0.9214
-Epoch 429/1000
-60000/60000 - 5s - loss: 0.1346 - accuracy: 0.9649 - val_loss: 0.3066 - val_accuracy: 0.9228
-Epoch 430/1000
-60000/60000 - 5s - loss: 0.1345 - accuracy: 0.9650 - val_loss: 0.3061 - val_accuracy: 0.9232
-Epoch 431/1000
-60000/60000 - 5s - loss: 0.1343 - accuracy: 0.9650 - val_loss: 0.3060 - val_accuracy: 0.9223
-Epoch 432/1000
-60000/60000 - 5s - loss: 0.1342 - accuracy: 0.9650 - val_loss: 0.3064 - val_accuracy: 0.9235
-Epoch 433/1000
-60000/60000 - 5s - loss: 0.1342 - accuracy: 0.9653 - val_loss: 0.3066 - val_accuracy: 0.9214
-Epoch 434/1000
-60000/60000 - 5s - loss: 0.1340 - accuracy: 0.9653 - val_loss: 0.3068 - val_accuracy: 0.9223
-Epoch 435/1000
-60000/60000 - 5s - loss: 0.1339 - accuracy: 0.9653 - val_loss: 0.3063 - val_accuracy: 0.9235
-Epoch 436/1000
-60000/60000 - 5s - loss: 0.1338 - accuracy: 0.9653 - val_loss: 0.3072 - val_accuracy: 0.9222
-Epoch 437/1000
-60000/60000 - 5s - loss: 0.1338 - accuracy: 0.9654 - val_loss: 0.3062 - val_accuracy: 0.9220
-Epoch 438/1000
-60000/60000 - 5s - loss: 0.1336 - accuracy: 0.9656 - val_loss: 0.3069 - val_accuracy: 0.9213
-Epoch 439/1000
-60000/60000 - 5s - loss: 0.1336 - accuracy: 0.9654 - val_loss: 0.3069 - val_accuracy: 0.9228
-Epoch 440/1000
-60000/60000 - 5s - loss: 0.1335 - accuracy: 0.9654 - val_loss: 0.3072 - val_accuracy: 0.9228
-Epoch 441/1000
-60000/60000 - 5s - loss: 0.1334 - accuracy: 0.9655 - val_loss: 0.3075 - val_accuracy: 0.9224
-Epoch 442/1000
-60000/60000 - 5s - loss: 0.1333 - accuracy: 0.9654 - val_loss: 0.3077 - val_accuracy: 0.9226
-Epoch 443/1000
-60000/60000 - 5s - loss: 0.1332 - accuracy: 0.9656 - val_loss: 0.3071 - val_accuracy: 0.9227
-Epoch 444/1000
-60000/60000 - 5s - loss: 0.1329 - accuracy: 0.9655 - val_loss: 0.3084 - val_accuracy: 0.9225
-Epoch 445/1000
-60000/60000 - 5s - loss: 0.1328 - accuracy: 0.9658 - val_loss: 0.3075 - val_accuracy: 0.9229
-Epoch 446/1000
-60000/60000 - 5s - loss: 0.1329 - accuracy: 0.9658 - val_loss: 0.3077 - val_accuracy: 0.9228
-Epoch 447/1000
-60000/60000 - 5s - loss: 0.1327 - accuracy: 0.9655 - val_loss: 0.3075 - val_accuracy: 0.9229
-Epoch 448/1000
-60000/60000 - 5s - loss: 0.1326 - accuracy: 0.9656 - val_loss: 0.3078 - val_accuracy: 0.9232
-Epoch 449/1000
-60000/60000 - 5s - loss: 0.1326 - accuracy: 0.9659 - val_loss: 0.3087 - val_accuracy: 0.9225
-Epoch 450/1000
-60000/60000 - 5s - loss: 0.1323 - accuracy: 0.9661 - val_loss: 0.3084 - val_accuracy: 0.9229
-Epoch 451/1000
-60000/60000 - 5s - loss: 0.1323 - accuracy: 0.9658 - val_loss: 0.3080 - val_accuracy: 0.9214
-Epoch 452/1000
-60000/60000 - 5s - loss: 0.1321 - accuracy: 0.9660 - val_loss: 0.3090 - val_accuracy: 0.9228
-Epoch 453/1000
-60000/60000 - 5s - loss: 0.1322 - accuracy: 0.9659 - val_loss: 0.3084 - val_accuracy: 0.9228
-Epoch 454/1000
-60000/60000 - 5s - loss: 0.1320 - accuracy: 0.9660 - val_loss: 0.3086 - val_accuracy: 0.9229
-Epoch 455/1000
-60000/60000 - 5s - loss: 0.1319 - accuracy: 0.9664 - val_loss: 0.3086 - val_accuracy: 0.9223
-Epoch 456/1000
-60000/60000 - 5s - loss: 0.1318 - accuracy: 0.9663 - val_loss: 0.3094 - val_accuracy: 0.9228
-Epoch 457/1000
-60000/60000 - 5s - loss: 0.1316 - accuracy: 0.9659 - val_loss: 0.3083 - val_accuracy: 0.9237
-Epoch 458/1000
-60000/60000 - 5s - loss: 0.1315 - accuracy: 0.9661 - val_loss: 0.3083 - val_accuracy: 0.9230
-Epoch 459/1000
-60000/60000 - 5s - loss: 0.1314 - accuracy: 0.9664 - val_loss: 0.3088 - val_accuracy: 0.9224
-Epoch 460/1000
-60000/60000 - 5s - loss: 0.1314 - accuracy: 0.9663 - val_loss: 0.3082 - val_accuracy: 0.9233
-Epoch 461/1000
-60000/60000 - 5s - loss: 0.1313 - accuracy: 0.9664 - val_loss: 0.3092 - val_accuracy: 0.9228
-Epoch 462/1000
-60000/60000 - 5s - loss: 0.1311 - accuracy: 0.9664 - val_loss: 0.3088 - val_accuracy: 0.9230
-Epoch 463/1000
-60000/60000 - 5s - loss: 0.1310 - accuracy: 0.9663 - val_loss: 0.3095 - val_accuracy: 0.9230
-Epoch 464/1000
-60000/60000 - 5s - loss: 0.1309 - accuracy: 0.9667 - val_loss: 0.3099 - val_accuracy: 0.9225
-Epoch 465/1000
-60000/60000 - 5s - loss: 0.1308 - accuracy: 0.9663 - val_loss: 0.3093 - val_accuracy: 0.9232
-Epoch 466/1000
-60000/60000 - 5s - loss: 0.1306 - accuracy: 0.9660 - val_loss: 0.3096 - val_accuracy: 0.9228
-Epoch 467/1000
-60000/60000 - 5s - loss: 0.1305 - accuracy: 0.9666 - val_loss: 0.3091 - val_accuracy: 0.9221
-Epoch 468/1000
-60000/60000 - 5s - loss: 0.1304 - accuracy: 0.9663 - val_loss: 0.3101 - val_accuracy: 0.9224
-Epoch 469/1000
-60000/60000 - 5s - loss: 0.1303 - accuracy: 0.9665 - val_loss: 0.3097 - val_accuracy: 0.9223
-Epoch 470/1000
-60000/60000 - 5s - loss: 0.1302 - accuracy: 0.9667 - val_loss: 0.3111 - val_accuracy: 0.9223
-Epoch 471/1000
-60000/60000 - 5s - loss: 0.1302 - accuracy: 0.9666 - val_loss: 0.3106 - val_accuracy: 0.9230
-Epoch 472/1000
-60000/60000 - 5s - loss: 0.1303 - accuracy: 0.9665 - val_loss: 0.3101 - val_accuracy: 0.9227
-Epoch 473/1000
-60000/60000 - 5s - loss: 0.1300 - accuracy: 0.9671 - val_loss: 0.3102 - val_accuracy: 0.9229
-Epoch 474/1000
-60000/60000 - 5s - loss: 0.1300 - accuracy: 0.9665 - val_loss: 0.3098 - val_accuracy: 0.9229
-Epoch 475/1000
-60000/60000 - 5s - loss: 0.1297 - accuracy: 0.9668 - val_loss: 0.3109 - val_accuracy: 0.9228
-Epoch 476/1000
-60000/60000 - 5s - loss: 0.1297 - accuracy: 0.9669 - val_loss: 0.3105 - val_accuracy: 0.9232
-Epoch 477/1000
-60000/60000 - 5s - loss: 0.1296 - accuracy: 0.9667 - val_loss: 0.3106 - val_accuracy: 0.9223
-Epoch 478/1000
-60000/60000 - 5s - loss: 0.1295 - accuracy: 0.9669 - val_loss: 0.3107 - val_accuracy: 0.9226
-Epoch 479/1000
-60000/60000 - 5s - loss: 0.1293 - accuracy: 0.9669 - val_loss: 0.3109 - val_accuracy: 0.9212
-Epoch 480/1000
-60000/60000 - 5s - loss: 0.1294 - accuracy: 0.9667 - val_loss: 0.3116 - val_accuracy: 0.9226
-Epoch 481/1000
-60000/60000 - 5s - loss: 0.1292 - accuracy: 0.9671 - val_loss: 0.3113 - val_accuracy: 0.9228
-Epoch 482/1000
-60000/60000 - 5s - loss: 0.1292 - accuracy: 0.9668 - val_loss: 0.3114 - val_accuracy: 0.9219
-Epoch 483/1000
-60000/60000 - 5s - loss: 0.1291 - accuracy: 0.9670 - val_loss: 0.3116 - val_accuracy: 0.9229
-Epoch 484/1000
-60000/60000 - 5s - loss: 0.1292 - accuracy: 0.9670 - val_loss: 0.3111 - val_accuracy: 0.9229
-Epoch 485/1000
-60000/60000 - 5s - loss: 0.1289 - accuracy: 0.9671 - val_loss: 0.3118 - val_accuracy: 0.9225
-Epoch 486/1000
-60000/60000 - 5s - loss: 0.1289 - accuracy: 0.9671 - val_loss: 0.3114 - val_accuracy: 0.9227
-Epoch 487/1000
-60000/60000 - 5s - loss: 0.1287 - accuracy: 0.9670 - val_loss: 0.3118 - val_accuracy: 0.9221
-Epoch 488/1000
-60000/60000 - 5s - loss: 0.1286 - accuracy: 0.9675 - val_loss: 0.3122 - val_accuracy: 0.9225
-Epoch 489/1000
-60000/60000 - 5s - loss: 0.1286 - accuracy: 0.9671 - val_loss: 0.3121 - val_accuracy: 0.9224
-Epoch 490/1000
-60000/60000 - 5s - loss: 0.1284 - accuracy: 0.9674 - val_loss: 0.3117 - val_accuracy: 0.9224
-Epoch 491/1000
-60000/60000 - 5s - loss: 0.1284 - accuracy: 0.9671 - val_loss: 0.3125 - val_accuracy: 0.9227
-Epoch 492/1000
-60000/60000 - 5s - loss: 0.1283 - accuracy: 0.9674 - val_loss: 0.3126 - val_accuracy: 0.9228
-Epoch 493/1000
-60000/60000 - 5s - loss: 0.1283 - accuracy: 0.9672 - val_loss: 0.3123 - val_accuracy: 0.9227
-Epoch 494/1000
-60000/60000 - 5s - loss: 0.1283 - accuracy: 0.9672 - val_loss: 0.3127 - val_accuracy: 0.9223
-Epoch 495/1000
-60000/60000 - 5s - loss: 0.1280 - accuracy: 0.9672 - val_loss: 0.3132 - val_accuracy: 0.9225
-Epoch 496/1000
-60000/60000 - 5s - loss: 0.1281 - accuracy: 0.9673 - val_loss: 0.3134 - val_accuracy: 0.9224
-Epoch 497/1000
-60000/60000 - 5s - loss: 0.1279 - accuracy: 0.9672 - val_loss: 0.3129 - val_accuracy: 0.9227
-Epoch 498/1000
-60000/60000 - 5s - loss: 0.1277 - accuracy: 0.9672 - val_loss: 0.3140 - val_accuracy: 0.9225
-Epoch 499/1000
-60000/60000 - 5s - loss: 0.1277 - accuracy: 0.9672 - val_loss: 0.3132 - val_accuracy: 0.9232
-Epoch 500/1000
-60000/60000 - 5s - loss: 0.1276 - accuracy: 0.9674 - val_loss: 0.3135 - val_accuracy: 0.9220
-Epoch 501/1000
-60000/60000 - 5s - loss: 0.1276 - accuracy: 0.9675 - val_loss: 0.3136 - val_accuracy: 0.9232
-Epoch 502/1000
-60000/60000 - 5s - loss: 0.1275 - accuracy: 0.9675 - val_loss: 0.3141 - val_accuracy: 0.9226
-Epoch 503/1000
-60000/60000 - 5s - loss: 0.1273 - accuracy: 0.9675 - val_loss: 0.3137 - val_accuracy: 0.9223
-Epoch 504/1000
-60000/60000 - 5s - loss: 0.1274 - accuracy: 0.9675 - val_loss: 0.3136 - val_accuracy: 0.9230
-Epoch 505/1000
-60000/60000 - 5s - loss: 0.1272 - accuracy: 0.9675 - val_loss: 0.3141 - val_accuracy: 0.9230
-Epoch 506/1000
-60000/60000 - 5s - loss: 0.1272 - accuracy: 0.9674 - val_loss: 0.3141 - val_accuracy: 0.9220
-Epoch 507/1000
-60000/60000 - 5s - loss: 0.1271 - accuracy: 0.9678 - val_loss: 0.3143 - val_accuracy: 0.9230
-Epoch 508/1000
-60000/60000 - 5s - loss: 0.1269 - accuracy: 0.9677 - val_loss: 0.3144 - val_accuracy: 0.9229
-Epoch 509/1000
-60000/60000 - 5s - loss: 0.1269 - accuracy: 0.9679 - val_loss: 0.3146 - val_accuracy: 0.9229
-Epoch 510/1000
-60000/60000 - 5s - loss: 0.1268 - accuracy: 0.9677 - val_loss: 0.3146 - val_accuracy: 0.9218
-Epoch 511/1000
-60000/60000 - 5s - loss: 0.1266 - accuracy: 0.9675 - val_loss: 0.3147 - val_accuracy: 0.9224
-Epoch 512/1000
-60000/60000 - 5s - loss: 0.1267 - accuracy: 0.9675 - val_loss: 0.3154 - val_accuracy: 0.9227
-Epoch 513/1000
-60000/60000 - 5s - loss: 0.1265 - accuracy: 0.9677 - val_loss: 0.3154 - val_accuracy: 0.9221
-Epoch 514/1000
-60000/60000 - 5s - loss: 0.1264 - accuracy: 0.9677 - val_loss: 0.3152 - val_accuracy: 0.9224
-Epoch 515/1000
-60000/60000 - 5s - loss: 0.1263 - accuracy: 0.9675 - val_loss: 0.3154 - val_accuracy: 0.9225
-Epoch 516/1000
-60000/60000 - 5s - loss: 0.1263 - accuracy: 0.9676 - val_loss: 0.3156 - val_accuracy: 0.9225
-Epoch 517/1000
-60000/60000 - 5s - loss: 0.1263 - accuracy: 0.9679 - val_loss: 0.3156 - val_accuracy: 0.9234
-Epoch 518/1000
-60000/60000 - 5s - loss: 0.1261 - accuracy: 0.9677 - val_loss: 0.3167 - val_accuracy: 0.9235
-Epoch 519/1000
-60000/60000 - 5s - loss: 0.1261 - accuracy: 0.9679 - val_loss: 0.3160 - val_accuracy: 0.9213
-Epoch 520/1000
-60000/60000 - 5s - loss: 0.1260 - accuracy: 0.9680 - val_loss: 0.3163 - val_accuracy: 0.9228
-Epoch 521/1000
-60000/60000 - 5s - loss: 0.1258 - accuracy: 0.9679 - val_loss: 0.3162 - val_accuracy: 0.9223
-Epoch 522/1000
-60000/60000 - 5s - loss: 0.1258 - accuracy: 0.9679 - val_loss: 0.3164 - val_accuracy: 0.9213
-Epoch 523/1000
-60000/60000 - 5s - loss: 0.1257 - accuracy: 0.9678 - val_loss: 0.3169 - val_accuracy: 0.9233
-Epoch 524/1000
-60000/60000 - 5s - loss: 0.1256 - accuracy: 0.9680 - val_loss: 0.3167 - val_accuracy: 0.9222
-Epoch 525/1000
-60000/60000 - 5s - loss: 0.1256 - accuracy: 0.9677 - val_loss: 0.3170 - val_accuracy: 0.9229
-Epoch 526/1000
-60000/60000 - 5s - loss: 0.1255 - accuracy: 0.9678 - val_loss: 0.3167 - val_accuracy: 0.9219
-Epoch 527/1000
-60000/60000 - 5s - loss: 0.1255 - accuracy: 0.9681 - val_loss: 0.3169 - val_accuracy: 0.9223
-Epoch 528/1000
-60000/60000 - 5s - loss: 0.1253 - accuracy: 0.9681 - val_loss: 0.3174 - val_accuracy: 0.9229
-Epoch 529/1000
-60000/60000 - 5s - loss: 0.1253 - accuracy: 0.9681 - val_loss: 0.3171 - val_accuracy: 0.9216
-Epoch 530/1000
-60000/60000 - 5s - loss: 0.1252 - accuracy: 0.9679 - val_loss: 0.3173 - val_accuracy: 0.9218
-Epoch 531/1000
-60000/60000 - 5s - loss: 0.1252 - accuracy: 0.9682 - val_loss: 0.3168 - val_accuracy: 0.9221
-Epoch 532/1000
-60000/60000 - 5s - loss: 0.1250 - accuracy: 0.9682 - val_loss: 0.3173 - val_accuracy: 0.9228
-Epoch 533/1000
-60000/60000 - 5s - loss: 0.1249 - accuracy: 0.9682 - val_loss: 0.3170 - val_accuracy: 0.9223
-Epoch 534/1000
-60000/60000 - 5s - loss: 0.1249 - accuracy: 0.9685 - val_loss: 0.3185 - val_accuracy: 0.9222
-Epoch 535/1000
-60000/60000 - 5s - loss: 0.1248 - accuracy: 0.9681 - val_loss: 0.3186 - val_accuracy: 0.9221
-Epoch 536/1000
-60000/60000 - 5s - loss: 0.1247 - accuracy: 0.9681 - val_loss: 0.3191 - val_accuracy: 0.9221
-Epoch 537/1000
-60000/60000 - 5s - loss: 0.1247 - accuracy: 0.9679 - val_loss: 0.3182 - val_accuracy: 0.9215
-Epoch 538/1000
-60000/60000 - 5s - loss: 0.1247 - accuracy: 0.9681 - val_loss: 0.3186 - val_accuracy: 0.9218
-Epoch 539/1000
-60000/60000 - 5s - loss: 0.1245 - accuracy: 0.9680 - val_loss: 0.3187 - val_accuracy: 0.9213
-Epoch 540/1000
-60000/60000 - 5s - loss: 0.1243 - accuracy: 0.9683 - val_loss: 0.3192 - val_accuracy: 0.9224
-Epoch 541/1000
-60000/60000 - 5s - loss: 0.1244 - accuracy: 0.9684 - val_loss: 0.3187 - val_accuracy: 0.9227
-Epoch 542/1000
-60000/60000 - 5s - loss: 0.1243 - accuracy: 0.9685 - val_loss: 0.3198 - val_accuracy: 0.9225
-Epoch 543/1000
-60000/60000 - 5s - loss: 0.1241 - accuracy: 0.9681 - val_loss: 0.3191 - val_accuracy: 0.9225
-Epoch 544/1000
-60000/60000 - 5s - loss: 0.1241 - accuracy: 0.9684 - val_loss: 0.3194 - val_accuracy: 0.9222
-Epoch 545/1000
-60000/60000 - 5s - loss: 0.1240 - accuracy: 0.9679 - val_loss: 0.3192 - val_accuracy: 0.9221
-Epoch 546/1000
-60000/60000 - 5s - loss: 0.1239 - accuracy: 0.9682 - val_loss: 0.3194 - val_accuracy: 0.9223
-Epoch 547/1000
-60000/60000 - 5s - loss: 0.1239 - accuracy: 0.9683 - val_loss: 0.3195 - val_accuracy: 0.9213
-Epoch 548/1000
-60000/60000 - 5s - loss: 0.1239 - accuracy: 0.9686 - val_loss: 0.3194 - val_accuracy: 0.9229
-Epoch 549/1000
-60000/60000 - 5s - loss: 0.1237 - accuracy: 0.9684 - val_loss: 0.3200 - val_accuracy: 0.9217
-Epoch 550/1000
-60000/60000 - 5s - loss: 0.1238 - accuracy: 0.9685 - val_loss: 0.3205 - val_accuracy: 0.9219
-Epoch 551/1000
-60000/60000 - 5s - loss: 0.1237 - accuracy: 0.9687 - val_loss: 0.3203 - val_accuracy: 0.9225
-Epoch 552/1000
-60000/60000 - 5s - loss: 0.1235 - accuracy: 0.9686 - val_loss: 0.3201 - val_accuracy: 0.9225
-Epoch 553/1000
-60000/60000 - 5s - loss: 0.1236 - accuracy: 0.9684 - val_loss: 0.3201 - val_accuracy: 0.9221
-Epoch 554/1000
-60000/60000 - 5s - loss: 0.1234 - accuracy: 0.9687 - val_loss: 0.3200 - val_accuracy: 0.9216
-Epoch 555/1000
-60000/60000 - 5s - loss: 0.1234 - accuracy: 0.9684 - val_loss: 0.3206 - val_accuracy: 0.9227
-Epoch 556/1000
-60000/60000 - 5s - loss: 0.1232 - accuracy: 0.9686 - val_loss: 0.3204 - val_accuracy: 0.9228
-Epoch 557/1000
-60000/60000 - 5s - loss: 0.1232 - accuracy: 0.9686 - val_loss: 0.3203 - val_accuracy: 0.9225
-Epoch 558/1000
-60000/60000 - 5s - loss: 0.1231 - accuracy: 0.9686 - val_loss: 0.3206 - val_accuracy: 0.9222
-Epoch 559/1000
-60000/60000 - 5s - loss: 0.1230 - accuracy: 0.9686 - val_loss: 0.3213 - val_accuracy: 0.9213
-Epoch 560/1000
-60000/60000 - 5s - loss: 0.1230 - accuracy: 0.9687 - val_loss: 0.3206 - val_accuracy: 0.9220
-Epoch 561/1000
-60000/60000 - 5s - loss: 0.1228 - accuracy: 0.9691 - val_loss: 0.3208 - val_accuracy: 0.9218
-Epoch 562/1000
-60000/60000 - 5s - loss: 0.1227 - accuracy: 0.9686 - val_loss: 0.3211 - val_accuracy: 0.9214
-Epoch 563/1000
-60000/60000 - 5s - loss: 0.1228 - accuracy: 0.9687 - val_loss: 0.3215 - val_accuracy: 0.9226
-Epoch 564/1000
-60000/60000 - 5s - loss: 0.1227 - accuracy: 0.9688 - val_loss: 0.3206 - val_accuracy: 0.9226
-Epoch 565/1000
-60000/60000 - 5s - loss: 0.1225 - accuracy: 0.9690 - val_loss: 0.3215 - val_accuracy: 0.9224
-Epoch 566/1000
-60000/60000 - 5s - loss: 0.1225 - accuracy: 0.9689 - val_loss: 0.3223 - val_accuracy: 0.9218
-Epoch 567/1000
-60000/60000 - 5s - loss: 0.1226 - accuracy: 0.9686 - val_loss: 0.3220 - val_accuracy: 0.9216
-Epoch 568/1000
-60000/60000 - 5s - loss: 0.1225 - accuracy: 0.9687 - val_loss: 0.3217 - val_accuracy: 0.9217
-Epoch 569/1000
-60000/60000 - 5s - loss: 0.1224 - accuracy: 0.9692 - val_loss: 0.3220 - val_accuracy: 0.9217
-Epoch 570/1000
-60000/60000 - 5s - loss: 0.1223 - accuracy: 0.9689 - val_loss: 0.3218 - val_accuracy: 0.9209
-Epoch 571/1000
-60000/60000 - 5s - loss: 0.1221 - accuracy: 0.9689 - val_loss: 0.3221 - val_accuracy: 0.9216
-Epoch 572/1000
-60000/60000 - 5s - loss: 0.1222 - accuracy: 0.9692 - val_loss: 0.3229 - val_accuracy: 0.9205
-Epoch 573/1000
-60000/60000 - 5s - loss: 0.1221 - accuracy: 0.9693 - val_loss: 0.3226 - val_accuracy: 0.9224
-Epoch 574/1000
-60000/60000 - 5s - loss: 0.1219 - accuracy: 0.9692 - val_loss: 0.3235 - val_accuracy: 0.9218
-Epoch 575/1000
-60000/60000 - 5s - loss: 0.1219 - accuracy: 0.9693 - val_loss: 0.3224 - val_accuracy: 0.9216
-Epoch 576/1000
-60000/60000 - 5s - loss: 0.1218 - accuracy: 0.9688 - val_loss: 0.3227 - val_accuracy: 0.9223
-Epoch 577/1000
-60000/60000 - 5s - loss: 0.1219 - accuracy: 0.9692 - val_loss: 0.3230 - val_accuracy: 0.9220
-Epoch 578/1000
-60000/60000 - 5s - loss: 0.1218 - accuracy: 0.9691 - val_loss: 0.3230 - val_accuracy: 0.9210
-Epoch 579/1000
-60000/60000 - 5s - loss: 0.1217 - accuracy: 0.9693 - val_loss: 0.3234 - val_accuracy: 0.9221
-Epoch 580/1000
-60000/60000 - 5s - loss: 0.1215 - accuracy: 0.9694 - val_loss: 0.3239 - val_accuracy: 0.9215
-Epoch 581/1000
-60000/60000 - 5s - loss: 0.1216 - accuracy: 0.9691 - val_loss: 0.3238 - val_accuracy: 0.9216
-Epoch 582/1000
-60000/60000 - 5s - loss: 0.1215 - accuracy: 0.9694 - val_loss: 0.3231 - val_accuracy: 0.9216
-Epoch 583/1000
-60000/60000 - 5s - loss: 0.1214 - accuracy: 0.9692 - val_loss: 0.3235 - val_accuracy: 0.9213
-Epoch 584/1000
-60000/60000 - 5s - loss: 0.1213 - accuracy: 0.9693 - val_loss: 0.3245 - val_accuracy: 0.9219
-Epoch 585/1000
-60000/60000 - 5s - loss: 0.1213 - accuracy: 0.9697 - val_loss: 0.3237 - val_accuracy: 0.9216
-Epoch 586/1000
-60000/60000 - 5s - loss: 0.1213 - accuracy: 0.9692 - val_loss: 0.3242 - val_accuracy: 0.9219
-Epoch 587/1000
-60000/60000 - 5s - loss: 0.1212 - accuracy: 0.9694 - val_loss: 0.3243 - val_accuracy: 0.9213
-Epoch 588/1000
-60000/60000 - 5s - loss: 0.1211 - accuracy: 0.9693 - val_loss: 0.3241 - val_accuracy: 0.9219
-Epoch 589/1000
-60000/60000 - 5s - loss: 0.1211 - accuracy: 0.9694 - val_loss: 0.3245 - val_accuracy: 0.9215
-Epoch 590/1000
-60000/60000 - 5s - loss: 0.1209 - accuracy: 0.9692 - val_loss: 0.3242 - val_accuracy: 0.9209
-Epoch 591/1000
-60000/60000 - 5s - loss: 0.1209 - accuracy: 0.9693 - val_loss: 0.3248 - val_accuracy: 0.9210
-Epoch 592/1000
-60000/60000 - 5s - loss: 0.1208 - accuracy: 0.9693 - val_loss: 0.3244 - val_accuracy: 0.9218
-Epoch 593/1000
-60000/60000 - 5s - loss: 0.1208 - accuracy: 0.9696 - val_loss: 0.3254 - val_accuracy: 0.9218
-Epoch 594/1000
-60000/60000 - 5s - loss: 0.1208 - accuracy: 0.9696 - val_loss: 0.3243 - val_accuracy: 0.9208
-Epoch 595/1000
-60000/60000 - 5s - loss: 0.1207 - accuracy: 0.9695 - val_loss: 0.3256 - val_accuracy: 0.9207
-Epoch 596/1000
-60000/60000 - 5s - loss: 0.1206 - accuracy: 0.9694 - val_loss: 0.3251 - val_accuracy: 0.9209
-Epoch 597/1000
-60000/60000 - 5s - loss: 0.1206 - accuracy: 0.9696 - val_loss: 0.3262 - val_accuracy: 0.9209
-Epoch 598/1000
-60000/60000 - 5s - loss: 0.1205 - accuracy: 0.9697 - val_loss: 0.3256 - val_accuracy: 0.9217
-Epoch 599/1000
-60000/60000 - 5s - loss: 0.1204 - accuracy: 0.9696 - val_loss: 0.3255 - val_accuracy: 0.9214
-Epoch 600/1000
-60000/60000 - 5s - loss: 0.1204 - accuracy: 0.9694 - val_loss: 0.3262 - val_accuracy: 0.9223
-Epoch 601/1000
-60000/60000 - 5s - loss: 0.1204 - accuracy: 0.9695 - val_loss: 0.3258 - val_accuracy: 0.9219
-Epoch 602/1000
-60000/60000 - 5s - loss: 0.1203 - accuracy: 0.9699 - val_loss: 0.3258 - val_accuracy: 0.9203
-Epoch 603/1000
-60000/60000 - 5s - loss: 0.1203 - accuracy: 0.9695 - val_loss: 0.3262 - val_accuracy: 0.9213
-Epoch 604/1000
-60000/60000 - 5s - loss: 0.1202 - accuracy: 0.9698 - val_loss: 0.3259 - val_accuracy: 0.9201
-Epoch 605/1000
-60000/60000 - 5s - loss: 0.1200 - accuracy: 0.9698 - val_loss: 0.3268 - val_accuracy: 0.9210
-Epoch 606/1000
-60000/60000 - 5s - loss: 0.1201 - accuracy: 0.9699 - val_loss: 0.3265 - val_accuracy: 0.9203
-Epoch 607/1000
-60000/60000 - 5s - loss: 0.1200 - accuracy: 0.9698 - val_loss: 0.3271 - val_accuracy: 0.9205
-Epoch 608/1000
-60000/60000 - 5s - loss: 0.1199 - accuracy: 0.9698 - val_loss: 0.3268 - val_accuracy: 0.9199
-Epoch 609/1000
-60000/60000 - 5s - loss: 0.1198 - accuracy: 0.9697 - val_loss: 0.3262 - val_accuracy: 0.9207
-Epoch 610/1000
-60000/60000 - 5s - loss: 0.1198 - accuracy: 0.9698 - val_loss: 0.3270 - val_accuracy: 0.9210
-Epoch 611/1000
-60000/60000 - 5s - loss: 0.1197 - accuracy: 0.9698 - val_loss: 0.3275 - val_accuracy: 0.9207
-Epoch 612/1000
-60000/60000 - 5s - loss: 0.1196 - accuracy: 0.9698 - val_loss: 0.3271 - val_accuracy: 0.9215
-Epoch 613/1000
-60000/60000 - 5s - loss: 0.1196 - accuracy: 0.9700 - val_loss: 0.3271 - val_accuracy: 0.9202
-Epoch 614/1000
-60000/60000 - 5s - loss: 0.1196 - accuracy: 0.9700 - val_loss: 0.3279 - val_accuracy: 0.9211
-Epoch 615/1000
-60000/60000 - 5s - loss: 0.1196 - accuracy: 0.9700 - val_loss: 0.3276 - val_accuracy: 0.9200
-Epoch 616/1000
-60000/60000 - 5s - loss: 0.1194 - accuracy: 0.9697 - val_loss: 0.3279 - val_accuracy: 0.9208
-Epoch 617/1000
-60000/60000 - 5s - loss: 0.1194 - accuracy: 0.9701 - val_loss: 0.3278 - val_accuracy: 0.9208
-Epoch 618/1000
-60000/60000 - 5s - loss: 0.1193 - accuracy: 0.9699 - val_loss: 0.3278 - val_accuracy: 0.9209
-Epoch 619/1000
-60000/60000 - 5s - loss: 0.1192 - accuracy: 0.9704 - val_loss: 0.3288 - val_accuracy: 0.9209
-Epoch 620/1000
-60000/60000 - 5s - loss: 0.1192 - accuracy: 0.9703 - val_loss: 0.3285 - val_accuracy: 0.9202
-Epoch 621/1000
-60000/60000 - 5s - loss: 0.1192 - accuracy: 0.9701 - val_loss: 0.3281 - val_accuracy: 0.9198
-Epoch 622/1000
-60000/60000 - 5s - loss: 0.1191 - accuracy: 0.9701 - val_loss: 0.3282 - val_accuracy: 0.9209
-Epoch 623/1000
-60000/60000 - 5s - loss: 0.1190 - accuracy: 0.9699 - val_loss: 0.3288 - val_accuracy: 0.9204
-Epoch 624/1000
-60000/60000 - 5s - loss: 0.1190 - accuracy: 0.9701 - val_loss: 0.3285 - val_accuracy: 0.9199
-Epoch 625/1000
-60000/60000 - 5s - loss: 0.1189 - accuracy: 0.9701 - val_loss: 0.3293 - val_accuracy: 0.9206
-Epoch 626/1000
-60000/60000 - 5s - loss: 0.1188 - accuracy: 0.9702 - val_loss: 0.3289 - val_accuracy: 0.9198
-Epoch 627/1000
-60000/60000 - 5s - loss: 0.1188 - accuracy: 0.9702 - val_loss: 0.3296 - val_accuracy: 0.9204
-Epoch 628/1000
-60000/60000 - 4s - loss: 0.1188 - accuracy: 0.9702 - val_loss: 0.3295 - val_accuracy: 0.9200
-Epoch 629/1000
-60000/60000 - 4s - loss: 0.1187 - accuracy: 0.9701 - val_loss: 0.3291 - val_accuracy: 0.9202
-Epoch 630/1000
-60000/60000 - 5s - loss: 0.1187 - accuracy: 0.9698 - val_loss: 0.3299 - val_accuracy: 0.9206
-Epoch 631/1000
-60000/60000 - 5s - loss: 0.1186 - accuracy: 0.9704 - val_loss: 0.3301 - val_accuracy: 0.9196
-Epoch 632/1000
-60000/60000 - 5s - loss: 0.1185 - accuracy: 0.9701 - val_loss: 0.3294 - val_accuracy: 0.9212
-Epoch 633/1000
-60000/60000 - 5s - loss: 0.1185 - accuracy: 0.9702 - val_loss: 0.3303 - val_accuracy: 0.9204
-Epoch 634/1000
-60000/60000 - 5s - loss: 0.1183 - accuracy: 0.9703 - val_loss: 0.3300 - val_accuracy: 0.9204
-Epoch 635/1000
-60000/60000 - 5s - loss: 0.1184 - accuracy: 0.9702 - val_loss: 0.3303 - val_accuracy: 0.9199
-Epoch 636/1000
-60000/60000 - 5s - loss: 0.1183 - accuracy: 0.9704 - val_loss: 0.3304 - val_accuracy: 0.9212
-Epoch 637/1000
-60000/60000 - 5s - loss: 0.1183 - accuracy: 0.9705 - val_loss: 0.3304 - val_accuracy: 0.9208
-Epoch 638/1000
-60000/60000 - 5s - loss: 0.1183 - accuracy: 0.9703 - val_loss: 0.3301 - val_accuracy: 0.9208
-Epoch 639/1000
-60000/60000 - 5s - loss: 0.1181 - accuracy: 0.9704 - val_loss: 0.3308 - val_accuracy: 0.9205
-Epoch 640/1000
-60000/60000 - 5s - loss: 0.1181 - accuracy: 0.9703 - val_loss: 0.3309 - val_accuracy: 0.9202
-Epoch 641/1000
-60000/60000 - 5s - loss: 0.1181 - accuracy: 0.9700 - val_loss: 0.3316 - val_accuracy: 0.9203
-Epoch 642/1000
-60000/60000 - 5s - loss: 0.1179 - accuracy: 0.9703 - val_loss: 0.3312 - val_accuracy: 0.9212
-Epoch 643/1000
-60000/60000 - 5s - loss: 0.1179 - accuracy: 0.9703 - val_loss: 0.3314 - val_accuracy: 0.9201
-Epoch 644/1000
-60000/60000 - 5s - loss: 0.1179 - accuracy: 0.9704 - val_loss: 0.3312 - val_accuracy: 0.9205
-Epoch 645/1000
-60000/60000 - 5s - loss: 0.1178 - accuracy: 0.9706 - val_loss: 0.3314 - val_accuracy: 0.9202
-Epoch 646/1000
-60000/60000 - 5s - loss: 0.1178 - accuracy: 0.9703 - val_loss: 0.3320 - val_accuracy: 0.9191
-Epoch 647/1000
-60000/60000 - 5s - loss: 0.1177 - accuracy: 0.9703 - val_loss: 0.3313 - val_accuracy: 0.9205
-Epoch 648/1000
-60000/60000 - 5s - loss: 0.1178 - accuracy: 0.9704 - val_loss: 0.3320 - val_accuracy: 0.9196
-Epoch 649/1000
-60000/60000 - 5s - loss: 0.1177 - accuracy: 0.9702 - val_loss: 0.3320 - val_accuracy: 0.9209
-Epoch 650/1000
-60000/60000 - 5s - loss: 0.1176 - accuracy: 0.9706 - val_loss: 0.3316 - val_accuracy: 0.9209
-Epoch 651/1000
-60000/60000 - 5s - loss: 0.1175 - accuracy: 0.9705 - val_loss: 0.3323 - val_accuracy: 0.9200
-Epoch 652/1000
-60000/60000 - 5s - loss: 0.1175 - accuracy: 0.9705 - val_loss: 0.3315 - val_accuracy: 0.9204
-Epoch 653/1000
-60000/60000 - 5s - loss: 0.1175 - accuracy: 0.9704 - val_loss: 0.3325 - val_accuracy: 0.9207
-Epoch 654/1000
-60000/60000 - 5s - loss: 0.1173 - accuracy: 0.9708 - val_loss: 0.3327 - val_accuracy: 0.9204
-Epoch 655/1000
-60000/60000 - 5s - loss: 0.1173 - accuracy: 0.9707 - val_loss: 0.3325 - val_accuracy: 0.9210
-Epoch 656/1000
-60000/60000 - 5s - loss: 0.1174 - accuracy: 0.9705 - val_loss: 0.3322 - val_accuracy: 0.9200
-Epoch 657/1000
-60000/60000 - 5s - loss: 0.1173 - accuracy: 0.9705 - val_loss: 0.3326 - val_accuracy: 0.9210
-Epoch 658/1000
-60000/60000 - 5s - loss: 0.1172 - accuracy: 0.9704 - val_loss: 0.3332 - val_accuracy: 0.9199
-Epoch 659/1000
-60000/60000 - 5s - loss: 0.1171 - accuracy: 0.9707 - val_loss: 0.3329 - val_accuracy: 0.9202
-Epoch 660/1000
-60000/60000 - 5s - loss: 0.1171 - accuracy: 0.9707 - val_loss: 0.3327 - val_accuracy: 0.9200
-Epoch 661/1000
-60000/60000 - 5s - loss: 0.1171 - accuracy: 0.9705 - val_loss: 0.3330 - val_accuracy: 0.9208
-Epoch 662/1000
-60000/60000 - 5s - loss: 0.1169 - accuracy: 0.9705 - val_loss: 0.3328 - val_accuracy: 0.9197
-Epoch 663/1000
-60000/60000 - 5s - loss: 0.1169 - accuracy: 0.9705 - val_loss: 0.3334 - val_accuracy: 0.9206
-Epoch 664/1000
-60000/60000 - 5s - loss: 0.1169 - accuracy: 0.9704 - val_loss: 0.3324 - val_accuracy: 0.9205
-Epoch 665/1000
-60000/60000 - 5s - loss: 0.1168 - accuracy: 0.9707 - val_loss: 0.3337 - val_accuracy: 0.9194
-Epoch 666/1000
-60000/60000 - 5s - loss: 0.1168 - accuracy: 0.9707 - val_loss: 0.3335 - val_accuracy: 0.9208
-Epoch 667/1000
-60000/60000 - 5s - loss: 0.1167 - accuracy: 0.9706 - val_loss: 0.3338 - val_accuracy: 0.9203
-Epoch 668/1000
-60000/60000 - 5s - loss: 0.1167 - accuracy: 0.9708 - val_loss: 0.3340 - val_accuracy: 0.9206
-Epoch 669/1000
-60000/60000 - 5s - loss: 0.1165 - accuracy: 0.9707 - val_loss: 0.3338 - val_accuracy: 0.9198
-Epoch 670/1000
-60000/60000 - 5s - loss: 0.1164 - accuracy: 0.9706 - val_loss: 0.3338 - val_accuracy: 0.9206
-Epoch 671/1000
-60000/60000 - 5s - loss: 0.1165 - accuracy: 0.9707 - val_loss: 0.3341 - val_accuracy: 0.9206
-Epoch 672/1000
-60000/60000 - 5s - loss: 0.1164 - accuracy: 0.9708 - val_loss: 0.3341 - val_accuracy: 0.9204
-Epoch 673/1000
-60000/60000 - 5s - loss: 0.1163 - accuracy: 0.9708 - val_loss: 0.3341 - val_accuracy: 0.9201
-Epoch 674/1000
-60000/60000 - 5s - loss: 0.1164 - accuracy: 0.9709 - val_loss: 0.3336 - val_accuracy: 0.9198
-Epoch 675/1000
-60000/60000 - 5s - loss: 0.1163 - accuracy: 0.9706 - val_loss: 0.3340 - val_accuracy: 0.9196
-Epoch 676/1000
-60000/60000 - 5s - loss: 0.1162 - accuracy: 0.9708 - val_loss: 0.3343 - val_accuracy: 0.9197
-Epoch 677/1000
-60000/60000 - 5s - loss: 0.1161 - accuracy: 0.9708 - val_loss: 0.3343 - val_accuracy: 0.9196
-Epoch 678/1000
-60000/60000 - 5s - loss: 0.1162 - accuracy: 0.9705 - val_loss: 0.3350 - val_accuracy: 0.9205
-Epoch 679/1000
-60000/60000 - 5s - loss: 0.1160 - accuracy: 0.9709 - val_loss: 0.3344 - val_accuracy: 0.9205
-Epoch 680/1000
-60000/60000 - 5s - loss: 0.1160 - accuracy: 0.9709 - val_loss: 0.3343 - val_accuracy: 0.9194
-Epoch 681/1000
-60000/60000 - 5s - loss: 0.1159 - accuracy: 0.9709 - val_loss: 0.3354 - val_accuracy: 0.9204
-Epoch 682/1000
-60000/60000 - 5s - loss: 0.1159 - accuracy: 0.9710 - val_loss: 0.3349 - val_accuracy: 0.9199
-Epoch 683/1000
-60000/60000 - 5s - loss: 0.1159 - accuracy: 0.9711 - val_loss: 0.3351 - val_accuracy: 0.9204
-Epoch 684/1000
-60000/60000 - 5s - loss: 0.1158 - accuracy: 0.9710 - val_loss: 0.3353 - val_accuracy: 0.9201
-Epoch 685/1000
-60000/60000 - 5s - loss: 0.1157 - accuracy: 0.9711 - val_loss: 0.3356 - val_accuracy: 0.9200
-Epoch 686/1000
-60000/60000 - 5s - loss: 0.1157 - accuracy: 0.9708 - val_loss: 0.3358 - val_accuracy: 0.9204
-Epoch 687/1000
-60000/60000 - 5s - loss: 0.1157 - accuracy: 0.9709 - val_loss: 0.3357 - val_accuracy: 0.9206
-Epoch 688/1000
-60000/60000 - 5s - loss: 0.1156 - accuracy: 0.9710 - val_loss: 0.3354 - val_accuracy: 0.9199
-Epoch 689/1000
-60000/60000 - 5s - loss: 0.1155 - accuracy: 0.9710 - val_loss: 0.3358 - val_accuracy: 0.9199
-Epoch 690/1000
-60000/60000 - 5s - loss: 0.1155 - accuracy: 0.9710 - val_loss: 0.3361 - val_accuracy: 0.9202
-Epoch 691/1000
-60000/60000 - 5s - loss: 0.1154 - accuracy: 0.9711 - val_loss: 0.3359 - val_accuracy: 0.9203
-Epoch 692/1000
-60000/60000 - 5s - loss: 0.1154 - accuracy: 0.9714 - val_loss: 0.3365 - val_accuracy: 0.9203
-Epoch 693/1000
-60000/60000 - 5s - loss: 0.1153 - accuracy: 0.9708 - val_loss: 0.3361 - val_accuracy: 0.9205
-Epoch 694/1000
-60000/60000 - 5s - loss: 0.1153 - accuracy: 0.9711 - val_loss: 0.3359 - val_accuracy: 0.9205
-Epoch 695/1000
-60000/60000 - 5s - loss: 0.1151 - accuracy: 0.9713 - val_loss: 0.3355 - val_accuracy: 0.9198
-Epoch 696/1000
-60000/60000 - 5s - loss: 0.1151 - accuracy: 0.9712 - val_loss: 0.3363 - val_accuracy: 0.9205
-Epoch 697/1000
-60000/60000 - 5s - loss: 0.1151 - accuracy: 0.9714 - val_loss: 0.3363 - val_accuracy: 0.9197
-Epoch 698/1000
-60000/60000 - 5s - loss: 0.1150 - accuracy: 0.9711 - val_loss: 0.3361 - val_accuracy: 0.9203
-Epoch 699/1000
-60000/60000 - 5s - loss: 0.1150 - accuracy: 0.9712 - val_loss: 0.3362 - val_accuracy: 0.9203
-Epoch 700/1000
-60000/60000 - 5s - loss: 0.1149 - accuracy: 0.9711 - val_loss: 0.3364 - val_accuracy: 0.9200
-Epoch 701/1000
-60000/60000 - 5s - loss: 0.1149 - accuracy: 0.9714 - val_loss: 0.3370 - val_accuracy: 0.9201
-Epoch 702/1000
-60000/60000 - 5s - loss: 0.1148 - accuracy: 0.9715 - val_loss: 0.3372 - val_accuracy: 0.9200
-Epoch 703/1000
-60000/60000 - 5s - loss: 0.1148 - accuracy: 0.9715 - val_loss: 0.3370 - val_accuracy: 0.9193
-Epoch 704/1000
-60000/60000 - 5s - loss: 0.1147 - accuracy: 0.9712 - val_loss: 0.3364 - val_accuracy: 0.9196
-Epoch 705/1000
-60000/60000 - 5s - loss: 0.1148 - accuracy: 0.9714 - val_loss: 0.3372 - val_accuracy: 0.9190
-Epoch 706/1000
-60000/60000 - 5s - loss: 0.1147 - accuracy: 0.9712 - val_loss: 0.3371 - val_accuracy: 0.9207
-Epoch 707/1000
-60000/60000 - 5s - loss: 0.1147 - accuracy: 0.9713 - val_loss: 0.3376 - val_accuracy: 0.9193
-Epoch 708/1000
-60000/60000 - 5s - loss: 0.1147 - accuracy: 0.9712 - val_loss: 0.3371 - val_accuracy: 0.9196
-Epoch 709/1000
-60000/60000 - 5s - loss: 0.1146 - accuracy: 0.9711 - val_loss: 0.3372 - val_accuracy: 0.9197
-Epoch 710/1000
-60000/60000 - 5s - loss: 0.1145 - accuracy: 0.9712 - val_loss: 0.3377 - val_accuracy: 0.9191
-Epoch 711/1000
-60000/60000 - 5s - loss: 0.1145 - accuracy: 0.9716 - val_loss: 0.3375 - val_accuracy: 0.9198
-Epoch 712/1000
-60000/60000 - 5s - loss: 0.1144 - accuracy: 0.9716 - val_loss: 0.3378 - val_accuracy: 0.9194
-Epoch 713/1000
-60000/60000 - 5s - loss: 0.1144 - accuracy: 0.9717 - val_loss: 0.3380 - val_accuracy: 0.9198
-Epoch 714/1000
-60000/60000 - 5s - loss: 0.1142 - accuracy: 0.9716 - val_loss: 0.3376 - val_accuracy: 0.9193
-Epoch 715/1000
-60000/60000 - 5s - loss: 0.1143 - accuracy: 0.9716 - val_loss: 0.3372 - val_accuracy: 0.9204
-Epoch 716/1000
-60000/60000 - 5s - loss: 0.1142 - accuracy: 0.9717 - val_loss: 0.3380 - val_accuracy: 0.9206
-Epoch 717/1000
-60000/60000 - 5s - loss: 0.1142 - accuracy: 0.9712 - val_loss: 0.3378 - val_accuracy: 0.9202
-Epoch 718/1000
-60000/60000 - 5s - loss: 0.1141 - accuracy: 0.9717 - val_loss: 0.3373 - val_accuracy: 0.9208
-Epoch 719/1000
-60000/60000 - 5s - loss: 0.1140 - accuracy: 0.9714 - val_loss: 0.3380 - val_accuracy: 0.9205
-Epoch 720/1000
-60000/60000 - 5s - loss: 0.1139 - accuracy: 0.9719 - val_loss: 0.3381 - val_accuracy: 0.9203
-Epoch 721/1000
-60000/60000 - 5s - loss: 0.1140 - accuracy: 0.9717 - val_loss: 0.3380 - val_accuracy: 0.9195
-Epoch 722/1000
-60000/60000 - 5s - loss: 0.1138 - accuracy: 0.9716 - val_loss: 0.3379 - val_accuracy: 0.9208
-Epoch 723/1000
-60000/60000 - 5s - loss: 0.1139 - accuracy: 0.9720 - val_loss: 0.3378 - val_accuracy: 0.9207
-Epoch 724/1000
-60000/60000 - 5s - loss: 0.1139 - accuracy: 0.9715 - val_loss: 0.3375 - val_accuracy: 0.9201
-Epoch 725/1000
-60000/60000 - 5s - loss: 0.1138 - accuracy: 0.9712 - val_loss: 0.3384 - val_accuracy: 0.9198
-Epoch 726/1000
-60000/60000 - 5s - loss: 0.1137 - accuracy: 0.9718 - val_loss: 0.3382 - val_accuracy: 0.9195
-Epoch 727/1000
-60000/60000 - 5s - loss: 0.1137 - accuracy: 0.9715 - val_loss: 0.3386 - val_accuracy: 0.9197
-Epoch 728/1000
-60000/60000 - 5s - loss: 0.1136 - accuracy: 0.9714 - val_loss: 0.3380 - val_accuracy: 0.9193
-Epoch 729/1000
-60000/60000 - 5s - loss: 0.1136 - accuracy: 0.9717 - val_loss: 0.3376 - val_accuracy: 0.9197
-Epoch 730/1000
-60000/60000 - 5s - loss: 0.1136 - accuracy: 0.9717 - val_loss: 0.3383 - val_accuracy: 0.9211
-Epoch 731/1000
-60000/60000 - 5s - loss: 0.1135 - accuracy: 0.9720 - val_loss: 0.3381 - val_accuracy: 0.9209
-Epoch 732/1000
-60000/60000 - 5s - loss: 0.1134 - accuracy: 0.9715 - val_loss: 0.3390 - val_accuracy: 0.9206
-Epoch 733/1000
-60000/60000 - 5s - loss: 0.1133 - accuracy: 0.9718 - val_loss: 0.3388 - val_accuracy: 0.9204
-Epoch 734/1000
-60000/60000 - 5s - loss: 0.1133 - accuracy: 0.9719 - val_loss: 0.3388 - val_accuracy: 0.9194
-Epoch 735/1000
-60000/60000 - 5s - loss: 0.1133 - accuracy: 0.9717 - val_loss: 0.3381 - val_accuracy: 0.9192
-Epoch 736/1000
-60000/60000 - 5s - loss: 0.1133 - accuracy: 0.9721 - val_loss: 0.3388 - val_accuracy: 0.9204
-Epoch 737/1000
-60000/60000 - 5s - loss: 0.1132 - accuracy: 0.9718 - val_loss: 0.3391 - val_accuracy: 0.9208
-Epoch 738/1000
-60000/60000 - 5s - loss: 0.1132 - accuracy: 0.9722 - val_loss: 0.3388 - val_accuracy: 0.9197
-Epoch 739/1000
-60000/60000 - 5s - loss: 0.1131 - accuracy: 0.9716 - val_loss: 0.3392 - val_accuracy: 0.9193
-Epoch 740/1000
-60000/60000 - 5s - loss: 0.1131 - accuracy: 0.9718 - val_loss: 0.3395 - val_accuracy: 0.9208
-Epoch 741/1000
-60000/60000 - 5s - loss: 0.1130 - accuracy: 0.9717 - val_loss: 0.3390 - val_accuracy: 0.9210
-Epoch 742/1000
-60000/60000 - 5s - loss: 0.1130 - accuracy: 0.9720 - val_loss: 0.3394 - val_accuracy: 0.9194
-Epoch 743/1000
-60000/60000 - 5s - loss: 0.1129 - accuracy: 0.9720 - val_loss: 0.3398 - val_accuracy: 0.9207
-Epoch 744/1000
-60000/60000 - 5s - loss: 0.1129 - accuracy: 0.9719 - val_loss: 0.3393 - val_accuracy: 0.9199
-Epoch 745/1000
-60000/60000 - 5s - loss: 0.1128 - accuracy: 0.9719 - val_loss: 0.3387 - val_accuracy: 0.9197
-Epoch 746/1000
-60000/60000 - 5s - loss: 0.1128 - accuracy: 0.9721 - val_loss: 0.3392 - val_accuracy: 0.9192
-Epoch 747/1000
-60000/60000 - 5s - loss: 0.1128 - accuracy: 0.9718 - val_loss: 0.3398 - val_accuracy: 0.9200
-Epoch 748/1000
-60000/60000 - 5s - loss: 0.1128 - accuracy: 0.9720 - val_loss: 0.3400 - val_accuracy: 0.9196
-Epoch 749/1000
-60000/60000 - 5s - loss: 0.1127 - accuracy: 0.9719 - val_loss: 0.3395 - val_accuracy: 0.9209
-Epoch 750/1000
-60000/60000 - 5s - loss: 0.1127 - accuracy: 0.9722 - val_loss: 0.3397 - val_accuracy: 0.9201
-Epoch 751/1000
-60000/60000 - 5s - loss: 0.1127 - accuracy: 0.9721 - val_loss: 0.3397 - val_accuracy: 0.9197
-Epoch 752/1000
-60000/60000 - 5s - loss: 0.1126 - accuracy: 0.9720 - val_loss: 0.3401 - val_accuracy: 0.9200
-Epoch 753/1000
-60000/60000 - 5s - loss: 0.1125 - accuracy: 0.9720 - val_loss: 0.3399 - val_accuracy: 0.9196
-Epoch 754/1000
-60000/60000 - 5s - loss: 0.1125 - accuracy: 0.9721 - val_loss: 0.3407 - val_accuracy: 0.9204
-Epoch 755/1000
-60000/60000 - 5s - loss: 0.1124 - accuracy: 0.9721 - val_loss: 0.3408 - val_accuracy: 0.9202
-Epoch 756/1000
-60000/60000 - 5s - loss: 0.1124 - accuracy: 0.9720 - val_loss: 0.3397 - val_accuracy: 0.9199
-Epoch 757/1000
-60000/60000 - 5s - loss: 0.1124 - accuracy: 0.9723 - val_loss: 0.3405 - val_accuracy: 0.9199
-Epoch 758/1000
-60000/60000 - 5s - loss: 0.1123 - accuracy: 0.9718 - val_loss: 0.3406 - val_accuracy: 0.9200
-Epoch 759/1000
-60000/60000 - 5s - loss: 0.1122 - accuracy: 0.9721 - val_loss: 0.3407 - val_accuracy: 0.9192
-Epoch 760/1000
-60000/60000 - 5s - loss: 0.1122 - accuracy: 0.9721 - val_loss: 0.3401 - val_accuracy: 0.9202
-Epoch 761/1000
-60000/60000 - 5s - loss: 0.1122 - accuracy: 0.9717 - val_loss: 0.3410 - val_accuracy: 0.9186
-Epoch 762/1000
-60000/60000 - 5s - loss: 0.1122 - accuracy: 0.9718 - val_loss: 0.3411 - val_accuracy: 0.9204
-Epoch 763/1000
-60000/60000 - 5s - loss: 0.1121 - accuracy: 0.9722 - val_loss: 0.3419 - val_accuracy: 0.9187
-Epoch 764/1000
-60000/60000 - 5s - loss: 0.1121 - accuracy: 0.9722 - val_loss: 0.3408 - val_accuracy: 0.9202
-Epoch 765/1000
-60000/60000 - 5s - loss: 0.1121 - accuracy: 0.9725 - val_loss: 0.3415 - val_accuracy: 0.9195
-Epoch 766/1000
-60000/60000 - 5s - loss: 0.1120 - accuracy: 0.9721 - val_loss: 0.3417 - val_accuracy: 0.9201
-Epoch 767/1000
-60000/60000 - 5s - loss: 0.1120 - accuracy: 0.9718 - val_loss: 0.3413 - val_accuracy: 0.9193
-Epoch 768/1000
-60000/60000 - 5s - loss: 0.1119 - accuracy: 0.9725 - val_loss: 0.3415 - val_accuracy: 0.9193
-Epoch 769/1000
-60000/60000 - 5s - loss: 0.1119 - accuracy: 0.9720 - val_loss: 0.3407 - val_accuracy: 0.9200
-Epoch 770/1000
-60000/60000 - 5s - loss: 0.1118 - accuracy: 0.9718 - val_loss: 0.3415 - val_accuracy: 0.9195
-Epoch 771/1000
-60000/60000 - 5s - loss: 0.1118 - accuracy: 0.9724 - val_loss: 0.3415 - val_accuracy: 0.9198
-Epoch 772/1000
-60000/60000 - 5s - loss: 0.1117 - accuracy: 0.9721 - val_loss: 0.3415 - val_accuracy: 0.9191
-Epoch 773/1000
-60000/60000 - 5s - loss: 0.1117 - accuracy: 0.9722 - val_loss: 0.3421 - val_accuracy: 0.9198
-Epoch 774/1000
-60000/60000 - 5s - loss: 0.1117 - accuracy: 0.9724 - val_loss: 0.3415 - val_accuracy: 0.9203
-Epoch 775/1000
-60000/60000 - 5s - loss: 0.1117 - accuracy: 0.9722 - val_loss: 0.3416 - val_accuracy: 0.9203
-Epoch 776/1000
-60000/60000 - 5s - loss: 0.1116 - accuracy: 0.9721 - val_loss: 0.3417 - val_accuracy: 0.9202
-Epoch 777/1000
-60000/60000 - 5s - loss: 0.1115 - accuracy: 0.9723 - val_loss: 0.3422 - val_accuracy: 0.9188
-Epoch 778/1000
-60000/60000 - 5s - loss: 0.1114 - accuracy: 0.9724 - val_loss: 0.3419 - val_accuracy: 0.9192
-Epoch 779/1000
-60000/60000 - 5s - loss: 0.1115 - accuracy: 0.9723 - val_loss: 0.3418 - val_accuracy: 0.9193
-Epoch 780/1000
-60000/60000 - 5s - loss: 0.1114 - accuracy: 0.9722 - val_loss: 0.3422 - val_accuracy: 0.9188
-Epoch 781/1000
-60000/60000 - 5s - loss: 0.1114 - accuracy: 0.9724 - val_loss: 0.3422 - val_accuracy: 0.9198
-Epoch 782/1000
-60000/60000 - 5s - loss: 0.1114 - accuracy: 0.9725 - val_loss: 0.3416 - val_accuracy: 0.9208
-Epoch 783/1000
-60000/60000 - 5s - loss: 0.1114 - accuracy: 0.9724 - val_loss: 0.3427 - val_accuracy: 0.9201
-Epoch 784/1000
-60000/60000 - 5s - loss: 0.1113 - accuracy: 0.9722 - val_loss: 0.3415 - val_accuracy: 0.9196
-Epoch 785/1000
-60000/60000 - 5s - loss: 0.1113 - accuracy: 0.9724 - val_loss: 0.3419 - val_accuracy: 0.9199
-Epoch 786/1000
-60000/60000 - 5s - loss: 0.1112 - accuracy: 0.9724 - val_loss: 0.3422 - val_accuracy: 0.9192
-Epoch 787/1000
-60000/60000 - 5s - loss: 0.1112 - accuracy: 0.9723 - val_loss: 0.3423 - val_accuracy: 0.9196
-Epoch 788/1000
-60000/60000 - 5s - loss: 0.1112 - accuracy: 0.9722 - val_loss: 0.3426 - val_accuracy: 0.9206
-Epoch 789/1000
-60000/60000 - 5s - loss: 0.1111 - accuracy: 0.9725 - val_loss: 0.3421 - val_accuracy: 0.9197
-Epoch 790/1000
-60000/60000 - 5s - loss: 0.1111 - accuracy: 0.9725 - val_loss: 0.3426 - val_accuracy: 0.9204
-Epoch 791/1000
-60000/60000 - 5s - loss: 0.1110 - accuracy: 0.9724 - val_loss: 0.3422 - val_accuracy: 0.9200
-Epoch 792/1000
-60000/60000 - 5s - loss: 0.1109 - accuracy: 0.9726 - val_loss: 0.3428 - val_accuracy: 0.9200
-Epoch 793/1000
-60000/60000 - 5s - loss: 0.1109 - accuracy: 0.9723 - val_loss: 0.3423 - val_accuracy: 0.9206
-Epoch 794/1000
-60000/60000 - 5s - loss: 0.1110 - accuracy: 0.9725 - val_loss: 0.3429 - val_accuracy: 0.9191
-Epoch 795/1000
-60000/60000 - 5s - loss: 0.1109 - accuracy: 0.9726 - val_loss: 0.3426 - val_accuracy: 0.9210
-Epoch 796/1000
-60000/60000 - 5s - loss: 0.1108 - accuracy: 0.9727 - val_loss: 0.3429 - val_accuracy: 0.9198
-Epoch 797/1000
-60000/60000 - 5s - loss: 0.1108 - accuracy: 0.9726 - val_loss: 0.3427 - val_accuracy: 0.9190
-Epoch 798/1000
-60000/60000 - 5s - loss: 0.1107 - accuracy: 0.9725 - val_loss: 0.3432 - val_accuracy: 0.9197
-Epoch 799/1000
-60000/60000 - 5s - loss: 0.1107 - accuracy: 0.9727 - val_loss: 0.3426 - val_accuracy: 0.9198
-Epoch 800/1000
-60000/60000 - 5s - loss: 0.1107 - accuracy: 0.9729 - val_loss: 0.3431 - val_accuracy: 0.9189
-Epoch 801/1000
-60000/60000 - 5s - loss: 0.1107 - accuracy: 0.9727 - val_loss: 0.3432 - val_accuracy: 0.9199
-Epoch 802/1000
-60000/60000 - 5s - loss: 0.1105 - accuracy: 0.9725 - val_loss: 0.3434 - val_accuracy: 0.9202
-Epoch 803/1000
-60000/60000 - 5s - loss: 0.1106 - accuracy: 0.9727 - val_loss: 0.3431 - val_accuracy: 0.9197
-Epoch 804/1000
-60000/60000 - 5s - loss: 0.1105 - accuracy: 0.9728 - val_loss: 0.3434 - val_accuracy: 0.9192
-Epoch 805/1000
-60000/60000 - 5s - loss: 0.1104 - accuracy: 0.9728 - val_loss: 0.3438 - val_accuracy: 0.9205
-Epoch 806/1000
-60000/60000 - 5s - loss: 0.1105 - accuracy: 0.9726 - val_loss: 0.3438 - val_accuracy: 0.9196
-Epoch 807/1000
-60000/60000 - 5s - loss: 0.1104 - accuracy: 0.9724 - val_loss: 0.3439 - val_accuracy: 0.9197
-Epoch 808/1000
-60000/60000 - 5s - loss: 0.1104 - accuracy: 0.9729 - val_loss: 0.3438 - val_accuracy: 0.9198
-Epoch 809/1000
-60000/60000 - 5s - loss: 0.1103 - accuracy: 0.9726 - val_loss: 0.3436 - val_accuracy: 0.9196
-Epoch 810/1000
-60000/60000 - 5s - loss: 0.1103 - accuracy: 0.9726 - val_loss: 0.3439 - val_accuracy: 0.9208
-Epoch 811/1000
-60000/60000 - 5s - loss: 0.1103 - accuracy: 0.9730 - val_loss: 0.3439 - val_accuracy: 0.9213
-Epoch 812/1000
-60000/60000 - 5s - loss: 0.1103 - accuracy: 0.9728 - val_loss: 0.3443 - val_accuracy: 0.9204
-Epoch 813/1000
-60000/60000 - 5s - loss: 0.1102 - accuracy: 0.9729 - val_loss: 0.3439 - val_accuracy: 0.9206
-Epoch 814/1000
-60000/60000 - 5s - loss: 0.1101 - accuracy: 0.9730 - val_loss: 0.3443 - val_accuracy: 0.9202
-Epoch 815/1000
-60000/60000 - 5s - loss: 0.1102 - accuracy: 0.9726 - val_loss: 0.3439 - val_accuracy: 0.9205
-Epoch 816/1000
-60000/60000 - 5s - loss: 0.1101 - accuracy: 0.9730 - val_loss: 0.3437 - val_accuracy: 0.9202
-Epoch 817/1000
-60000/60000 - 5s - loss: 0.1101 - accuracy: 0.9729 - val_loss: 0.3445 - val_accuracy: 0.9193
-Epoch 818/1000
-60000/60000 - 5s - loss: 0.1101 - accuracy: 0.9728 - val_loss: 0.3443 - val_accuracy: 0.9199
-Epoch 819/1000
-60000/60000 - 5s - loss: 0.1100 - accuracy: 0.9728 - val_loss: 0.3441 - val_accuracy: 0.9196
-Epoch 820/1000
-60000/60000 - 5s - loss: 0.1100 - accuracy: 0.9728 - val_loss: 0.3440 - val_accuracy: 0.9196
-Epoch 821/1000
-60000/60000 - 5s - loss: 0.1100 - accuracy: 0.9729 - val_loss: 0.3446 - val_accuracy: 0.9197
-Epoch 822/1000
-60000/60000 - 5s - loss: 0.1098 - accuracy: 0.9731 - val_loss: 0.3450 - val_accuracy: 0.9204
-Epoch 823/1000
-60000/60000 - 5s - loss: 0.1098 - accuracy: 0.9729 - val_loss: 0.3444 - val_accuracy: 0.9195
-Epoch 824/1000
-60000/60000 - 5s - loss: 0.1098 - accuracy: 0.9729 - val_loss: 0.3449 - val_accuracy: 0.9195
-Epoch 825/1000
-60000/60000 - 5s - loss: 0.1097 - accuracy: 0.9729 - val_loss: 0.3446 - val_accuracy: 0.9208
-Epoch 826/1000
-60000/60000 - 5s - loss: 0.1098 - accuracy: 0.9730 - val_loss: 0.3447 - val_accuracy: 0.9207
-Epoch 827/1000
-60000/60000 - 5s - loss: 0.1098 - accuracy: 0.9729 - val_loss: 0.3444 - val_accuracy: 0.9205
-Epoch 828/1000
-60000/60000 - 5s - loss: 0.1097 - accuracy: 0.9733 - val_loss: 0.3442 - val_accuracy: 0.9204
-Epoch 829/1000
-60000/60000 - 5s - loss: 0.1097 - accuracy: 0.9728 - val_loss: 0.3450 - val_accuracy: 0.9197
-Epoch 830/1000
-60000/60000 - 5s - loss: 0.1097 - accuracy: 0.9730 - val_loss: 0.3446 - val_accuracy: 0.9200
-Epoch 831/1000
-60000/60000 - 5s - loss: 0.1096 - accuracy: 0.9732 - val_loss: 0.3451 - val_accuracy: 0.9196
-Epoch 832/1000
-60000/60000 - 5s - loss: 0.1095 - accuracy: 0.9730 - val_loss: 0.3450 - val_accuracy: 0.9202
-Epoch 833/1000
-60000/60000 - 5s - loss: 0.1096 - accuracy: 0.9731 - val_loss: 0.3447 - val_accuracy: 0.9204
-Epoch 834/1000
-60000/60000 - 5s - loss: 0.1095 - accuracy: 0.9730 - val_loss: 0.3446 - val_accuracy: 0.9199
-Epoch 835/1000
-60000/60000 - 5s - loss: 0.1094 - accuracy: 0.9731 - val_loss: 0.3452 - val_accuracy: 0.9197
-Epoch 836/1000
-60000/60000 - 5s - loss: 0.1095 - accuracy: 0.9731 - val_loss: 0.3454 - val_accuracy: 0.9196
-Epoch 837/1000
-60000/60000 - 5s - loss: 0.1094 - accuracy: 0.9733 - val_loss: 0.3455 - val_accuracy: 0.9201
-Epoch 838/1000
-60000/60000 - 5s - loss: 0.1093 - accuracy: 0.9731 - val_loss: 0.3454 - val_accuracy: 0.9204
-Epoch 839/1000
-60000/60000 - 5s - loss: 0.1093 - accuracy: 0.9733 - val_loss: 0.3454 - val_accuracy: 0.9201
-Epoch 840/1000
-60000/60000 - 5s - loss: 0.1093 - accuracy: 0.9730 - val_loss: 0.3448 - val_accuracy: 0.9198
-Epoch 841/1000
-60000/60000 - 5s - loss: 0.1092 - accuracy: 0.9731 - val_loss: 0.3452 - val_accuracy: 0.9210
-Epoch 842/1000
-60000/60000 - 5s - loss: 0.1091 - accuracy: 0.9732 - val_loss: 0.3455 - val_accuracy: 0.9198
-Epoch 843/1000
-60000/60000 - 5s - loss: 0.1092 - accuracy: 0.9731 - val_loss: 0.3458 - val_accuracy: 0.9195
-Epoch 844/1000
-60000/60000 - 5s - loss: 0.1092 - accuracy: 0.9732 - val_loss: 0.3453 - val_accuracy: 0.9198
-Epoch 845/1000
-60000/60000 - 5s - loss: 0.1091 - accuracy: 0.9730 - val_loss: 0.3456 - val_accuracy: 0.9201
-Epoch 846/1000
-60000/60000 - 5s - loss: 0.1091 - accuracy: 0.9735 - val_loss: 0.3462 - val_accuracy: 0.9194
-Epoch 847/1000
-60000/60000 - 5s - loss: 0.1091 - accuracy: 0.9732 - val_loss: 0.3457 - val_accuracy: 0.9196
-Epoch 848/1000
-60000/60000 - 5s - loss: 0.1090 - accuracy: 0.9731 - val_loss: 0.3461 - val_accuracy: 0.9200
-Epoch 849/1000
-60000/60000 - 5s - loss: 0.1089 - accuracy: 0.9732 - val_loss: 0.3457 - val_accuracy: 0.9203
-Epoch 850/1000
-60000/60000 - 5s - loss: 0.1089 - accuracy: 0.9734 - val_loss: 0.3462 - val_accuracy: 0.9201
-Epoch 851/1000
-60000/60000 - 5s - loss: 0.1089 - accuracy: 0.9734 - val_loss: 0.3470 - val_accuracy: 0.9204
-Epoch 852/1000
-60000/60000 - 5s - loss: 0.1089 - accuracy: 0.9735 - val_loss: 0.3464 - val_accuracy: 0.9197
-Epoch 853/1000
-60000/60000 - 5s - loss: 0.1088 - accuracy: 0.9732 - val_loss: 0.3455 - val_accuracy: 0.9198
-Epoch 854/1000
-60000/60000 - 5s - loss: 0.1088 - accuracy: 0.9735 - val_loss: 0.3462 - val_accuracy: 0.9204
-Epoch 855/1000
-60000/60000 - 5s - loss: 0.1088 - accuracy: 0.9736 - val_loss: 0.3464 - val_accuracy: 0.9209
-Epoch 856/1000
-60000/60000 - 5s - loss: 0.1087 - accuracy: 0.9734 - val_loss: 0.3467 - val_accuracy: 0.9205
-Epoch 857/1000
-60000/60000 - 5s - loss: 0.1087 - accuracy: 0.9734 - val_loss: 0.3471 - val_accuracy: 0.9207
-Epoch 858/1000
-60000/60000 - 5s - loss: 0.1087 - accuracy: 0.9735 - val_loss: 0.3467 - val_accuracy: 0.9199
-Epoch 859/1000
-60000/60000 - 5s - loss: 0.1087 - accuracy: 0.9733 - val_loss: 0.3469 - val_accuracy: 0.9202
-Epoch 860/1000
-60000/60000 - 5s - loss: 0.1085 - accuracy: 0.9735 - val_loss: 0.3472 - val_accuracy: 0.9205
-Epoch 861/1000
-60000/60000 - 5s - loss: 0.1086 - accuracy: 0.9733 - val_loss: 0.3473 - val_accuracy: 0.9192
-Epoch 862/1000
-60000/60000 - 5s - loss: 0.1085 - accuracy: 0.9736 - val_loss: 0.3468 - val_accuracy: 0.9197
-Epoch 863/1000
-60000/60000 - 5s - loss: 0.1085 - accuracy: 0.9736 - val_loss: 0.3472 - val_accuracy: 0.9199
-Epoch 864/1000
-60000/60000 - 5s - loss: 0.1084 - accuracy: 0.9736 - val_loss: 0.3475 - val_accuracy: 0.9196
-Epoch 865/1000
-60000/60000 - 5s - loss: 0.1084 - accuracy: 0.9736 - val_loss: 0.3472 - val_accuracy: 0.9206
-Epoch 866/1000
-60000/60000 - 5s - loss: 0.1084 - accuracy: 0.9735 - val_loss: 0.3472 - val_accuracy: 0.9200
-Epoch 867/1000
-60000/60000 - 5s - loss: 0.1084 - accuracy: 0.9735 - val_loss: 0.3475 - val_accuracy: 0.9201
-Epoch 868/1000
-60000/60000 - 5s - loss: 0.1083 - accuracy: 0.9736 - val_loss: 0.3475 - val_accuracy: 0.9202
-Epoch 869/1000
-60000/60000 - 5s - loss: 0.1084 - accuracy: 0.9731 - val_loss: 0.3479 - val_accuracy: 0.9205
-Epoch 870/1000
-60000/60000 - 5s - loss: 0.1083 - accuracy: 0.9737 - val_loss: 0.3483 - val_accuracy: 0.9194
-Epoch 871/1000
-60000/60000 - 5s - loss: 0.1083 - accuracy: 0.9733 - val_loss: 0.3475 - val_accuracy: 0.9198
-Epoch 872/1000
-60000/60000 - 5s - loss: 0.1083 - accuracy: 0.9735 - val_loss: 0.3480 - val_accuracy: 0.9197
-Epoch 873/1000
-60000/60000 - 5s - loss: 0.1082 - accuracy: 0.9737 - val_loss: 0.3478 - val_accuracy: 0.9198
-Epoch 874/1000
-60000/60000 - 5s - loss: 0.1082 - accuracy: 0.9735 - val_loss: 0.3484 - val_accuracy: 0.9204
-Epoch 875/1000
-60000/60000 - 5s - loss: 0.1081 - accuracy: 0.9734 - val_loss: 0.3479 - val_accuracy: 0.9210
-Epoch 876/1000
-60000/60000 - 5s - loss: 0.1082 - accuracy: 0.9736 - val_loss: 0.3477 - val_accuracy: 0.9198
-Epoch 877/1000
-60000/60000 - 5s - loss: 0.1081 - accuracy: 0.9737 - val_loss: 0.3478 - val_accuracy: 0.9201
-Epoch 878/1000
-60000/60000 - 5s - loss: 0.1080 - accuracy: 0.9733 - val_loss: 0.3483 - val_accuracy: 0.9202
-Epoch 879/1000
-60000/60000 - 5s - loss: 0.1080 - accuracy: 0.9735 - val_loss: 0.3484 - val_accuracy: 0.9189
-Epoch 880/1000
-60000/60000 - 5s - loss: 0.1080 - accuracy: 0.9734 - val_loss: 0.3483 - val_accuracy: 0.9196
-Epoch 881/1000
-60000/60000 - 5s - loss: 0.1079 - accuracy: 0.9737 - val_loss: 0.3483 - val_accuracy: 0.9190
-Epoch 882/1000
-60000/60000 - 5s - loss: 0.1079 - accuracy: 0.9733 - val_loss: 0.3484 - val_accuracy: 0.9192
-Epoch 883/1000
-60000/60000 - 5s - loss: 0.1080 - accuracy: 0.9738 - val_loss: 0.3485 - val_accuracy: 0.9195
-Epoch 884/1000
-60000/60000 - 5s - loss: 0.1079 - accuracy: 0.9737 - val_loss: 0.3488 - val_accuracy: 0.9195
-Epoch 885/1000
-60000/60000 - 5s - loss: 0.1079 - accuracy: 0.9738 - val_loss: 0.3492 - val_accuracy: 0.9203
-Epoch 886/1000
-60000/60000 - 5s - loss: 0.1078 - accuracy: 0.9736 - val_loss: 0.3489 - val_accuracy: 0.9202
-Epoch 887/1000
-60000/60000 - 5s - loss: 0.1078 - accuracy: 0.9737 - val_loss: 0.3486 - val_accuracy: 0.9197
-Epoch 888/1000
-60000/60000 - 5s - loss: 0.1077 - accuracy: 0.9739 - val_loss: 0.3489 - val_accuracy: 0.9193
-Epoch 889/1000
-60000/60000 - 5s - loss: 0.1077 - accuracy: 0.9738 - val_loss: 0.3490 - val_accuracy: 0.9199
-Epoch 890/1000
-60000/60000 - 5s - loss: 0.1078 - accuracy: 0.9736 - val_loss: 0.3488 - val_accuracy: 0.9194
-Epoch 891/1000
-60000/60000 - 5s - loss: 0.1077 - accuracy: 0.9736 - val_loss: 0.3488 - val_accuracy: 0.9196
-Epoch 892/1000
-60000/60000 - 5s - loss: 0.1077 - accuracy: 0.9737 - val_loss: 0.3490 - val_accuracy: 0.9200
-Epoch 893/1000
-60000/60000 - 5s - loss: 0.1076 - accuracy: 0.9738 - val_loss: 0.3489 - val_accuracy: 0.9193
-Epoch 894/1000
-60000/60000 - 5s - loss: 0.1076 - accuracy: 0.9738 - val_loss: 0.3492 - val_accuracy: 0.9193
-Epoch 895/1000
-60000/60000 - 5s - loss: 0.1075 - accuracy: 0.9736 - val_loss: 0.3494 - val_accuracy: 0.9199
-Epoch 896/1000
-60000/60000 - 5s - loss: 0.1076 - accuracy: 0.9737 - val_loss: 0.3491 - val_accuracy: 0.9198
-Epoch 897/1000
-60000/60000 - 5s - loss: 0.1075 - accuracy: 0.9737 - val_loss: 0.3494 - val_accuracy: 0.9195
-Epoch 898/1000
-60000/60000 - 5s - loss: 0.1075 - accuracy: 0.9736 - val_loss: 0.3499 - val_accuracy: 0.9192
-Epoch 899/1000
-60000/60000 - 5s - loss: 0.1074 - accuracy: 0.9737 - val_loss: 0.3493 - val_accuracy: 0.9195
-Epoch 900/1000
-60000/60000 - 5s - loss: 0.1074 - accuracy: 0.9740 - val_loss: 0.3498 - val_accuracy: 0.9191
-Epoch 901/1000
-60000/60000 - 5s - loss: 0.1074 - accuracy: 0.9737 - val_loss: 0.3498 - val_accuracy: 0.9196
-Epoch 902/1000
-60000/60000 - 5s - loss: 0.1074 - accuracy: 0.9738 - val_loss: 0.3501 - val_accuracy: 0.9195
-Epoch 903/1000
-60000/60000 - 5s - loss: 0.1074 - accuracy: 0.9739 - val_loss: 0.3500 - val_accuracy: 0.9193
-Epoch 904/1000
-60000/60000 - 5s - loss: 0.1074 - accuracy: 0.9737 - val_loss: 0.3498 - val_accuracy: 0.9196
-Epoch 905/1000
-60000/60000 - 5s - loss: 0.1073 - accuracy: 0.9736 - val_loss: 0.3500 - val_accuracy: 0.9203
-Epoch 906/1000
-60000/60000 - 5s - loss: 0.1073 - accuracy: 0.9737 - val_loss: 0.3504 - val_accuracy: 0.9198
-Epoch 907/1000
-60000/60000 - 5s - loss: 0.1072 - accuracy: 0.9740 - val_loss: 0.3507 - val_accuracy: 0.9203
-Epoch 908/1000
-60000/60000 - 5s - loss: 0.1073 - accuracy: 0.9739 - val_loss: 0.3507 - val_accuracy: 0.9196
-Epoch 909/1000
-60000/60000 - 5s - loss: 0.1073 - accuracy: 0.9737 - val_loss: 0.3499 - val_accuracy: 0.9198
-Epoch 910/1000
-60000/60000 - 5s - loss: 0.1072 - accuracy: 0.9739 - val_loss: 0.3507 - val_accuracy: 0.9197
-Epoch 911/1000
-60000/60000 - 5s - loss: 0.1072 - accuracy: 0.9738 - val_loss: 0.3505 - val_accuracy: 0.9189
-Epoch 912/1000
-60000/60000 - 5s - loss: 0.1072 - accuracy: 0.9739 - val_loss: 0.3507 - val_accuracy: 0.9197
-Epoch 913/1000
-60000/60000 - 5s - loss: 0.1071 - accuracy: 0.9737 - val_loss: 0.3502 - val_accuracy: 0.9196
-Epoch 914/1000
-60000/60000 - 5s - loss: 0.1071 - accuracy: 0.9737 - val_loss: 0.3506 - val_accuracy: 0.9195
-Epoch 915/1000
-60000/60000 - 5s - loss: 0.1071 - accuracy: 0.9736 - val_loss: 0.3511 - val_accuracy: 0.9199
-Epoch 916/1000
-60000/60000 - 5s - loss: 0.1070 - accuracy: 0.9739 - val_loss: 0.3504 - val_accuracy: 0.9192
-Epoch 917/1000
-60000/60000 - 5s - loss: 0.1071 - accuracy: 0.9738 - val_loss: 0.3509 - val_accuracy: 0.9199
-Epoch 918/1000
-60000/60000 - 5s - loss: 0.1070 - accuracy: 0.9740 - val_loss: 0.3509 - val_accuracy: 0.9189
-Epoch 919/1000
-60000/60000 - 5s - loss: 0.1070 - accuracy: 0.9740 - val_loss: 0.3506 - val_accuracy: 0.9193
-Epoch 920/1000
-60000/60000 - 5s - loss: 0.1069 - accuracy: 0.9742 - val_loss: 0.3506 - val_accuracy: 0.9201
-Epoch 921/1000
-60000/60000 - 5s - loss: 0.1069 - accuracy: 0.9740 - val_loss: 0.3512 - val_accuracy: 0.9191
-Epoch 922/1000
-60000/60000 - 5s - loss: 0.1069 - accuracy: 0.9739 - val_loss: 0.3515 - val_accuracy: 0.9196
-Epoch 923/1000
-60000/60000 - 5s - loss: 0.1069 - accuracy: 0.9739 - val_loss: 0.3505 - val_accuracy: 0.9197
-Epoch 924/1000
-60000/60000 - 5s - loss: 0.1068 - accuracy: 0.9741 - val_loss: 0.3515 - val_accuracy: 0.9189
-Epoch 925/1000
-60000/60000 - 5s - loss: 0.1068 - accuracy: 0.9740 - val_loss: 0.3516 - val_accuracy: 0.9187
-Epoch 926/1000
-60000/60000 - 5s - loss: 0.1068 - accuracy: 0.9737 - val_loss: 0.3508 - val_accuracy: 0.9192
-Epoch 927/1000
-60000/60000 - 5s - loss: 0.1068 - accuracy: 0.9739 - val_loss: 0.3518 - val_accuracy: 0.9195
-Epoch 928/1000
-60000/60000 - 5s - loss: 0.1067 - accuracy: 0.9742 - val_loss: 0.3518 - val_accuracy: 0.9190
-Epoch 929/1000
-60000/60000 - 5s - loss: 0.1067 - accuracy: 0.9739 - val_loss: 0.3515 - val_accuracy: 0.9190
-Epoch 930/1000
-60000/60000 - 5s - loss: 0.1067 - accuracy: 0.9740 - val_loss: 0.3514 - val_accuracy: 0.9196
-Epoch 931/1000
-60000/60000 - 5s - loss: 0.1067 - accuracy: 0.9741 - val_loss: 0.3517 - val_accuracy: 0.9196
-Epoch 932/1000
-60000/60000 - 5s - loss: 0.1067 - accuracy: 0.9740 - val_loss: 0.3520 - val_accuracy: 0.9191
-Epoch 933/1000
-60000/60000 - 5s - loss: 0.1066 - accuracy: 0.9742 - val_loss: 0.3520 - val_accuracy: 0.9194
-Epoch 934/1000
-60000/60000 - 5s - loss: 0.1066 - accuracy: 0.9741 - val_loss: 0.3518 - val_accuracy: 0.9186
-Epoch 935/1000
-60000/60000 - 5s - loss: 0.1065 - accuracy: 0.9741 - val_loss: 0.3522 - val_accuracy: 0.9199
-Epoch 936/1000
-60000/60000 - 5s - loss: 0.1066 - accuracy: 0.9739 - val_loss: 0.3519 - val_accuracy: 0.9189
-Epoch 937/1000
-60000/60000 - 5s - loss: 0.1066 - accuracy: 0.9743 - val_loss: 0.3525 - val_accuracy: 0.9190
-Epoch 938/1000
-60000/60000 - 5s - loss: 0.1065 - accuracy: 0.9742 - val_loss: 0.3522 - val_accuracy: 0.9193
-Epoch 939/1000
-60000/60000 - 5s - loss: 0.1064 - accuracy: 0.9740 - val_loss: 0.3518 - val_accuracy: 0.9197
-Epoch 940/1000
-60000/60000 - 5s - loss: 0.1065 - accuracy: 0.9741 - val_loss: 0.3520 - val_accuracy: 0.9193
-Epoch 941/1000
-60000/60000 - 5s - loss: 0.1064 - accuracy: 0.9740 - val_loss: 0.3519 - val_accuracy: 0.9196
-Epoch 942/1000
-60000/60000 - 5s - loss: 0.1065 - accuracy: 0.9741 - val_loss: 0.3526 - val_accuracy: 0.9193
-Epoch 943/1000
-60000/60000 - 5s - loss: 0.1064 - accuracy: 0.9741 - val_loss: 0.3519 - val_accuracy: 0.9190
-Epoch 944/1000
-60000/60000 - 5s - loss: 0.1063 - accuracy: 0.9741 - val_loss: 0.3527 - val_accuracy: 0.9191
-Epoch 945/1000
-60000/60000 - 5s - loss: 0.1064 - accuracy: 0.9741 - val_loss: 0.3525 - val_accuracy: 0.9191
-Epoch 946/1000
-60000/60000 - 5s - loss: 0.1063 - accuracy: 0.9740 - val_loss: 0.3532 - val_accuracy: 0.9185
-Epoch 947/1000
-60000/60000 - 5s - loss: 0.1063 - accuracy: 0.9741 - val_loss: 0.3522 - val_accuracy: 0.9196
-Epoch 948/1000
-60000/60000 - 5s - loss: 0.1063 - accuracy: 0.9741 - val_loss: 0.3529 - val_accuracy: 0.9186
-Epoch 949/1000
-60000/60000 - 5s - loss: 0.1063 - accuracy: 0.9742 - val_loss: 0.3527 - val_accuracy: 0.9186
-Epoch 950/1000
-60000/60000 - 5s - loss: 0.1062 - accuracy: 0.9741 - val_loss: 0.3531 - val_accuracy: 0.9190
-Epoch 951/1000
-60000/60000 - 5s - loss: 0.1062 - accuracy: 0.9742 - val_loss: 0.3533 - val_accuracy: 0.9189
-Epoch 952/1000
-60000/60000 - 5s - loss: 0.1062 - accuracy: 0.9743 - val_loss: 0.3529 - val_accuracy: 0.9186
-Epoch 953/1000
-60000/60000 - 5s - loss: 0.1062 - accuracy: 0.9743 - val_loss: 0.3530 - val_accuracy: 0.9190
-Epoch 954/1000
-60000/60000 - 5s - loss: 0.1062 - accuracy: 0.9740 - val_loss: 0.3530 - val_accuracy: 0.9189
-Epoch 955/1000
-60000/60000 - 5s - loss: 0.1061 - accuracy: 0.9742 - val_loss: 0.3535 - val_accuracy: 0.9187
-Epoch 956/1000
-60000/60000 - 5s - loss: 0.1061 - accuracy: 0.9741 - val_loss: 0.3533 - val_accuracy: 0.9190
-Epoch 957/1000
-60000/60000 - 5s - loss: 0.1061 - accuracy: 0.9741 - val_loss: 0.3534 - val_accuracy: 0.9194
-Epoch 958/1000
-60000/60000 - 5s - loss: 0.1060 - accuracy: 0.9742 - val_loss: 0.3531 - val_accuracy: 0.9195
-Epoch 959/1000
-60000/60000 - 5s - loss: 0.1060 - accuracy: 0.9741 - val_loss: 0.3532 - val_accuracy: 0.9192
-Epoch 960/1000
-60000/60000 - 5s - loss: 0.1061 - accuracy: 0.9741 - val_loss: 0.3534 - val_accuracy: 0.9188
-Epoch 961/1000
-60000/60000 - 5s - loss: 0.1059 - accuracy: 0.9743 - val_loss: 0.3538 - val_accuracy: 0.9193
-Epoch 962/1000
-60000/60000 - 5s - loss: 0.1060 - accuracy: 0.9742 - val_loss: 0.3540 - val_accuracy: 0.9190
-Epoch 963/1000
-60000/60000 - 5s - loss: 0.1059 - accuracy: 0.9743 - val_loss: 0.3541 - val_accuracy: 0.9192
-Epoch 964/1000
-60000/60000 - 5s - loss: 0.1059 - accuracy: 0.9742 - val_loss: 0.3537 - val_accuracy: 0.9189
-Epoch 965/1000
-60000/60000 - 5s - loss: 0.1059 - accuracy: 0.9743 - val_loss: 0.3539 - val_accuracy: 0.9189
-Epoch 966/1000
-60000/60000 - 5s - loss: 0.1059 - accuracy: 0.9742 - val_loss: 0.3537 - val_accuracy: 0.9183
-Epoch 967/1000
-60000/60000 - 5s - loss: 0.1059 - accuracy: 0.9741 - val_loss: 0.3538 - val_accuracy: 0.9186
-Epoch 968/1000
-60000/60000 - 5s - loss: 0.1058 - accuracy: 0.9740 - val_loss: 0.3539 - val_accuracy: 0.9187
-Epoch 969/1000
-60000/60000 - 5s - loss: 0.1058 - accuracy: 0.9743 - val_loss: 0.3542 - val_accuracy: 0.9187
-Epoch 970/1000
-60000/60000 - 5s - loss: 0.1058 - accuracy: 0.9742 - val_loss: 0.3540 - val_accuracy: 0.9193
-Epoch 971/1000
-60000/60000 - 5s - loss: 0.1058 - accuracy: 0.9742 - val_loss: 0.3540 - val_accuracy: 0.9189
-Epoch 972/1000
-60000/60000 - 5s - loss: 0.1058 - accuracy: 0.9742 - val_loss: 0.3539 - val_accuracy: 0.9185
-Epoch 973/1000
-60000/60000 - 5s - loss: 0.1057 - accuracy: 0.9744 - val_loss: 0.3542 - val_accuracy: 0.9185
-Epoch 974/1000
-60000/60000 - 5s - loss: 0.1057 - accuracy: 0.9743 - val_loss: 0.3544 - val_accuracy: 0.9196
-Epoch 975/1000
-60000/60000 - 5s - loss: 0.1056 - accuracy: 0.9744 - val_loss: 0.3540 - val_accuracy: 0.9186
-Epoch 976/1000
-60000/60000 - 5s - loss: 0.1056 - accuracy: 0.9741 - val_loss: 0.3544 - val_accuracy: 0.9187
-Epoch 977/1000
-60000/60000 - 5s - loss: 0.1057 - accuracy: 0.9741 - val_loss: 0.3544 - val_accuracy: 0.9186
-Epoch 978/1000
-60000/60000 - 5s - loss: 0.1057 - accuracy: 0.9744 - val_loss: 0.3543 - val_accuracy: 0.9186
-Epoch 979/1000
-60000/60000 - 5s - loss: 0.1056 - accuracy: 0.9743 - val_loss: 0.3544 - val_accuracy: 0.9188
-Epoch 980/1000
-60000/60000 - 5s - loss: 0.1055 - accuracy: 0.9743 - val_loss: 0.3545 - val_accuracy: 0.9194
-Epoch 981/1000
-60000/60000 - 5s - loss: 0.1056 - accuracy: 0.9745 - val_loss: 0.3543 - val_accuracy: 0.9183
-Epoch 982/1000
-60000/60000 - 5s - loss: 0.1056 - accuracy: 0.9742 - val_loss: 0.3548 - val_accuracy: 0.9186
-Epoch 983/1000
-60000/60000 - 5s - loss: 0.1055 - accuracy: 0.9741 - val_loss: 0.3546 - val_accuracy: 0.9182
-Epoch 984/1000
-60000/60000 - 5s - loss: 0.1055 - accuracy: 0.9744 - val_loss: 0.3555 - val_accuracy: 0.9184
-Epoch 985/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9744 - val_loss: 0.3539 - val_accuracy: 0.9184
-Epoch 986/1000
-60000/60000 - 5s - loss: 0.1055 - accuracy: 0.9742 - val_loss: 0.3550 - val_accuracy: 0.9188
-Epoch 987/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9745 - val_loss: 0.3545 - val_accuracy: 0.9190
-Epoch 988/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9743 - val_loss: 0.3547 - val_accuracy: 0.9182
-Epoch 989/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9743 - val_loss: 0.3552 - val_accuracy: 0.9186
-Epoch 990/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9743 - val_loss: 0.3551 - val_accuracy: 0.9182
-Epoch 991/1000
-60000/60000 - 5s - loss: 0.1053 - accuracy: 0.9742 - val_loss: 0.3552 - val_accuracy: 0.9186
-Epoch 992/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9744 - val_loss: 0.3552 - val_accuracy: 0.9185
-Epoch 993/1000
-60000/60000 - 5s - loss: 0.1053 - accuracy: 0.9746 - val_loss: 0.3550 - val_accuracy: 0.9183
-Epoch 994/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9744 - val_loss: 0.3557 - val_accuracy: 0.9185
-Epoch 995/1000
-60000/60000 - 5s - loss: 0.1054 - accuracy: 0.9745 - val_loss: 0.3552 - val_accuracy: 0.9184
-Epoch 996/1000
-60000/60000 - 5s - loss: 0.1053 - accuracy: 0.9745 - val_loss: 0.3548 - val_accuracy: 0.9183
-Epoch 997/1000
-60000/60000 - 5s - loss: 0.1053 - accuracy: 0.9743 - val_loss: 0.3552 - val_accuracy: 0.9186
-Epoch 998/1000
-60000/60000 - 5s - loss: 0.1052 - accuracy: 0.9746 - val_loss: 0.3557 - val_accuracy: 0.9185
-Epoch 999/1000
-60000/60000 - 5s - loss: 0.1053 - accuracy: 0.9743 - val_loss: 0.3553 - val_accuracy: 0.9181
-Epoch 1000/1000
-60000/60000 - 5s - loss: 0.1051 - accuracy: 0.9744 - val_loss: 0.3556 - val_accuracy: 0.9189
-Test loss was 0.2972, test accuracy was 0.9239
diff --git a/MNIST/nonlinear_withsigmoid.txt b/MNIST/nonlinear_withsigmoid.txt
deleted file mode 100644
index efa6d8c..0000000
--- a/MNIST/nonlinear_withsigmoid.txt
+++ /dev/null
@@ -1,2005 +0,0 @@
-X_train shape: (60000, 784)
-60000 train samples
-10000 test samples
-Train on 60000 samples, validate on 10000 samples
-Epoch 1/1000
-60000/60000 - 9s - loss: 0.6632 - accuracy: 0.8073 - val_loss: 0.4125 - val_accuracy: 0.8811
-Epoch 2/1000
-60000/60000 - 6s - loss: 0.3864 - accuracy: 0.8889 - val_loss: 0.3437 - val_accuracy: 0.9008
-Epoch 3/1000
-60000/60000 - 6s - loss: 0.3298 - accuracy: 0.9053 - val_loss: 0.3090 - val_accuracy: 0.9077
-Epoch 4/1000
-60000/60000 - 6s - loss: 0.2969 - accuracy: 0.9142 - val_loss: 0.2858 - val_accuracy: 0.9154
-Epoch 5/1000
-60000/60000 - 6s - loss: 0.2737 - accuracy: 0.9216 - val_loss: 0.2694 - val_accuracy: 0.9222
-Epoch 6/1000
-60000/60000 - 6s - loss: 0.2557 - accuracy: 0.9260 - val_loss: 0.2568 - val_accuracy: 0.9241
-Epoch 7/1000
-60000/60000 - 6s - loss: 0.2413 - accuracy: 0.9311 - val_loss: 0.2480 - val_accuracy: 0.9266
-Epoch 8/1000
-60000/60000 - 6s - loss: 0.2296 - accuracy: 0.9346 - val_loss: 0.2381 - val_accuracy: 0.9299
-Epoch 9/1000
-60000/60000 - 6s - loss: 0.2197 - accuracy: 0.9376 - val_loss: 0.2308 - val_accuracy: 0.9308
-Epoch 10/1000
-60000/60000 - 6s - loss: 0.2108 - accuracy: 0.9400 - val_loss: 0.2269 - val_accuracy: 0.9312
-Epoch 11/1000
-60000/60000 - 6s - loss: 0.2033 - accuracy: 0.9421 - val_loss: 0.2207 - val_accuracy: 0.9347
-Epoch 12/1000
-60000/60000 - 6s - loss: 0.1969 - accuracy: 0.9438 - val_loss: 0.2177 - val_accuracy: 0.9346
-Epoch 13/1000
-60000/60000 - 6s - loss: 0.1906 - accuracy: 0.9456 - val_loss: 0.2138 - val_accuracy: 0.9362
-Epoch 14/1000
-60000/60000 - 6s - loss: 0.1852 - accuracy: 0.9470 - val_loss: 0.2101 - val_accuracy: 0.9372
-Epoch 15/1000
-60000/60000 - 6s - loss: 0.1803 - accuracy: 0.9488 - val_loss: 0.2087 - val_accuracy: 0.9374
-Epoch 16/1000
-60000/60000 - 6s - loss: 0.1757 - accuracy: 0.9497 - val_loss: 0.2045 - val_accuracy: 0.9381
-Epoch 17/1000
-60000/60000 - 6s - loss: 0.1715 - accuracy: 0.9506 - val_loss: 0.2035 - val_accuracy: 0.9382
-Epoch 18/1000
-60000/60000 - 6s - loss: 0.1676 - accuracy: 0.9518 - val_loss: 0.2010 - val_accuracy: 0.9388
-Epoch 19/1000
-60000/60000 - 6s - loss: 0.1640 - accuracy: 0.9525 - val_loss: 0.1997 - val_accuracy: 0.9392
-Epoch 20/1000
-60000/60000 - 6s - loss: 0.1605 - accuracy: 0.9537 - val_loss: 0.1985 - val_accuracy: 0.9395
-Epoch 21/1000
-60000/60000 - 6s - loss: 0.1570 - accuracy: 0.9545 - val_loss: 0.1982 - val_accuracy: 0.9389
-Epoch 22/1000
-60000/60000 - 6s - loss: 0.1540 - accuracy: 0.9561 - val_loss: 0.1968 - val_accuracy: 0.9399
-Epoch 23/1000
-60000/60000 - 6s - loss: 0.1511 - accuracy: 0.9567 - val_loss: 0.1947 - val_accuracy: 0.9406
-Epoch 24/1000
-60000/60000 - 6s - loss: 0.1481 - accuracy: 0.9573 - val_loss: 0.1945 - val_accuracy: 0.9405
-Epoch 25/1000
-60000/60000 - 6s - loss: 0.1458 - accuracy: 0.9582 - val_loss: 0.1936 - val_accuracy: 0.9395
-Epoch 26/1000
-60000/60000 - 6s - loss: 0.1430 - accuracy: 0.9591 - val_loss: 0.1927 - val_accuracy: 0.9405
-Epoch 27/1000
-60000/60000 - 6s - loss: 0.1404 - accuracy: 0.9596 - val_loss: 0.1909 - val_accuracy: 0.9411
-Epoch 28/1000
-60000/60000 - 6s - loss: 0.1383 - accuracy: 0.9605 - val_loss: 0.1917 - val_accuracy: 0.9418
-Epoch 29/1000
-60000/60000 - 6s - loss: 0.1362 - accuracy: 0.9608 - val_loss: 0.1896 - val_accuracy: 0.9428
-Epoch 30/1000
-60000/60000 - 6s - loss: 0.1340 - accuracy: 0.9614 - val_loss: 0.1889 - val_accuracy: 0.9427
-Epoch 31/1000
-60000/60000 - 6s - loss: 0.1319 - accuracy: 0.9618 - val_loss: 0.1891 - val_accuracy: 0.9411
-Epoch 32/1000
-60000/60000 - 6s - loss: 0.1297 - accuracy: 0.9632 - val_loss: 0.1893 - val_accuracy: 0.9418
-Epoch 33/1000
-60000/60000 - 6s - loss: 0.1280 - accuracy: 0.9632 - val_loss: 0.1895 - val_accuracy: 0.9428
-Epoch 34/1000
-60000/60000 - 6s - loss: 0.1263 - accuracy: 0.9634 - val_loss: 0.1885 - val_accuracy: 0.9424
-Epoch 35/1000
-60000/60000 - 6s - loss: 0.1243 - accuracy: 0.9646 - val_loss: 0.1885 - val_accuracy: 0.9426
-Epoch 36/1000
-60000/60000 - 6s - loss: 0.1224 - accuracy: 0.9654 - val_loss: 0.1885 - val_accuracy: 0.9425
-Epoch 37/1000
-60000/60000 - 6s - loss: 0.1207 - accuracy: 0.9659 - val_loss: 0.1901 - val_accuracy: 0.9427
-Epoch 38/1000
-60000/60000 - 6s - loss: 0.1191 - accuracy: 0.9662 - val_loss: 0.1872 - val_accuracy: 0.9429
-Epoch 39/1000
-60000/60000 - 6s - loss: 0.1174 - accuracy: 0.9665 - val_loss: 0.1893 - val_accuracy: 0.9418
-Epoch 40/1000
-60000/60000 - 6s - loss: 0.1156 - accuracy: 0.9676 - val_loss: 0.1864 - val_accuracy: 0.9450
-Epoch 41/1000
-60000/60000 - 6s - loss: 0.1145 - accuracy: 0.9675 - val_loss: 0.1871 - val_accuracy: 0.9441
-Epoch 42/1000
-60000/60000 - 6s - loss: 0.1130 - accuracy: 0.9683 - val_loss: 0.1860 - val_accuracy: 0.9441
-Epoch 43/1000
-60000/60000 - 6s - loss: 0.1115 - accuracy: 0.9682 - val_loss: 0.1863 - val_accuracy: 0.9434
-Epoch 44/1000
-60000/60000 - 6s - loss: 0.1104 - accuracy: 0.9689 - val_loss: 0.1870 - val_accuracy: 0.9439
-Epoch 45/1000
-60000/60000 - 6s - loss: 0.1088 - accuracy: 0.9695 - val_loss: 0.1859 - val_accuracy: 0.9449
-Epoch 46/1000
-60000/60000 - 6s - loss: 0.1074 - accuracy: 0.9703 - val_loss: 0.1855 - val_accuracy: 0.9448
-Epoch 47/1000
-60000/60000 - 6s - loss: 0.1063 - accuracy: 0.9703 - val_loss: 0.1842 - val_accuracy: 0.9455
-Epoch 48/1000
-60000/60000 - 6s - loss: 0.1049 - accuracy: 0.9708 - val_loss: 0.1842 - val_accuracy: 0.9461
-Epoch 49/1000
-60000/60000 - 6s - loss: 0.1037 - accuracy: 0.9712 - val_loss: 0.1847 - val_accuracy: 0.9451
-Epoch 50/1000
-60000/60000 - 6s - loss: 0.1026 - accuracy: 0.9713 - val_loss: 0.1845 - val_accuracy: 0.9455
-Epoch 51/1000
-60000/60000 - 6s - loss: 0.1014 - accuracy: 0.9719 - val_loss: 0.1840 - val_accuracy: 0.9460
-Epoch 52/1000
-60000/60000 - 6s - loss: 0.1004 - accuracy: 0.9719 - val_loss: 0.1856 - val_accuracy: 0.9459
-Epoch 53/1000
-60000/60000 - 6s - loss: 0.0991 - accuracy: 0.9726 - val_loss: 0.1858 - val_accuracy: 0.9443
-Epoch 54/1000
-60000/60000 - 6s - loss: 0.0980 - accuracy: 0.9730 - val_loss: 0.1854 - val_accuracy: 0.9465
-Epoch 55/1000
-60000/60000 - 6s - loss: 0.0971 - accuracy: 0.9733 - val_loss: 0.1841 - val_accuracy: 0.9459
-Epoch 56/1000
-60000/60000 - 6s - loss: 0.0960 - accuracy: 0.9736 - val_loss: 0.1860 - val_accuracy: 0.9453
-Epoch 57/1000
-60000/60000 - 6s - loss: 0.0950 - accuracy: 0.9741 - val_loss: 0.1850 - val_accuracy: 0.9451
-Epoch 58/1000
-60000/60000 - 6s - loss: 0.0940 - accuracy: 0.9744 - val_loss: 0.1839 - val_accuracy: 0.9454
-Epoch 59/1000
-60000/60000 - 6s - loss: 0.0930 - accuracy: 0.9745 - val_loss: 0.1839 - val_accuracy: 0.9456
-Epoch 60/1000
-60000/60000 - 6s - loss: 0.0921 - accuracy: 0.9747 - val_loss: 0.1856 - val_accuracy: 0.9462
-Epoch 61/1000
-60000/60000 - 6s - loss: 0.0912 - accuracy: 0.9750 - val_loss: 0.1836 - val_accuracy: 0.9459
-Epoch 62/1000
-60000/60000 - 6s - loss: 0.0902 - accuracy: 0.9755 - val_loss: 0.1851 - val_accuracy: 0.9466
-Epoch 63/1000
-60000/60000 - 6s - loss: 0.0893 - accuracy: 0.9759 - val_loss: 0.1838 - val_accuracy: 0.9458
-Epoch 64/1000
-60000/60000 - 6s - loss: 0.0883 - accuracy: 0.9755 - val_loss: 0.1850 - val_accuracy: 0.9455
-Epoch 65/1000
-60000/60000 - 6s - loss: 0.0878 - accuracy: 0.9766 - val_loss: 0.1849 - val_accuracy: 0.9455
-Epoch 66/1000
-60000/60000 - 6s - loss: 0.0868 - accuracy: 0.9769 - val_loss: 0.1843 - val_accuracy: 0.9460
-Epoch 67/1000
-60000/60000 - 6s - loss: 0.0860 - accuracy: 0.9766 - val_loss: 0.1848 - val_accuracy: 0.9458
-Epoch 68/1000
-60000/60000 - 6s - loss: 0.0852 - accuracy: 0.9772 - val_loss: 0.1855 - val_accuracy: 0.9464
-Epoch 69/1000
-60000/60000 - 6s - loss: 0.0842 - accuracy: 0.9773 - val_loss: 0.1871 - val_accuracy: 0.9439
-Epoch 70/1000
-60000/60000 - 6s - loss: 0.0835 - accuracy: 0.9774 - val_loss: 0.1877 - val_accuracy: 0.9448
-Epoch 71/1000
-60000/60000 - 6s - loss: 0.0827 - accuracy: 0.9778 - val_loss: 0.1849 - val_accuracy: 0.9466
-Epoch 72/1000
-60000/60000 - 6s - loss: 0.0821 - accuracy: 0.9778 - val_loss: 0.1845 - val_accuracy: 0.9465
-Epoch 73/1000
-60000/60000 - 6s - loss: 0.0815 - accuracy: 0.9782 - val_loss: 0.1866 - val_accuracy: 0.9454
-Epoch 74/1000
-60000/60000 - 6s - loss: 0.0806 - accuracy: 0.9784 - val_loss: 0.1872 - val_accuracy: 0.9453
-Epoch 75/1000
-60000/60000 - 6s - loss: 0.0797 - accuracy: 0.9786 - val_loss: 0.1858 - val_accuracy: 0.9459
-Epoch 76/1000
-60000/60000 - 6s - loss: 0.0792 - accuracy: 0.9788 - val_loss: 0.1861 - val_accuracy: 0.9460
-Epoch 77/1000
-60000/60000 - 6s - loss: 0.0784 - accuracy: 0.9789 - val_loss: 0.1896 - val_accuracy: 0.9451
-Epoch 78/1000
-60000/60000 - 6s - loss: 0.0779 - accuracy: 0.9795 - val_loss: 0.1876 - val_accuracy: 0.9461
-Epoch 79/1000
-60000/60000 - 6s - loss: 0.0771 - accuracy: 0.9797 - val_loss: 0.1884 - val_accuracy: 0.9458
-Epoch 80/1000
-60000/60000 - 6s - loss: 0.0766 - accuracy: 0.9797 - val_loss: 0.1877 - val_accuracy: 0.9458
-Epoch 81/1000
-60000/60000 - 6s - loss: 0.0759 - accuracy: 0.9798 - val_loss: 0.1873 - val_accuracy: 0.9468
-Epoch 82/1000
-60000/60000 - 6s - loss: 0.0751 - accuracy: 0.9804 - val_loss: 0.1892 - val_accuracy: 0.9458
-Epoch 83/1000
-60000/60000 - 6s - loss: 0.0745 - accuracy: 0.9803 - val_loss: 0.1885 - val_accuracy: 0.9461
-Epoch 84/1000
-60000/60000 - 6s - loss: 0.0740 - accuracy: 0.9811 - val_loss: 0.1882 - val_accuracy: 0.9466
-Epoch 85/1000
-60000/60000 - 6s - loss: 0.0732 - accuracy: 0.9812 - val_loss: 0.1876 - val_accuracy: 0.9471
-Epoch 86/1000
-60000/60000 - 6s - loss: 0.0726 - accuracy: 0.9814 - val_loss: 0.1899 - val_accuracy: 0.9458
-Epoch 87/1000
-60000/60000 - 6s - loss: 0.0720 - accuracy: 0.9812 - val_loss: 0.1908 - val_accuracy: 0.9462
-Epoch 88/1000
-60000/60000 - 6s - loss: 0.0715 - accuracy: 0.9815 - val_loss: 0.1909 - val_accuracy: 0.9452
-Epoch 89/1000
-60000/60000 - 6s - loss: 0.0710 - accuracy: 0.9816 - val_loss: 0.1906 - val_accuracy: 0.9463
-Epoch 90/1000
-60000/60000 - 6s - loss: 0.0704 - accuracy: 0.9820 - val_loss: 0.1909 - val_accuracy: 0.9465
-Epoch 91/1000
-60000/60000 - 6s - loss: 0.0697 - accuracy: 0.9818 - val_loss: 0.1896 - val_accuracy: 0.9457
-Epoch 92/1000
-60000/60000 - 6s - loss: 0.0694 - accuracy: 0.9819 - val_loss: 0.1915 - val_accuracy: 0.9469
-Epoch 93/1000
-60000/60000 - 6s - loss: 0.0686 - accuracy: 0.9824 - val_loss: 0.1933 - val_accuracy: 0.9454
-Epoch 94/1000
-60000/60000 - 6s - loss: 0.0681 - accuracy: 0.9828 - val_loss: 0.1926 - val_accuracy: 0.9459
-Epoch 95/1000
-60000/60000 - 6s - loss: 0.0677 - accuracy: 0.9825 - val_loss: 0.1926 - val_accuracy: 0.9457
-Epoch 96/1000
-60000/60000 - 6s - loss: 0.0672 - accuracy: 0.9826 - val_loss: 0.1923 - val_accuracy: 0.9462
-Epoch 97/1000
-60000/60000 - 6s - loss: 0.0667 - accuracy: 0.9831 - val_loss: 0.1924 - val_accuracy: 0.9457
-Epoch 98/1000
-60000/60000 - 6s - loss: 0.0661 - accuracy: 0.9833 - val_loss: 0.1951 - val_accuracy: 0.9452
-Epoch 99/1000
-60000/60000 - 6s - loss: 0.0657 - accuracy: 0.9833 - val_loss: 0.1931 - val_accuracy: 0.9467
-Epoch 100/1000
-60000/60000 - 6s - loss: 0.0651 - accuracy: 0.9833 - val_loss: 0.1951 - val_accuracy: 0.9468
-Epoch 101/1000
-60000/60000 - 6s - loss: 0.0645 - accuracy: 0.9835 - val_loss: 0.1939 - val_accuracy: 0.9457
-Epoch 102/1000
-60000/60000 - 6s - loss: 0.0641 - accuracy: 0.9837 - val_loss: 0.1962 - val_accuracy: 0.9462
-Epoch 103/1000
-60000/60000 - 6s - loss: 0.0637 - accuracy: 0.9834 - val_loss: 0.1946 - val_accuracy: 0.9461
-Epoch 104/1000
-60000/60000 - 6s - loss: 0.0633 - accuracy: 0.9839 - val_loss: 0.1962 - val_accuracy: 0.9463
-Epoch 105/1000
-60000/60000 - 6s - loss: 0.0628 - accuracy: 0.9842 - val_loss: 0.1972 - val_accuracy: 0.9472
-Epoch 106/1000
-60000/60000 - 6s - loss: 0.0624 - accuracy: 0.9843 - val_loss: 0.1986 - val_accuracy: 0.9454
-Epoch 107/1000
-60000/60000 - 6s - loss: 0.0618 - accuracy: 0.9842 - val_loss: 0.1967 - val_accuracy: 0.9456
-Epoch 108/1000
-60000/60000 - 6s - loss: 0.0615 - accuracy: 0.9842 - val_loss: 0.1989 - val_accuracy: 0.9457
-Epoch 109/1000
-60000/60000 - 6s - loss: 0.0610 - accuracy: 0.9845 - val_loss: 0.1980 - val_accuracy: 0.9455
-Epoch 110/1000
-60000/60000 - 6s - loss: 0.0605 - accuracy: 0.9848 - val_loss: 0.1977 - val_accuracy: 0.9459
-Epoch 111/1000
-60000/60000 - 6s - loss: 0.0602 - accuracy: 0.9850 - val_loss: 0.2003 - val_accuracy: 0.9459
-Epoch 112/1000
-60000/60000 - 6s - loss: 0.0599 - accuracy: 0.9852 - val_loss: 0.1989 - val_accuracy: 0.9456
-Epoch 113/1000
-60000/60000 - 6s - loss: 0.0592 - accuracy: 0.9852 - val_loss: 0.2008 - val_accuracy: 0.9451
-Epoch 114/1000
-60000/60000 - 6s - loss: 0.0587 - accuracy: 0.9855 - val_loss: 0.1997 - val_accuracy: 0.9461
-Epoch 115/1000
-60000/60000 - 7s - loss: 0.0584 - accuracy: 0.9854 - val_loss: 0.2018 - val_accuracy: 0.9448
-Epoch 116/1000
-60000/60000 - 6s - loss: 0.0580 - accuracy: 0.9859 - val_loss: 0.2004 - val_accuracy: 0.9452
-Epoch 117/1000
-60000/60000 - 6s - loss: 0.0578 - accuracy: 0.9858 - val_loss: 0.2032 - val_accuracy: 0.9446
-Epoch 118/1000
-60000/60000 - 6s - loss: 0.0573 - accuracy: 0.9861 - val_loss: 0.2016 - val_accuracy: 0.9463
-Epoch 119/1000
-60000/60000 - 7s - loss: 0.0569 - accuracy: 0.9861 - val_loss: 0.2030 - val_accuracy: 0.9458
-Epoch 120/1000
-60000/60000 - 6s - loss: 0.0565 - accuracy: 0.9861 - val_loss: 0.2018 - val_accuracy: 0.9459
-Epoch 121/1000
-60000/60000 - 6s - loss: 0.0563 - accuracy: 0.9864 - val_loss: 0.2031 - val_accuracy: 0.9454
-Epoch 122/1000
-60000/60000 - 6s - loss: 0.0558 - accuracy: 0.9864 - val_loss: 0.2037 - val_accuracy: 0.9458
-Epoch 123/1000
-60000/60000 - 7s - loss: 0.0554 - accuracy: 0.9869 - val_loss: 0.2042 - val_accuracy: 0.9449
-Epoch 124/1000
-60000/60000 - 7s - loss: 0.0551 - accuracy: 0.9870 - val_loss: 0.2060 - val_accuracy: 0.9454
-Epoch 125/1000
-60000/60000 - 7s - loss: 0.0546 - accuracy: 0.9871 - val_loss: 0.2045 - val_accuracy: 0.9449
-Epoch 126/1000
-60000/60000 - 7s - loss: 0.0544 - accuracy: 0.9872 - val_loss: 0.2069 - val_accuracy: 0.9456
-Epoch 127/1000
-60000/60000 - 6s - loss: 0.0540 - accuracy: 0.9872 - val_loss: 0.2049 - val_accuracy: 0.9449
-Epoch 128/1000
-60000/60000 - 6s - loss: 0.0536 - accuracy: 0.9875 - val_loss: 0.2060 - val_accuracy: 0.9448
-Epoch 129/1000
-60000/60000 - 6s - loss: 0.0534 - accuracy: 0.9875 - val_loss: 0.2068 - val_accuracy: 0.9446
-Epoch 130/1000
-60000/60000 - 6s - loss: 0.0529 - accuracy: 0.9875 - val_loss: 0.2060 - val_accuracy: 0.9454
-Epoch 131/1000
-60000/60000 - 6s - loss: 0.0528 - accuracy: 0.9877 - val_loss: 0.2070 - val_accuracy: 0.9446
-Epoch 132/1000
-60000/60000 - 6s - loss: 0.0523 - accuracy: 0.9880 - val_loss: 0.2076 - val_accuracy: 0.9447
-Epoch 133/1000
-60000/60000 - 6s - loss: 0.0519 - accuracy: 0.9880 - val_loss: 0.2086 - val_accuracy: 0.9455
-Epoch 134/1000
-60000/60000 - 6s - loss: 0.0516 - accuracy: 0.9880 - val_loss: 0.2088 - val_accuracy: 0.9440
-Epoch 135/1000
-60000/60000 - 6s - loss: 0.0512 - accuracy: 0.9884 - val_loss: 0.2099 - val_accuracy: 0.9439
-Epoch 136/1000
-60000/60000 - 6s - loss: 0.0510 - accuracy: 0.9883 - val_loss: 0.2101 - val_accuracy: 0.9441
-Epoch 137/1000
-60000/60000 - 6s - loss: 0.0508 - accuracy: 0.9887 - val_loss: 0.2080 - val_accuracy: 0.9451
-Epoch 138/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9888 - val_loss: 0.2104 - val_accuracy: 0.9442
-Epoch 139/1000
-60000/60000 - 6s - loss: 0.0501 - accuracy: 0.9887 - val_loss: 0.2107 - val_accuracy: 0.9450
-Epoch 140/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9888 - val_loss: 0.2108 - val_accuracy: 0.9443
-Epoch 141/1000
-60000/60000 - 6s - loss: 0.0495 - accuracy: 0.9888 - val_loss: 0.2112 - val_accuracy: 0.9442
-Epoch 142/1000
-60000/60000 - 6s - loss: 0.0491 - accuracy: 0.9891 - val_loss: 0.2119 - val_accuracy: 0.9445
-Epoch 143/1000
-60000/60000 - 6s - loss: 0.0489 - accuracy: 0.9888 - val_loss: 0.2118 - val_accuracy: 0.9454
-Epoch 144/1000
-60000/60000 - 6s - loss: 0.0486 - accuracy: 0.9891 - val_loss: 0.2138 - val_accuracy: 0.9439
-Epoch 145/1000
-60000/60000 - 6s - loss: 0.0482 - accuracy: 0.9892 - val_loss: 0.2140 - val_accuracy: 0.9430
-Epoch 146/1000
-60000/60000 - 6s - loss: 0.0480 - accuracy: 0.9893 - val_loss: 0.2134 - val_accuracy: 0.9445
-Epoch 147/1000
-60000/60000 - 6s - loss: 0.0477 - accuracy: 0.9898 - val_loss: 0.2148 - val_accuracy: 0.9439
-Epoch 148/1000
-60000/60000 - 6s - loss: 0.0474 - accuracy: 0.9895 - val_loss: 0.2148 - val_accuracy: 0.9448
-Epoch 149/1000
-60000/60000 - 6s - loss: 0.0472 - accuracy: 0.9897 - val_loss: 0.2156 - val_accuracy: 0.9439
-Epoch 150/1000
-60000/60000 - 6s - loss: 0.0469 - accuracy: 0.9898 - val_loss: 0.2158 - val_accuracy: 0.9432
-Epoch 151/1000
-60000/60000 - 6s - loss: 0.0467 - accuracy: 0.9898 - val_loss: 0.2154 - val_accuracy: 0.9437
-Epoch 152/1000
-60000/60000 - 6s - loss: 0.0463 - accuracy: 0.9901 - val_loss: 0.2177 - val_accuracy: 0.9430
-Epoch 153/1000
-60000/60000 - 6s - loss: 0.0463 - accuracy: 0.9899 - val_loss: 0.2162 - val_accuracy: 0.9440
-Epoch 154/1000
-60000/60000 - 6s - loss: 0.0459 - accuracy: 0.9902 - val_loss: 0.2164 - val_accuracy: 0.9433
-Epoch 155/1000
-60000/60000 - 6s - loss: 0.0456 - accuracy: 0.9901 - val_loss: 0.2167 - val_accuracy: 0.9425
-Epoch 156/1000
-60000/60000 - 6s - loss: 0.0453 - accuracy: 0.9904 - val_loss: 0.2184 - val_accuracy: 0.9431
-Epoch 157/1000
-60000/60000 - 6s - loss: 0.0452 - accuracy: 0.9906 - val_loss: 0.2185 - val_accuracy: 0.9430
-Epoch 158/1000
-60000/60000 - 6s - loss: 0.0448 - accuracy: 0.9906 - val_loss: 0.2181 - val_accuracy: 0.9429
-Epoch 159/1000
-60000/60000 - 6s - loss: 0.0445 - accuracy: 0.9905 - val_loss: 0.2197 - val_accuracy: 0.9427
-Epoch 160/1000
-60000/60000 - 6s - loss: 0.0444 - accuracy: 0.9905 - val_loss: 0.2215 - val_accuracy: 0.9427
-Epoch 161/1000
-60000/60000 - 6s - loss: 0.0441 - accuracy: 0.9907 - val_loss: 0.2200 - val_accuracy: 0.9427
-Epoch 162/1000
-60000/60000 - 6s - loss: 0.0438 - accuracy: 0.9913 - val_loss: 0.2215 - val_accuracy: 0.9423
-Epoch 163/1000
-60000/60000 - 6s - loss: 0.0436 - accuracy: 0.9908 - val_loss: 0.2194 - val_accuracy: 0.9437
-Epoch 164/1000
-60000/60000 - 6s - loss: 0.0434 - accuracy: 0.9911 - val_loss: 0.2210 - val_accuracy: 0.9426
-Epoch 165/1000
-60000/60000 - 6s - loss: 0.0431 - accuracy: 0.9913 - val_loss: 0.2225 - val_accuracy: 0.9421
-Epoch 166/1000
-60000/60000 - 6s - loss: 0.0429 - accuracy: 0.9915 - val_loss: 0.2234 - val_accuracy: 0.9427
-Epoch 167/1000
-60000/60000 - 6s - loss: 0.0426 - accuracy: 0.9914 - val_loss: 0.2221 - val_accuracy: 0.9427
-Epoch 168/1000
-60000/60000 - 6s - loss: 0.0424 - accuracy: 0.9913 - val_loss: 0.2233 - val_accuracy: 0.9430
-Epoch 169/1000
-60000/60000 - 6s - loss: 0.0422 - accuracy: 0.9915 - val_loss: 0.2239 - val_accuracy: 0.9422
-Epoch 170/1000
-60000/60000 - 6s - loss: 0.0419 - accuracy: 0.9917 - val_loss: 0.2250 - val_accuracy: 0.9410
-Epoch 171/1000
-60000/60000 - 6s - loss: 0.0417 - accuracy: 0.9917 - val_loss: 0.2243 - val_accuracy: 0.9422
-Epoch 172/1000
-60000/60000 - 6s - loss: 0.0414 - accuracy: 0.9916 - val_loss: 0.2259 - val_accuracy: 0.9419
-Epoch 173/1000
-60000/60000 - 6s - loss: 0.0412 - accuracy: 0.9917 - val_loss: 0.2238 - val_accuracy: 0.9425
-Epoch 174/1000
-60000/60000 - 6s - loss: 0.0411 - accuracy: 0.9919 - val_loss: 0.2243 - val_accuracy: 0.9433
-Epoch 175/1000
-60000/60000 - 6s - loss: 0.0409 - accuracy: 0.9918 - val_loss: 0.2258 - val_accuracy: 0.9414
-Epoch 176/1000
-60000/60000 - 6s - loss: 0.0407 - accuracy: 0.9920 - val_loss: 0.2265 - val_accuracy: 0.9426
-Epoch 177/1000
-60000/60000 - 6s - loss: 0.0404 - accuracy: 0.9919 - val_loss: 0.2276 - val_accuracy: 0.9419
-Epoch 178/1000
-60000/60000 - 6s - loss: 0.0402 - accuracy: 0.9920 - val_loss: 0.2272 - val_accuracy: 0.9428
-Epoch 179/1000
-60000/60000 - 6s - loss: 0.0401 - accuracy: 0.9920 - val_loss: 0.2274 - val_accuracy: 0.9422
-Epoch 180/1000
-60000/60000 - 6s - loss: 0.0397 - accuracy: 0.9923 - val_loss: 0.2291 - val_accuracy: 0.9416
-Epoch 181/1000
-60000/60000 - 6s - loss: 0.0395 - accuracy: 0.9924 - val_loss: 0.2287 - val_accuracy: 0.9413
-Epoch 182/1000
-60000/60000 - 6s - loss: 0.0394 - accuracy: 0.9923 - val_loss: 0.2303 - val_accuracy: 0.9413
-Epoch 183/1000
-60000/60000 - 6s - loss: 0.0391 - accuracy: 0.9925 - val_loss: 0.2290 - val_accuracy: 0.9428
-Epoch 184/1000
-60000/60000 - 6s - loss: 0.0390 - accuracy: 0.9923 - val_loss: 0.2298 - val_accuracy: 0.9425
-Epoch 185/1000
-60000/60000 - 6s - loss: 0.0388 - accuracy: 0.9926 - val_loss: 0.2312 - val_accuracy: 0.9411
-Epoch 186/1000
-60000/60000 - 6s - loss: 0.0385 - accuracy: 0.9928 - val_loss: 0.2322 - val_accuracy: 0.9406
-Epoch 187/1000
-60000/60000 - 6s - loss: 0.0383 - accuracy: 0.9927 - val_loss: 0.2320 - val_accuracy: 0.9409
-Epoch 188/1000
-60000/60000 - 6s - loss: 0.0381 - accuracy: 0.9928 - val_loss: 0.2326 - val_accuracy: 0.9416
-Epoch 189/1000
-60000/60000 - 6s - loss: 0.0379 - accuracy: 0.9930 - val_loss: 0.2331 - val_accuracy: 0.9416
-Epoch 190/1000
-60000/60000 - 6s - loss: 0.0377 - accuracy: 0.9930 - val_loss: 0.2325 - val_accuracy: 0.9417
-Epoch 191/1000
-60000/60000 - 6s - loss: 0.0375 - accuracy: 0.9929 - val_loss: 0.2351 - val_accuracy: 0.9407
-Epoch 192/1000
-60000/60000 - 6s - loss: 0.0374 - accuracy: 0.9929 - val_loss: 0.2348 - val_accuracy: 0.9413
-Epoch 193/1000
-60000/60000 - 6s - loss: 0.0372 - accuracy: 0.9930 - val_loss: 0.2347 - val_accuracy: 0.9405
-Epoch 194/1000
-60000/60000 - 6s - loss: 0.0370 - accuracy: 0.9929 - val_loss: 0.2356 - val_accuracy: 0.9410
-Epoch 195/1000
-60000/60000 - 6s - loss: 0.0368 - accuracy: 0.9930 - val_loss: 0.2349 - val_accuracy: 0.9405
-Epoch 196/1000
-60000/60000 - 6s - loss: 0.0366 - accuracy: 0.9933 - val_loss: 0.2359 - val_accuracy: 0.9417
-Epoch 197/1000
-60000/60000 - 6s - loss: 0.0362 - accuracy: 0.9933 - val_loss: 0.2377 - val_accuracy: 0.9407
-Epoch 198/1000
-60000/60000 - 6s - loss: 0.0363 - accuracy: 0.9933 - val_loss: 0.2354 - val_accuracy: 0.9412
-Epoch 199/1000
-60000/60000 - 6s - loss: 0.0360 - accuracy: 0.9935 - val_loss: 0.2370 - val_accuracy: 0.9407
-Epoch 200/1000
-60000/60000 - 6s - loss: 0.0359 - accuracy: 0.9935 - val_loss: 0.2372 - val_accuracy: 0.9408
-Epoch 201/1000
-60000/60000 - 6s - loss: 0.0357 - accuracy: 0.9936 - val_loss: 0.2381 - val_accuracy: 0.9417
-Epoch 202/1000
-60000/60000 - 6s - loss: 0.0355 - accuracy: 0.9934 - val_loss: 0.2382 - val_accuracy: 0.9403
-Epoch 203/1000
-60000/60000 - 6s - loss: 0.0353 - accuracy: 0.9937 - val_loss: 0.2398 - val_accuracy: 0.9408
-Epoch 204/1000
-60000/60000 - 6s - loss: 0.0351 - accuracy: 0.9937 - val_loss: 0.2381 - val_accuracy: 0.9405
-Epoch 205/1000
-60000/60000 - 6s - loss: 0.0349 - accuracy: 0.9938 - val_loss: 0.2396 - val_accuracy: 0.9407
-Epoch 206/1000
-60000/60000 - 6s - loss: 0.0348 - accuracy: 0.9934 - val_loss: 0.2391 - val_accuracy: 0.9409
-Epoch 207/1000
-60000/60000 - 6s - loss: 0.0347 - accuracy: 0.9938 - val_loss: 0.2399 - val_accuracy: 0.9417
-Epoch 208/1000
-60000/60000 - 6s - loss: 0.0345 - accuracy: 0.9940 - val_loss: 0.2406 - val_accuracy: 0.9408
-Epoch 209/1000
-60000/60000 - 6s - loss: 0.0343 - accuracy: 0.9940 - val_loss: 0.2412 - val_accuracy: 0.9411
-Epoch 210/1000
-60000/60000 - 6s - loss: 0.0342 - accuracy: 0.9941 - val_loss: 0.2424 - val_accuracy: 0.9403
-Epoch 211/1000
-60000/60000 - 6s - loss: 0.0339 - accuracy: 0.9938 - val_loss: 0.2421 - val_accuracy: 0.9406
-Epoch 212/1000
-60000/60000 - 6s - loss: 0.0338 - accuracy: 0.9942 - val_loss: 0.2420 - val_accuracy: 0.9399
-Epoch 213/1000
-60000/60000 - 6s - loss: 0.0335 - accuracy: 0.9941 - val_loss: 0.2441 - val_accuracy: 0.9399
-Epoch 214/1000
-60000/60000 - 6s - loss: 0.0334 - accuracy: 0.9942 - val_loss: 0.2444 - val_accuracy: 0.9396
-Epoch 215/1000
-60000/60000 - 6s - loss: 0.0333 - accuracy: 0.9942 - val_loss: 0.2441 - val_accuracy: 0.9402
-Epoch 216/1000
-60000/60000 - 6s - loss: 0.0331 - accuracy: 0.9943 - val_loss: 0.2452 - val_accuracy: 0.9400
-Epoch 217/1000
-60000/60000 - 6s - loss: 0.0329 - accuracy: 0.9944 - val_loss: 0.2443 - val_accuracy: 0.9407
-Epoch 218/1000
-60000/60000 - 6s - loss: 0.0327 - accuracy: 0.9942 - val_loss: 0.2456 - val_accuracy: 0.9404
-Epoch 219/1000
-60000/60000 - 6s - loss: 0.0326 - accuracy: 0.9944 - val_loss: 0.2462 - val_accuracy: 0.9405
-Epoch 220/1000
-60000/60000 - 6s - loss: 0.0325 - accuracy: 0.9946 - val_loss: 0.2468 - val_accuracy: 0.9395
-Epoch 221/1000
-60000/60000 - 6s - loss: 0.0324 - accuracy: 0.9944 - val_loss: 0.2467 - val_accuracy: 0.9404
-Epoch 222/1000
-60000/60000 - 6s - loss: 0.0322 - accuracy: 0.9945 - val_loss: 0.2465 - val_accuracy: 0.9397
-Epoch 223/1000
-60000/60000 - 6s - loss: 0.0320 - accuracy: 0.9948 - val_loss: 0.2482 - val_accuracy: 0.9399
-Epoch 224/1000
-60000/60000 - 6s - loss: 0.0319 - accuracy: 0.9945 - val_loss: 0.2483 - val_accuracy: 0.9396
-Epoch 225/1000
-60000/60000 - 6s - loss: 0.0317 - accuracy: 0.9945 - val_loss: 0.2490 - val_accuracy: 0.9390
-Epoch 226/1000
-60000/60000 - 6s - loss: 0.0316 - accuracy: 0.9946 - val_loss: 0.2492 - val_accuracy: 0.9403
-Epoch 227/1000
-60000/60000 - 6s - loss: 0.0314 - accuracy: 0.9946 - val_loss: 0.2506 - val_accuracy: 0.9404
-Epoch 228/1000
-60000/60000 - 6s - loss: 0.0313 - accuracy: 0.9951 - val_loss: 0.2499 - val_accuracy: 0.9392
-Epoch 229/1000
-60000/60000 - 6s - loss: 0.0311 - accuracy: 0.9950 - val_loss: 0.2494 - val_accuracy: 0.9399
-Epoch 230/1000
-60000/60000 - 6s - loss: 0.0310 - accuracy: 0.9950 - val_loss: 0.2506 - val_accuracy: 0.9395
-Epoch 231/1000
-60000/60000 - 6s - loss: 0.0308 - accuracy: 0.9951 - val_loss: 0.2508 - val_accuracy: 0.9399
-Epoch 232/1000
-60000/60000 - 6s - loss: 0.0307 - accuracy: 0.9949 - val_loss: 0.2515 - val_accuracy: 0.9390
-Epoch 233/1000
-60000/60000 - 6s - loss: 0.0305 - accuracy: 0.9950 - val_loss: 0.2519 - val_accuracy: 0.9394
-Epoch 234/1000
-60000/60000 - 6s - loss: 0.0304 - accuracy: 0.9952 - val_loss: 0.2530 - val_accuracy: 0.9394
-Epoch 235/1000
-60000/60000 - 6s - loss: 0.0303 - accuracy: 0.9950 - val_loss: 0.2526 - val_accuracy: 0.9397
-Epoch 236/1000
-60000/60000 - 6s - loss: 0.0300 - accuracy: 0.9953 - val_loss: 0.2539 - val_accuracy: 0.9395
-Epoch 237/1000
-60000/60000 - 6s - loss: 0.0300 - accuracy: 0.9950 - val_loss: 0.2536 - val_accuracy: 0.9395
-Epoch 238/1000
-60000/60000 - 6s - loss: 0.0298 - accuracy: 0.9951 - val_loss: 0.2555 - val_accuracy: 0.9394
-Epoch 239/1000
-60000/60000 - 6s - loss: 0.0298 - accuracy: 0.9953 - val_loss: 0.2548 - val_accuracy: 0.9388
-Epoch 240/1000
-60000/60000 - 6s - loss: 0.0295 - accuracy: 0.9952 - val_loss: 0.2570 - val_accuracy: 0.9386
-Epoch 241/1000
-60000/60000 - 6s - loss: 0.0295 - accuracy: 0.9954 - val_loss: 0.2561 - val_accuracy: 0.9387
-Epoch 242/1000
-60000/60000 - 6s - loss: 0.0293 - accuracy: 0.9952 - val_loss: 0.2550 - val_accuracy: 0.9410
-Epoch 243/1000
-60000/60000 - 6s - loss: 0.0292 - accuracy: 0.9955 - val_loss: 0.2573 - val_accuracy: 0.9396
-Epoch 244/1000
-60000/60000 - 6s - loss: 0.0290 - accuracy: 0.9954 - val_loss: 0.2558 - val_accuracy: 0.9396
-Epoch 245/1000
-60000/60000 - 6s - loss: 0.0290 - accuracy: 0.9953 - val_loss: 0.2580 - val_accuracy: 0.9396
-Epoch 246/1000
-60000/60000 - 6s - loss: 0.0287 - accuracy: 0.9956 - val_loss: 0.2587 - val_accuracy: 0.9389
-Epoch 247/1000
-60000/60000 - 6s - loss: 0.0287 - accuracy: 0.9956 - val_loss: 0.2592 - val_accuracy: 0.9385
-Epoch 248/1000
-60000/60000 - 6s - loss: 0.0286 - accuracy: 0.9956 - val_loss: 0.2588 - val_accuracy: 0.9394
-Epoch 249/1000
-60000/60000 - 6s - loss: 0.0284 - accuracy: 0.9958 - val_loss: 0.2578 - val_accuracy: 0.9388
-Epoch 250/1000
-60000/60000 - 6s - loss: 0.0282 - accuracy: 0.9958 - val_loss: 0.2604 - val_accuracy: 0.9389
-Epoch 251/1000
-60000/60000 - 6s - loss: 0.0282 - accuracy: 0.9958 - val_loss: 0.2589 - val_accuracy: 0.9398
-Epoch 252/1000
-60000/60000 - 6s - loss: 0.0280 - accuracy: 0.9959 - val_loss: 0.2608 - val_accuracy: 0.9396
-Epoch 253/1000
-60000/60000 - 6s - loss: 0.0279 - accuracy: 0.9958 - val_loss: 0.2605 - val_accuracy: 0.9386
-Epoch 254/1000
-60000/60000 - 6s - loss: 0.0278 - accuracy: 0.9957 - val_loss: 0.2625 - val_accuracy: 0.9394
-Epoch 255/1000
-60000/60000 - 6s - loss: 0.0276 - accuracy: 0.9958 - val_loss: 0.2624 - val_accuracy: 0.9394
-Epoch 256/1000
-60000/60000 - 6s - loss: 0.0276 - accuracy: 0.9960 - val_loss: 0.2620 - val_accuracy: 0.9383
-Epoch 257/1000
-60000/60000 - 6s - loss: 0.0274 - accuracy: 0.9960 - val_loss: 0.2619 - val_accuracy: 0.9386
-Epoch 258/1000
-60000/60000 - 6s - loss: 0.0273 - accuracy: 0.9959 - val_loss: 0.2624 - val_accuracy: 0.9391
-Epoch 259/1000
-60000/60000 - 6s - loss: 0.0271 - accuracy: 0.9962 - val_loss: 0.2662 - val_accuracy: 0.9397
-Epoch 260/1000
-60000/60000 - 6s - loss: 0.0271 - accuracy: 0.9959 - val_loss: 0.2648 - val_accuracy: 0.9379
-Epoch 261/1000
-60000/60000 - 6s - loss: 0.0270 - accuracy: 0.9961 - val_loss: 0.2639 - val_accuracy: 0.9390
-Epoch 262/1000
-60000/60000 - 6s - loss: 0.0268 - accuracy: 0.9961 - val_loss: 0.2668 - val_accuracy: 0.9379
-Epoch 263/1000
-60000/60000 - 6s - loss: 0.0267 - accuracy: 0.9962 - val_loss: 0.2656 - val_accuracy: 0.9381
-Epoch 264/1000
-60000/60000 - 6s - loss: 0.0265 - accuracy: 0.9963 - val_loss: 0.2651 - val_accuracy: 0.9385
-Epoch 265/1000
-60000/60000 - 6s - loss: 0.0265 - accuracy: 0.9962 - val_loss: 0.2664 - val_accuracy: 0.9386
-Epoch 266/1000
-60000/60000 - 6s - loss: 0.0264 - accuracy: 0.9963 - val_loss: 0.2667 - val_accuracy: 0.9386
-Epoch 267/1000
-60000/60000 - 6s - loss: 0.0263 - accuracy: 0.9961 - val_loss: 0.2669 - val_accuracy: 0.9380
-Epoch 268/1000
-60000/60000 - 6s - loss: 0.0261 - accuracy: 0.9964 - val_loss: 0.2669 - val_accuracy: 0.9382
-Epoch 269/1000
-60000/60000 - 6s - loss: 0.0260 - accuracy: 0.9964 - val_loss: 0.2687 - val_accuracy: 0.9380
-Epoch 270/1000
-60000/60000 - 6s - loss: 0.0259 - accuracy: 0.9963 - val_loss: 0.2688 - val_accuracy: 0.9383
-Epoch 271/1000
-60000/60000 - 6s - loss: 0.0258 - accuracy: 0.9966 - val_loss: 0.2699 - val_accuracy: 0.9376
-Epoch 272/1000
-60000/60000 - 6s - loss: 0.0256 - accuracy: 0.9964 - val_loss: 0.2682 - val_accuracy: 0.9387
-Epoch 273/1000
-60000/60000 - 6s - loss: 0.0255 - accuracy: 0.9965 - val_loss: 0.2690 - val_accuracy: 0.9390
-Epoch 274/1000
-60000/60000 - 6s - loss: 0.0255 - accuracy: 0.9966 - val_loss: 0.2699 - val_accuracy: 0.9387
-Epoch 275/1000
-60000/60000 - 6s - loss: 0.0254 - accuracy: 0.9965 - val_loss: 0.2694 - val_accuracy: 0.9384
-Epoch 276/1000
-60000/60000 - 6s - loss: 0.0253 - accuracy: 0.9964 - val_loss: 0.2709 - val_accuracy: 0.9378
-Epoch 277/1000
-60000/60000 - 6s - loss: 0.0251 - accuracy: 0.9968 - val_loss: 0.2723 - val_accuracy: 0.9379
-Epoch 278/1000
-60000/60000 - 6s - loss: 0.0251 - accuracy: 0.9966 - val_loss: 0.2723 - val_accuracy: 0.9375
-Epoch 279/1000
-60000/60000 - 6s - loss: 0.0250 - accuracy: 0.9966 - val_loss: 0.2713 - val_accuracy: 0.9379
-Epoch 280/1000
-60000/60000 - 6s - loss: 0.0249 - accuracy: 0.9967 - val_loss: 0.2735 - val_accuracy: 0.9379
-Epoch 281/1000
-60000/60000 - 6s - loss: 0.0247 - accuracy: 0.9966 - val_loss: 0.2752 - val_accuracy: 0.9375
-Epoch 282/1000
-60000/60000 - 6s - loss: 0.0246 - accuracy: 0.9968 - val_loss: 0.2744 - val_accuracy: 0.9381
-Epoch 283/1000
-60000/60000 - 6s - loss: 0.0245 - accuracy: 0.9966 - val_loss: 0.2749 - val_accuracy: 0.9370
-Epoch 284/1000
-60000/60000 - 6s - loss: 0.0245 - accuracy: 0.9968 - val_loss: 0.2735 - val_accuracy: 0.9374
-Epoch 285/1000
-60000/60000 - 6s - loss: 0.0244 - accuracy: 0.9968 - val_loss: 0.2744 - val_accuracy: 0.9379
-Epoch 286/1000
-60000/60000 - 6s - loss: 0.0243 - accuracy: 0.9969 - val_loss: 0.2755 - val_accuracy: 0.9375
-Epoch 287/1000
-60000/60000 - 6s - loss: 0.0242 - accuracy: 0.9967 - val_loss: 0.2756 - val_accuracy: 0.9376
-Epoch 288/1000
-60000/60000 - 6s - loss: 0.0241 - accuracy: 0.9970 - val_loss: 0.2768 - val_accuracy: 0.9373
-Epoch 289/1000
-60000/60000 - 6s - loss: 0.0239 - accuracy: 0.9970 - val_loss: 0.2775 - val_accuracy: 0.9383
-Epoch 290/1000
-60000/60000 - 6s - loss: 0.0239 - accuracy: 0.9969 - val_loss: 0.2769 - val_accuracy: 0.9375
-Epoch 291/1000
-60000/60000 - 6s - loss: 0.0237 - accuracy: 0.9970 - val_loss: 0.2770 - val_accuracy: 0.9378
-Epoch 292/1000
-60000/60000 - 6s - loss: 0.0237 - accuracy: 0.9969 - val_loss: 0.2780 - val_accuracy: 0.9369
-Epoch 293/1000
-60000/60000 - 6s - loss: 0.0236 - accuracy: 0.9972 - val_loss: 0.2779 - val_accuracy: 0.9378
-Epoch 294/1000
-60000/60000 - 6s - loss: 0.0235 - accuracy: 0.9969 - val_loss: 0.2788 - val_accuracy: 0.9365
-Epoch 295/1000
-60000/60000 - 6s - loss: 0.0234 - accuracy: 0.9972 - val_loss: 0.2785 - val_accuracy: 0.9370
-Epoch 296/1000
-60000/60000 - 6s - loss: 0.0233 - accuracy: 0.9970 - val_loss: 0.2782 - val_accuracy: 0.9371
-Epoch 297/1000
-60000/60000 - 6s - loss: 0.0233 - accuracy: 0.9971 - val_loss: 0.2789 - val_accuracy: 0.9366
-Epoch 298/1000
-60000/60000 - 6s - loss: 0.0231 - accuracy: 0.9971 - val_loss: 0.2805 - val_accuracy: 0.9372
-Epoch 299/1000
-60000/60000 - 6s - loss: 0.0231 - accuracy: 0.9972 - val_loss: 0.2791 - val_accuracy: 0.9370
-Epoch 300/1000
-60000/60000 - 6s - loss: 0.0230 - accuracy: 0.9971 - val_loss: 0.2811 - val_accuracy: 0.9368
-Epoch 301/1000
-60000/60000 - 6s - loss: 0.0229 - accuracy: 0.9972 - val_loss: 0.2822 - val_accuracy: 0.9376
-Epoch 302/1000
-60000/60000 - 6s - loss: 0.0227 - accuracy: 0.9972 - val_loss: 0.2807 - val_accuracy: 0.9373
-Epoch 303/1000
-60000/60000 - 6s - loss: 0.0227 - accuracy: 0.9973 - val_loss: 0.2832 - val_accuracy: 0.9369
-Epoch 304/1000
-60000/60000 - 6s - loss: 0.0227 - accuracy: 0.9972 - val_loss: 0.2815 - val_accuracy: 0.9371
-Epoch 305/1000
-60000/60000 - 6s - loss: 0.0225 - accuracy: 0.9973 - val_loss: 0.2814 - val_accuracy: 0.9365
-Epoch 306/1000
-60000/60000 - 6s - loss: 0.0224 - accuracy: 0.9973 - val_loss: 0.2828 - val_accuracy: 0.9372
-Epoch 307/1000
-60000/60000 - 6s - loss: 0.0224 - accuracy: 0.9973 - val_loss: 0.2830 - val_accuracy: 0.9374
-Epoch 308/1000
-60000/60000 - 6s - loss: 0.0222 - accuracy: 0.9974 - val_loss: 0.2835 - val_accuracy: 0.9369
-Epoch 309/1000
-60000/60000 - 6s - loss: 0.0222 - accuracy: 0.9975 - val_loss: 0.2841 - val_accuracy: 0.9367
-Epoch 310/1000
-60000/60000 - 6s - loss: 0.0221 - accuracy: 0.9974 - val_loss: 0.2843 - val_accuracy: 0.9374
-Epoch 311/1000
-60000/60000 - 6s - loss: 0.0221 - accuracy: 0.9974 - val_loss: 0.2843 - val_accuracy: 0.9367
-Epoch 312/1000
-60000/60000 - 6s - loss: 0.0219 - accuracy: 0.9973 - val_loss: 0.2846 - val_accuracy: 0.9370
-Epoch 313/1000
-60000/60000 - 7s - loss: 0.0218 - accuracy: 0.9973 - val_loss: 0.2865 - val_accuracy: 0.9366
-Epoch 314/1000
-60000/60000 - 7s - loss: 0.0218 - accuracy: 0.9973 - val_loss: 0.2859 - val_accuracy: 0.9372
-Epoch 315/1000
-60000/60000 - 7s - loss: 0.0217 - accuracy: 0.9975 - val_loss: 0.2873 - val_accuracy: 0.9360
-Epoch 316/1000
-60000/60000 - 7s - loss: 0.0216 - accuracy: 0.9976 - val_loss: 0.2858 - val_accuracy: 0.9372
-Epoch 317/1000
-60000/60000 - 7s - loss: 0.0215 - accuracy: 0.9975 - val_loss: 0.2879 - val_accuracy: 0.9363
-Epoch 318/1000
-60000/60000 - 6s - loss: 0.0214 - accuracy: 0.9975 - val_loss: 0.2869 - val_accuracy: 0.9366
-Epoch 319/1000
-60000/60000 - 7s - loss: 0.0214 - accuracy: 0.9976 - val_loss: 0.2893 - val_accuracy: 0.9360
-Epoch 320/1000
-60000/60000 - 6s - loss: 0.0213 - accuracy: 0.9975 - val_loss: 0.2886 - val_accuracy: 0.9365
-Epoch 321/1000
-60000/60000 - 6s - loss: 0.0212 - accuracy: 0.9976 - val_loss: 0.2892 - val_accuracy: 0.9368
-Epoch 322/1000
-60000/60000 - 7s - loss: 0.0212 - accuracy: 0.9977 - val_loss: 0.2900 - val_accuracy: 0.9357
-Epoch 323/1000
-60000/60000 - 6s - loss: 0.0211 - accuracy: 0.9975 - val_loss: 0.2897 - val_accuracy: 0.9368
-Epoch 324/1000
-60000/60000 - 6s - loss: 0.0210 - accuracy: 0.9977 - val_loss: 0.2902 - val_accuracy: 0.9359
-Epoch 325/1000
-60000/60000 - 7s - loss: 0.0209 - accuracy: 0.9976 - val_loss: 0.2906 - val_accuracy: 0.9364
-Epoch 326/1000
-60000/60000 - 6s - loss: 0.0208 - accuracy: 0.9977 - val_loss: 0.2916 - val_accuracy: 0.9358
-Epoch 327/1000
-60000/60000 - 6s - loss: 0.0208 - accuracy: 0.9977 - val_loss: 0.2925 - val_accuracy: 0.9353
-Epoch 328/1000
-60000/60000 - 7s - loss: 0.0207 - accuracy: 0.9976 - val_loss: 0.2919 - val_accuracy: 0.9363
-Epoch 329/1000
-60000/60000 - 6s - loss: 0.0207 - accuracy: 0.9977 - val_loss: 0.2917 - val_accuracy: 0.9366
-Epoch 330/1000
-60000/60000 - 6s - loss: 0.0205 - accuracy: 0.9978 - val_loss: 0.2938 - val_accuracy: 0.9360
-Epoch 331/1000
-60000/60000 - 7s - loss: 0.0204 - accuracy: 0.9977 - val_loss: 0.2919 - val_accuracy: 0.9362
-Epoch 332/1000
-60000/60000 - 7s - loss: 0.0204 - accuracy: 0.9977 - val_loss: 0.2929 - val_accuracy: 0.9364
-Epoch 333/1000
-60000/60000 - 7s - loss: 0.0203 - accuracy: 0.9977 - val_loss: 0.2930 - val_accuracy: 0.9362
-Epoch 334/1000
-60000/60000 - 7s - loss: 0.0202 - accuracy: 0.9977 - val_loss: 0.2943 - val_accuracy: 0.9360
-Epoch 335/1000
-60000/60000 - 7s - loss: 0.0202 - accuracy: 0.9978 - val_loss: 0.2949 - val_accuracy: 0.9361
-Epoch 336/1000
-60000/60000 - 6s - loss: 0.0201 - accuracy: 0.9978 - val_loss: 0.2957 - val_accuracy: 0.9354
-Epoch 337/1000
-60000/60000 - 6s - loss: 0.0201 - accuracy: 0.9980 - val_loss: 0.2956 - val_accuracy: 0.9361
-Epoch 338/1000
-60000/60000 - 6s - loss: 0.0200 - accuracy: 0.9979 - val_loss: 0.2952 - val_accuracy: 0.9369
-Epoch 339/1000
-60000/60000 - 6s - loss: 0.0199 - accuracy: 0.9979 - val_loss: 0.2957 - val_accuracy: 0.9359
-Epoch 340/1000
-60000/60000 - 6s - loss: 0.0199 - accuracy: 0.9979 - val_loss: 0.2955 - val_accuracy: 0.9361
-Epoch 341/1000
-60000/60000 - 6s - loss: 0.0197 - accuracy: 0.9979 - val_loss: 0.2964 - val_accuracy: 0.9353
-Epoch 342/1000
-60000/60000 - 6s - loss: 0.0197 - accuracy: 0.9980 - val_loss: 0.2971 - val_accuracy: 0.9362
-Epoch 343/1000
-60000/60000 - 6s - loss: 0.0196 - accuracy: 0.9980 - val_loss: 0.2974 - val_accuracy: 0.9364
-Epoch 344/1000
-60000/60000 - 6s - loss: 0.0195 - accuracy: 0.9979 - val_loss: 0.2971 - val_accuracy: 0.9359
-Epoch 345/1000
-60000/60000 - 6s - loss: 0.0195 - accuracy: 0.9980 - val_loss: 0.2976 - val_accuracy: 0.9362
-Epoch 346/1000
-60000/60000 - 6s - loss: 0.0194 - accuracy: 0.9980 - val_loss: 0.2974 - val_accuracy: 0.9356
-Epoch 347/1000
-60000/60000 - 6s - loss: 0.0193 - accuracy: 0.9980 - val_loss: 0.2974 - val_accuracy: 0.9359
-Epoch 348/1000
-60000/60000 - 6s - loss: 0.0193 - accuracy: 0.9981 - val_loss: 0.2996 - val_accuracy: 0.9353
-Epoch 349/1000
-60000/60000 - 6s - loss: 0.0192 - accuracy: 0.9980 - val_loss: 0.2996 - val_accuracy: 0.9356
-Epoch 350/1000
-60000/60000 - 6s - loss: 0.0192 - accuracy: 0.9981 - val_loss: 0.2998 - val_accuracy: 0.9359
-Epoch 351/1000
-60000/60000 - 6s - loss: 0.0191 - accuracy: 0.9981 - val_loss: 0.3006 - val_accuracy: 0.9365
-Epoch 352/1000
-60000/60000 - 6s - loss: 0.0190 - accuracy: 0.9980 - val_loss: 0.3012 - val_accuracy: 0.9361
-Epoch 353/1000
-60000/60000 - 6s - loss: 0.0189 - accuracy: 0.9980 - val_loss: 0.2997 - val_accuracy: 0.9360
-Epoch 354/1000
-60000/60000 - 6s - loss: 0.0189 - accuracy: 0.9981 - val_loss: 0.3005 - val_accuracy: 0.9359
-Epoch 355/1000
-60000/60000 - 6s - loss: 0.0188 - accuracy: 0.9981 - val_loss: 0.3011 - val_accuracy: 0.9357
-Epoch 356/1000
-60000/60000 - 6s - loss: 0.0188 - accuracy: 0.9981 - val_loss: 0.3022 - val_accuracy: 0.9356
-Epoch 357/1000
-60000/60000 - 6s - loss: 0.0187 - accuracy: 0.9981 - val_loss: 0.3028 - val_accuracy: 0.9357
-Epoch 358/1000
-60000/60000 - 6s - loss: 0.0187 - accuracy: 0.9981 - val_loss: 0.3026 - val_accuracy: 0.9354
-Epoch 359/1000
-60000/60000 - 6s - loss: 0.0186 - accuracy: 0.9981 - val_loss: 0.3031 - val_accuracy: 0.9353
-Epoch 360/1000
-60000/60000 - 6s - loss: 0.0185 - accuracy: 0.9981 - val_loss: 0.3044 - val_accuracy: 0.9347
-Epoch 361/1000
-60000/60000 - 6s - loss: 0.0184 - accuracy: 0.9982 - val_loss: 0.3044 - val_accuracy: 0.9351
-Epoch 362/1000
-60000/60000 - 6s - loss: 0.0184 - accuracy: 0.9983 - val_loss: 0.3041 - val_accuracy: 0.9349
-Epoch 363/1000
-60000/60000 - 6s - loss: 0.0183 - accuracy: 0.9981 - val_loss: 0.3047 - val_accuracy: 0.9361
-Epoch 364/1000
-60000/60000 - 6s - loss: 0.0183 - accuracy: 0.9982 - val_loss: 0.3045 - val_accuracy: 0.9352
-Epoch 365/1000
-60000/60000 - 6s - loss: 0.0182 - accuracy: 0.9983 - val_loss: 0.3049 - val_accuracy: 0.9360
-Epoch 366/1000
-60000/60000 - 6s - loss: 0.0182 - accuracy: 0.9982 - val_loss: 0.3050 - val_accuracy: 0.9354
-Epoch 367/1000
-60000/60000 - 6s - loss: 0.0181 - accuracy: 0.9982 - val_loss: 0.3056 - val_accuracy: 0.9346
-Epoch 368/1000
-60000/60000 - 6s - loss: 0.0180 - accuracy: 0.9984 - val_loss: 0.3053 - val_accuracy: 0.9358
-Epoch 369/1000
-60000/60000 - 6s - loss: 0.0180 - accuracy: 0.9983 - val_loss: 0.3062 - val_accuracy: 0.9352
-Epoch 370/1000
-60000/60000 - 6s - loss: 0.0179 - accuracy: 0.9983 - val_loss: 0.3067 - val_accuracy: 0.9355
-Epoch 371/1000
-60000/60000 - 6s - loss: 0.0179 - accuracy: 0.9984 - val_loss: 0.3074 - val_accuracy: 0.9363
-Epoch 372/1000
-60000/60000 - 6s - loss: 0.0178 - accuracy: 0.9982 - val_loss: 0.3072 - val_accuracy: 0.9355
-Epoch 373/1000
-60000/60000 - 6s - loss: 0.0177 - accuracy: 0.9983 - val_loss: 0.3086 - val_accuracy: 0.9358
-Epoch 374/1000
-60000/60000 - 6s - loss: 0.0177 - accuracy: 0.9984 - val_loss: 0.3099 - val_accuracy: 0.9359
-Epoch 375/1000
-60000/60000 - 6s - loss: 0.0176 - accuracy: 0.9984 - val_loss: 0.3081 - val_accuracy: 0.9360
-Epoch 376/1000
-60000/60000 - 6s - loss: 0.0175 - accuracy: 0.9984 - val_loss: 0.3102 - val_accuracy: 0.9356
-Epoch 377/1000
-60000/60000 - 6s - loss: 0.0175 - accuracy: 0.9984 - val_loss: 0.3102 - val_accuracy: 0.9349
-Epoch 378/1000
-60000/60000 - 6s - loss: 0.0174 - accuracy: 0.9983 - val_loss: 0.3095 - val_accuracy: 0.9352
-Epoch 379/1000
-60000/60000 - 6s - loss: 0.0174 - accuracy: 0.9984 - val_loss: 0.3103 - val_accuracy: 0.9355
-Epoch 380/1000
-60000/60000 - 6s - loss: 0.0173 - accuracy: 0.9984 - val_loss: 0.3108 - val_accuracy: 0.9351
-Epoch 381/1000
-60000/60000 - 6s - loss: 0.0173 - accuracy: 0.9984 - val_loss: 0.3119 - val_accuracy: 0.9342
-Epoch 382/1000
-60000/60000 - 6s - loss: 0.0172 - accuracy: 0.9985 - val_loss: 0.3121 - val_accuracy: 0.9348
-Epoch 383/1000
-60000/60000 - 6s - loss: 0.0171 - accuracy: 0.9984 - val_loss: 0.3113 - val_accuracy: 0.9351
-Epoch 384/1000
-60000/60000 - 6s - loss: 0.0171 - accuracy: 0.9985 - val_loss: 0.3119 - val_accuracy: 0.9353
-Epoch 385/1000
-60000/60000 - 6s - loss: 0.0170 - accuracy: 0.9985 - val_loss: 0.3116 - val_accuracy: 0.9361
-Epoch 386/1000
-60000/60000 - 6s - loss: 0.0170 - accuracy: 0.9983 - val_loss: 0.3131 - val_accuracy: 0.9350
-Epoch 387/1000
-60000/60000 - 6s - loss: 0.0170 - accuracy: 0.9985 - val_loss: 0.3130 - val_accuracy: 0.9351
-Epoch 388/1000
-60000/60000 - 6s - loss: 0.0169 - accuracy: 0.9985 - val_loss: 0.3132 - val_accuracy: 0.9347
-Epoch 389/1000
-60000/60000 - 6s - loss: 0.0168 - accuracy: 0.9985 - val_loss: 0.3141 - val_accuracy: 0.9356
-Epoch 390/1000
-60000/60000 - 6s - loss: 0.0168 - accuracy: 0.9986 - val_loss: 0.3144 - val_accuracy: 0.9351
-Epoch 391/1000
-60000/60000 - 6s - loss: 0.0168 - accuracy: 0.9985 - val_loss: 0.3147 - val_accuracy: 0.9339
-Epoch 392/1000
-60000/60000 - 6s - loss: 0.0167 - accuracy: 0.9985 - val_loss: 0.3146 - val_accuracy: 0.9348
-Epoch 393/1000
-60000/60000 - 6s - loss: 0.0166 - accuracy: 0.9984 - val_loss: 0.3152 - val_accuracy: 0.9348
-Epoch 394/1000
-60000/60000 - 6s - loss: 0.0165 - accuracy: 0.9986 - val_loss: 0.3156 - val_accuracy: 0.9344
-Epoch 395/1000
-60000/60000 - 6s - loss: 0.0165 - accuracy: 0.9985 - val_loss: 0.3162 - val_accuracy: 0.9344
-Epoch 396/1000
-60000/60000 - 6s - loss: 0.0165 - accuracy: 0.9985 - val_loss: 0.3158 - val_accuracy: 0.9361
-Epoch 397/1000
-60000/60000 - 6s - loss: 0.0164 - accuracy: 0.9986 - val_loss: 0.3166 - val_accuracy: 0.9357
-Epoch 398/1000
-60000/60000 - 6s - loss: 0.0164 - accuracy: 0.9986 - val_loss: 0.3173 - val_accuracy: 0.9346
-Epoch 399/1000
-60000/60000 - 6s - loss: 0.0163 - accuracy: 0.9986 - val_loss: 0.3160 - val_accuracy: 0.9352
-Epoch 400/1000
-60000/60000 - 6s - loss: 0.0163 - accuracy: 0.9986 - val_loss: 0.3166 - val_accuracy: 0.9355
-Epoch 401/1000
-60000/60000 - 6s - loss: 0.0162 - accuracy: 0.9986 - val_loss: 0.3177 - val_accuracy: 0.9347
-Epoch 402/1000
-60000/60000 - 6s - loss: 0.0162 - accuracy: 0.9987 - val_loss: 0.3192 - val_accuracy: 0.9346
-Epoch 403/1000
-60000/60000 - 6s - loss: 0.0161 - accuracy: 0.9987 - val_loss: 0.3190 - val_accuracy: 0.9347
-Epoch 404/1000
-60000/60000 - 6s - loss: 0.0161 - accuracy: 0.9986 - val_loss: 0.3190 - val_accuracy: 0.9349
-Epoch 405/1000
-60000/60000 - 6s - loss: 0.0160 - accuracy: 0.9987 - val_loss: 0.3179 - val_accuracy: 0.9346
-Epoch 406/1000
-60000/60000 - 6s - loss: 0.0160 - accuracy: 0.9987 - val_loss: 0.3192 - val_accuracy: 0.9355
-Epoch 407/1000
-60000/60000 - 6s - loss: 0.0159 - accuracy: 0.9987 - val_loss: 0.3199 - val_accuracy: 0.9351
-Epoch 408/1000
-60000/60000 - 6s - loss: 0.0159 - accuracy: 0.9987 - val_loss: 0.3200 - val_accuracy: 0.9348
-Epoch 409/1000
-60000/60000 - 6s - loss: 0.0158 - accuracy: 0.9987 - val_loss: 0.3200 - val_accuracy: 0.9352
-Epoch 410/1000
-60000/60000 - 6s - loss: 0.0158 - accuracy: 0.9987 - val_loss: 0.3202 - val_accuracy: 0.9346
-Epoch 411/1000
-60000/60000 - 6s - loss: 0.0157 - accuracy: 0.9986 - val_loss: 0.3203 - val_accuracy: 0.9346
-Epoch 412/1000
-60000/60000 - 6s - loss: 0.0157 - accuracy: 0.9987 - val_loss: 0.3210 - val_accuracy: 0.9348
-Epoch 413/1000
-60000/60000 - 6s - loss: 0.0156 - accuracy: 0.9988 - val_loss: 0.3213 - val_accuracy: 0.9347
-Epoch 414/1000
-60000/60000 - 6s - loss: 0.0156 - accuracy: 0.9987 - val_loss: 0.3230 - val_accuracy: 0.9347
-Epoch 415/1000
-60000/60000 - 6s - loss: 0.0155 - accuracy: 0.9987 - val_loss: 0.3220 - val_accuracy: 0.9351
-Epoch 416/1000
-60000/60000 - 6s - loss: 0.0155 - accuracy: 0.9988 - val_loss: 0.3226 - val_accuracy: 0.9348
-Epoch 417/1000
-60000/60000 - 6s - loss: 0.0154 - accuracy: 0.9988 - val_loss: 0.3231 - val_accuracy: 0.9346
-Epoch 418/1000
-60000/60000 - 6s - loss: 0.0154 - accuracy: 0.9988 - val_loss: 0.3239 - val_accuracy: 0.9349
-Epoch 419/1000
-60000/60000 - 6s - loss: 0.0154 - accuracy: 0.9987 - val_loss: 0.3228 - val_accuracy: 0.9349
-Epoch 420/1000
-60000/60000 - 6s - loss: 0.0153 - accuracy: 0.9988 - val_loss: 0.3245 - val_accuracy: 0.9354
-Epoch 421/1000
-60000/60000 - 6s - loss: 0.0153 - accuracy: 0.9987 - val_loss: 0.3240 - val_accuracy: 0.9345
-Epoch 422/1000
-60000/60000 - 6s - loss: 0.0152 - accuracy: 0.9988 - val_loss: 0.3243 - val_accuracy: 0.9347
-Epoch 423/1000
-60000/60000 - 6s - loss: 0.0152 - accuracy: 0.9988 - val_loss: 0.3249 - val_accuracy: 0.9339
-Epoch 424/1000
-60000/60000 - 7s - loss: 0.0151 - accuracy: 0.9988 - val_loss: 0.3252 - val_accuracy: 0.9351
-Epoch 425/1000
-60000/60000 - 6s - loss: 0.0151 - accuracy: 0.9988 - val_loss: 0.3248 - val_accuracy: 0.9348
-Epoch 426/1000
-60000/60000 - 6s - loss: 0.0150 - accuracy: 0.9988 - val_loss: 0.3247 - val_accuracy: 0.9337
-Epoch 427/1000
-60000/60000 - 6s - loss: 0.0150 - accuracy: 0.9988 - val_loss: 0.3256 - val_accuracy: 0.9352
-Epoch 428/1000
-60000/60000 - 6s - loss: 0.0150 - accuracy: 0.9988 - val_loss: 0.3261 - val_accuracy: 0.9353
-Epoch 429/1000
-60000/60000 - 7s - loss: 0.0149 - accuracy: 0.9988 - val_loss: 0.3264 - val_accuracy: 0.9348
-Epoch 430/1000
-60000/60000 - 7s - loss: 0.0149 - accuracy: 0.9989 - val_loss: 0.3269 - val_accuracy: 0.9353
-Epoch 431/1000
-60000/60000 - 7s - loss: 0.0148 - accuracy: 0.9988 - val_loss: 0.3275 - val_accuracy: 0.9346
-Epoch 432/1000
-60000/60000 - 7s - loss: 0.0148 - accuracy: 0.9988 - val_loss: 0.3271 - val_accuracy: 0.9345
-Epoch 433/1000
-60000/60000 - 6s - loss: 0.0148 - accuracy: 0.9989 - val_loss: 0.3275 - val_accuracy: 0.9353
-Epoch 434/1000
-60000/60000 - 6s - loss: 0.0147 - accuracy: 0.9988 - val_loss: 0.3280 - val_accuracy: 0.9350
-Epoch 435/1000
-60000/60000 - 7s - loss: 0.0147 - accuracy: 0.9988 - val_loss: 0.3290 - val_accuracy: 0.9342
-Epoch 436/1000
-60000/60000 - 6s - loss: 0.0146 - accuracy: 0.9988 - val_loss: 0.3276 - val_accuracy: 0.9350
-Epoch 437/1000
-60000/60000 - 6s - loss: 0.0146 - accuracy: 0.9989 - val_loss: 0.3283 - val_accuracy: 0.9345
-Epoch 438/1000
-60000/60000 - 6s - loss: 0.0146 - accuracy: 0.9988 - val_loss: 0.3291 - val_accuracy: 0.9348
-Epoch 439/1000
-60000/60000 - 6s - loss: 0.0145 - accuracy: 0.9988 - val_loss: 0.3279 - val_accuracy: 0.9343
-Epoch 440/1000
-60000/60000 - 6s - loss: 0.0145 - accuracy: 0.9989 - val_loss: 0.3291 - val_accuracy: 0.9350
-Epoch 441/1000
-60000/60000 - 6s - loss: 0.0145 - accuracy: 0.9989 - val_loss: 0.3298 - val_accuracy: 0.9347
-Epoch 442/1000
-60000/60000 - 6s - loss: 0.0144 - accuracy: 0.9989 - val_loss: 0.3302 - val_accuracy: 0.9346
-Epoch 443/1000
-60000/60000 - 6s - loss: 0.0143 - accuracy: 0.9989 - val_loss: 0.3309 - val_accuracy: 0.9348
-Epoch 444/1000
-60000/60000 - 6s - loss: 0.0143 - accuracy: 0.9989 - val_loss: 0.3303 - val_accuracy: 0.9346
-Epoch 445/1000
-60000/60000 - 6s - loss: 0.0143 - accuracy: 0.9989 - val_loss: 0.3316 - val_accuracy: 0.9341
-Epoch 446/1000
-60000/60000 - 6s - loss: 0.0142 - accuracy: 0.9989 - val_loss: 0.3306 - val_accuracy: 0.9343
-Epoch 447/1000
-60000/60000 - 6s - loss: 0.0142 - accuracy: 0.9989 - val_loss: 0.3309 - val_accuracy: 0.9352
-Epoch 448/1000
-60000/60000 - 6s - loss: 0.0142 - accuracy: 0.9990 - val_loss: 0.3313 - val_accuracy: 0.9353
-Epoch 449/1000
-60000/60000 - 6s - loss: 0.0141 - accuracy: 0.9989 - val_loss: 0.3329 - val_accuracy: 0.9347
-Epoch 450/1000
-60000/60000 - 6s - loss: 0.0141 - accuracy: 0.9988 - val_loss: 0.3325 - val_accuracy: 0.9352
-Epoch 451/1000
-60000/60000 - 6s - loss: 0.0140 - accuracy: 0.9990 - val_loss: 0.3334 - val_accuracy: 0.9345
-Epoch 452/1000
-60000/60000 - 6s - loss: 0.0140 - accuracy: 0.9990 - val_loss: 0.3338 - val_accuracy: 0.9347
-Epoch 453/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9989 - val_loss: 0.3331 - val_accuracy: 0.9347
-Epoch 454/1000
-60000/60000 - 6s - loss: 0.0140 - accuracy: 0.9990 - val_loss: 0.3344 - val_accuracy: 0.9344
-Epoch 455/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9990 - val_loss: 0.3342 - val_accuracy: 0.9345
-Epoch 456/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9990 - val_loss: 0.3348 - val_accuracy: 0.9345
-Epoch 457/1000
-60000/60000 - 6s - loss: 0.0138 - accuracy: 0.9990 - val_loss: 0.3345 - val_accuracy: 0.9339
-Epoch 458/1000
-60000/60000 - 6s - loss: 0.0138 - accuracy: 0.9990 - val_loss: 0.3347 - val_accuracy: 0.9343
-Epoch 459/1000
-60000/60000 - 6s - loss: 0.0138 - accuracy: 0.9989 - val_loss: 0.3350 - val_accuracy: 0.9341
-Epoch 460/1000
-60000/60000 - 6s - loss: 0.0137 - accuracy: 0.9990 - val_loss: 0.3356 - val_accuracy: 0.9349
-Epoch 461/1000
-60000/60000 - 6s - loss: 0.0137 - accuracy: 0.9990 - val_loss: 0.3363 - val_accuracy: 0.9344
-Epoch 462/1000
-60000/60000 - 6s - loss: 0.0136 - accuracy: 0.9990 - val_loss: 0.3366 - val_accuracy: 0.9346
-Epoch 463/1000
-60000/60000 - 6s - loss: 0.0136 - accuracy: 0.9990 - val_loss: 0.3359 - val_accuracy: 0.9338
-Epoch 464/1000
-60000/60000 - 6s - loss: 0.0136 - accuracy: 0.9991 - val_loss: 0.3361 - val_accuracy: 0.9343
-Epoch 465/1000
-60000/60000 - 6s - loss: 0.0135 - accuracy: 0.9991 - val_loss: 0.3368 - val_accuracy: 0.9346
-Epoch 466/1000
-60000/60000 - 6s - loss: 0.0135 - accuracy: 0.9989 - val_loss: 0.3373 - val_accuracy: 0.9344
-Epoch 467/1000
-60000/60000 - 6s - loss: 0.0135 - accuracy: 0.9991 - val_loss: 0.3369 - val_accuracy: 0.9340
-Epoch 468/1000
-60000/60000 - 6s - loss: 0.0134 - accuracy: 0.9991 - val_loss: 0.3376 - val_accuracy: 0.9347
-Epoch 469/1000
-60000/60000 - 6s - loss: 0.0134 - accuracy: 0.9990 - val_loss: 0.3383 - val_accuracy: 0.9344
-Epoch 470/1000
-60000/60000 - 6s - loss: 0.0134 - accuracy: 0.9991 - val_loss: 0.3378 - val_accuracy: 0.9340
-Epoch 471/1000
-60000/60000 - 6s - loss: 0.0133 - accuracy: 0.9991 - val_loss: 0.3386 - val_accuracy: 0.9336
-Epoch 472/1000
-60000/60000 - 6s - loss: 0.0133 - accuracy: 0.9991 - val_loss: 0.3384 - val_accuracy: 0.9338
-Epoch 473/1000
-60000/60000 - 6s - loss: 0.0133 - accuracy: 0.9991 - val_loss: 0.3386 - val_accuracy: 0.9340
-Epoch 474/1000
-60000/60000 - 6s - loss: 0.0133 - accuracy: 0.9991 - val_loss: 0.3391 - val_accuracy: 0.9340
-Epoch 475/1000
-60000/60000 - 6s - loss: 0.0132 - accuracy: 0.9991 - val_loss: 0.3393 - val_accuracy: 0.9342
-Epoch 476/1000
-60000/60000 - 6s - loss: 0.0131 - accuracy: 0.9991 - val_loss: 0.3400 - val_accuracy: 0.9341
-Epoch 477/1000
-60000/60000 - 6s - loss: 0.0131 - accuracy: 0.9991 - val_loss: 0.3401 - val_accuracy: 0.9341
-Epoch 478/1000
-60000/60000 - 6s - loss: 0.0131 - accuracy: 0.9991 - val_loss: 0.3404 - val_accuracy: 0.9345
-Epoch 479/1000
-60000/60000 - 6s - loss: 0.0131 - accuracy: 0.9991 - val_loss: 0.3405 - val_accuracy: 0.9346
-Epoch 480/1000
-60000/60000 - 6s - loss: 0.0130 - accuracy: 0.9992 - val_loss: 0.3401 - val_accuracy: 0.9340
-Epoch 481/1000
-60000/60000 - 6s - loss: 0.0130 - accuracy: 0.9991 - val_loss: 0.3412 - val_accuracy: 0.9342
-Epoch 482/1000
-60000/60000 - 6s - loss: 0.0130 - accuracy: 0.9991 - val_loss: 0.3402 - val_accuracy: 0.9337
-Epoch 483/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9991 - val_loss: 0.3411 - val_accuracy: 0.9339
-Epoch 484/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9992 - val_loss: 0.3414 - val_accuracy: 0.9350
-Epoch 485/1000
-60000/60000 - 6s - loss: 0.0128 - accuracy: 0.9991 - val_loss: 0.3427 - val_accuracy: 0.9341
-Epoch 486/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9991 - val_loss: 0.3421 - val_accuracy: 0.9339
-Epoch 487/1000
-60000/60000 - 6s - loss: 0.0128 - accuracy: 0.9991 - val_loss: 0.3431 - val_accuracy: 0.9339
-Epoch 488/1000
-60000/60000 - 6s - loss: 0.0128 - accuracy: 0.9991 - val_loss: 0.3424 - val_accuracy: 0.9343
-Epoch 489/1000
-60000/60000 - 6s - loss: 0.0127 - accuracy: 0.9991 - val_loss: 0.3439 - val_accuracy: 0.9338
-Epoch 490/1000
-60000/60000 - 6s - loss: 0.0127 - accuracy: 0.9991 - val_loss: 0.3420 - val_accuracy: 0.9339
-Epoch 491/1000
-60000/60000 - 6s - loss: 0.0127 - accuracy: 0.9991 - val_loss: 0.3436 - val_accuracy: 0.9339
-Epoch 492/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9992 - val_loss: 0.3435 - val_accuracy: 0.9338
-Epoch 493/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9992 - val_loss: 0.3437 - val_accuracy: 0.9346
-Epoch 494/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9992 - val_loss: 0.3437 - val_accuracy: 0.9339
-Epoch 495/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9992 - val_loss: 0.3448 - val_accuracy: 0.9342
-Epoch 496/1000
-60000/60000 - 6s - loss: 0.0125 - accuracy: 0.9991 - val_loss: 0.3449 - val_accuracy: 0.9347
-Epoch 497/1000
-60000/60000 - 6s - loss: 0.0125 - accuracy: 0.9992 - val_loss: 0.3446 - val_accuracy: 0.9338
-Epoch 498/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9992 - val_loss: 0.3457 - val_accuracy: 0.9338
-Epoch 499/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9992 - val_loss: 0.3445 - val_accuracy: 0.9345
-Epoch 500/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9992 - val_loss: 0.3468 - val_accuracy: 0.9340
-Epoch 501/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9992 - val_loss: 0.3458 - val_accuracy: 0.9343
-Epoch 502/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9991 - val_loss: 0.3475 - val_accuracy: 0.9344
-Epoch 503/1000
-60000/60000 - 6s - loss: 0.0123 - accuracy: 0.9992 - val_loss: 0.3466 - val_accuracy: 0.9334
-Epoch 504/1000
-60000/60000 - 6s - loss: 0.0123 - accuracy: 0.9992 - val_loss: 0.3472 - val_accuracy: 0.9338
-Epoch 505/1000
-60000/60000 - 6s - loss: 0.0123 - accuracy: 0.9992 - val_loss: 0.3469 - val_accuracy: 0.9343
-Epoch 506/1000
-60000/60000 - 6s - loss: 0.0122 - accuracy: 0.9992 - val_loss: 0.3481 - val_accuracy: 0.9333
-Epoch 507/1000
-60000/60000 - 6s - loss: 0.0122 - accuracy: 0.9992 - val_loss: 0.3483 - val_accuracy: 0.9341
-Epoch 508/1000
-60000/60000 - 6s - loss: 0.0122 - accuracy: 0.9992 - val_loss: 0.3473 - val_accuracy: 0.9336
-Epoch 509/1000
-60000/60000 - 6s - loss: 0.0122 - accuracy: 0.9992 - val_loss: 0.3482 - val_accuracy: 0.9344
-Epoch 510/1000
-60000/60000 - 6s - loss: 0.0121 - accuracy: 0.9992 - val_loss: 0.3488 - val_accuracy: 0.9339
-Epoch 511/1000
-60000/60000 - 6s - loss: 0.0121 - accuracy: 0.9992 - val_loss: 0.3495 - val_accuracy: 0.9339
-Epoch 512/1000
-60000/60000 - 6s - loss: 0.0121 - accuracy: 0.9992 - val_loss: 0.3493 - val_accuracy: 0.9336
-Epoch 513/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9992 - val_loss: 0.3491 - val_accuracy: 0.9335
-Epoch 514/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9992 - val_loss: 0.3497 - val_accuracy: 0.9336
-Epoch 515/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9992 - val_loss: 0.3488 - val_accuracy: 0.9340
-Epoch 516/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9992 - val_loss: 0.3491 - val_accuracy: 0.9336
-Epoch 517/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9992 - val_loss: 0.3496 - val_accuracy: 0.9342
-Epoch 518/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9992 - val_loss: 0.3502 - val_accuracy: 0.9337
-Epoch 519/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9992 - val_loss: 0.3515 - val_accuracy: 0.9333
-Epoch 520/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9992 - val_loss: 0.3515 - val_accuracy: 0.9339
-Epoch 521/1000
-60000/60000 - 6s - loss: 0.0118 - accuracy: 0.9992 - val_loss: 0.3505 - val_accuracy: 0.9330
-Epoch 522/1000
-60000/60000 - 6s - loss: 0.0118 - accuracy: 0.9992 - val_loss: 0.3521 - val_accuracy: 0.9336
-Epoch 523/1000
-60000/60000 - 6s - loss: 0.0118 - accuracy: 0.9992 - val_loss: 0.3510 - val_accuracy: 0.9336
-Epoch 524/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9993 - val_loss: 0.3516 - val_accuracy: 0.9338
-Epoch 525/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9992 - val_loss: 0.3531 - val_accuracy: 0.9336
-Epoch 526/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9992 - val_loss: 0.3524 - val_accuracy: 0.9339
-Epoch 527/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9992 - val_loss: 0.3527 - val_accuracy: 0.9328
-Epoch 528/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9993 - val_loss: 0.3521 - val_accuracy: 0.9332
-Epoch 529/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9992 - val_loss: 0.3534 - val_accuracy: 0.9337
-Epoch 530/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9992 - val_loss: 0.3534 - val_accuracy: 0.9332
-Epoch 531/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9993 - val_loss: 0.3537 - val_accuracy: 0.9334
-Epoch 532/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9992 - val_loss: 0.3545 - val_accuracy: 0.9336
-Epoch 533/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9993 - val_loss: 0.3537 - val_accuracy: 0.9339
-Epoch 534/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9993 - val_loss: 0.3545 - val_accuracy: 0.9337
-Epoch 535/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9992 - val_loss: 0.3550 - val_accuracy: 0.9342
-Epoch 536/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9992 - val_loss: 0.3547 - val_accuracy: 0.9339
-Epoch 537/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9993 - val_loss: 0.3552 - val_accuracy: 0.9338
-Epoch 538/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9993 - val_loss: 0.3556 - val_accuracy: 0.9339
-Epoch 539/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9993 - val_loss: 0.3570 - val_accuracy: 0.9329
-Epoch 540/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9992 - val_loss: 0.3558 - val_accuracy: 0.9335
-Epoch 541/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9993 - val_loss: 0.3558 - val_accuracy: 0.9338
-Epoch 542/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9993 - val_loss: 0.3569 - val_accuracy: 0.9336
-Epoch 543/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9993 - val_loss: 0.3566 - val_accuracy: 0.9337
-Epoch 544/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9993 - val_loss: 0.3573 - val_accuracy: 0.9337
-Epoch 545/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9993 - val_loss: 0.3573 - val_accuracy: 0.9331
-Epoch 546/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9993 - val_loss: 0.3570 - val_accuracy: 0.9329
-Epoch 547/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9992 - val_loss: 0.3568 - val_accuracy: 0.9334
-Epoch 548/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9993 - val_loss: 0.3576 - val_accuracy: 0.9333
-Epoch 549/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9993 - val_loss: 0.3583 - val_accuracy: 0.9333
-Epoch 550/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9992 - val_loss: 0.3588 - val_accuracy: 0.9334
-Epoch 551/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9993 - val_loss: 0.3595 - val_accuracy: 0.9335
-Epoch 552/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9993 - val_loss: 0.3590 - val_accuracy: 0.9337
-Epoch 553/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9993 - val_loss: 0.3595 - val_accuracy: 0.9329
-Epoch 554/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9993 - val_loss: 0.3601 - val_accuracy: 0.9330
-Epoch 555/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9993 - val_loss: 0.3598 - val_accuracy: 0.9334
-Epoch 556/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9993 - val_loss: 0.3599 - val_accuracy: 0.9337
-Epoch 557/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9993 - val_loss: 0.3593 - val_accuracy: 0.9329
-Epoch 558/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9993 - val_loss: 0.3600 - val_accuracy: 0.9330
-Epoch 559/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9993 - val_loss: 0.3600 - val_accuracy: 0.9333
-Epoch 560/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9993 - val_loss: 0.3604 - val_accuracy: 0.9336
-Epoch 561/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9993 - val_loss: 0.3609 - val_accuracy: 0.9336
-Epoch 562/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9993 - val_loss: 0.3606 - val_accuracy: 0.9334
-Epoch 563/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9993 - val_loss: 0.3612 - val_accuracy: 0.9329
-Epoch 564/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9993 - val_loss: 0.3615 - val_accuracy: 0.9330
-Epoch 565/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9993 - val_loss: 0.3616 - val_accuracy: 0.9328
-Epoch 566/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9993 - val_loss: 0.3620 - val_accuracy: 0.9329
-Epoch 567/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9993 - val_loss: 0.3617 - val_accuracy: 0.9331
-Epoch 568/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9993 - val_loss: 0.3622 - val_accuracy: 0.9328
-Epoch 569/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9993 - val_loss: 0.3637 - val_accuracy: 0.9331
-Epoch 570/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9993 - val_loss: 0.3639 - val_accuracy: 0.9333
-Epoch 571/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9993 - val_loss: 0.3636 - val_accuracy: 0.9327
-Epoch 572/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9993 - val_loss: 0.3632 - val_accuracy: 0.9330
-Epoch 573/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9993 - val_loss: 0.3635 - val_accuracy: 0.9329
-Epoch 574/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9993 - val_loss: 0.3634 - val_accuracy: 0.9333
-Epoch 575/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9993 - val_loss: 0.3636 - val_accuracy: 0.9327
-Epoch 576/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9993 - val_loss: 0.3636 - val_accuracy: 0.9334
-Epoch 577/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9993 - val_loss: 0.3650 - val_accuracy: 0.9329
-Epoch 578/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9993 - val_loss: 0.3642 - val_accuracy: 0.9327
-Epoch 579/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9993 - val_loss: 0.3653 - val_accuracy: 0.9326
-Epoch 580/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9993 - val_loss: 0.3651 - val_accuracy: 0.9321
-Epoch 581/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9993 - val_loss: 0.3654 - val_accuracy: 0.9329
-Epoch 582/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9993 - val_loss: 0.3664 - val_accuracy: 0.9325
-Epoch 583/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9993 - val_loss: 0.3660 - val_accuracy: 0.9322
-Epoch 584/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9993 - val_loss: 0.3661 - val_accuracy: 0.9321
-Epoch 585/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9994 - val_loss: 0.3669 - val_accuracy: 0.9326
-Epoch 586/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9993 - val_loss: 0.3658 - val_accuracy: 0.9321
-Epoch 587/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9994 - val_loss: 0.3672 - val_accuracy: 0.9320
-Epoch 588/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9993 - val_loss: 0.3671 - val_accuracy: 0.9324
-Epoch 589/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9993 - val_loss: 0.3665 - val_accuracy: 0.9321
-Epoch 590/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9993 - val_loss: 0.3672 - val_accuracy: 0.9330
-Epoch 591/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9993 - val_loss: 0.3675 - val_accuracy: 0.9319
-Epoch 592/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9994 - val_loss: 0.3680 - val_accuracy: 0.9327
-Epoch 593/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9993 - val_loss: 0.3687 - val_accuracy: 0.9329
-Epoch 594/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9993 - val_loss: 0.3684 - val_accuracy: 0.9325
-Epoch 595/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9993 - val_loss: 0.3690 - val_accuracy: 0.9324
-Epoch 596/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9994 - val_loss: 0.3686 - val_accuracy: 0.9323
-Epoch 597/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9993 - val_loss: 0.3697 - val_accuracy: 0.9324
-Epoch 598/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9993 - val_loss: 0.3693 - val_accuracy: 0.9323
-Epoch 599/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9994 - val_loss: 0.3700 - val_accuracy: 0.9323
-Epoch 600/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9994 - val_loss: 0.3701 - val_accuracy: 0.9322
-Epoch 601/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9993 - val_loss: 0.3691 - val_accuracy: 0.9324
-Epoch 602/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9993 - val_loss: 0.3700 - val_accuracy: 0.9317
-Epoch 603/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9994 - val_loss: 0.3705 - val_accuracy: 0.9327
-Epoch 604/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9994 - val_loss: 0.3706 - val_accuracy: 0.9322
-Epoch 605/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9994 - val_loss: 0.3709 - val_accuracy: 0.9327
-Epoch 606/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9994 - val_loss: 0.3715 - val_accuracy: 0.9322
-Epoch 607/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9993 - val_loss: 0.3712 - val_accuracy: 0.9326
-Epoch 608/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9993 - val_loss: 0.3714 - val_accuracy: 0.9325
-Epoch 609/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9994 - val_loss: 0.3722 - val_accuracy: 0.9325
-Epoch 610/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9994 - val_loss: 0.3722 - val_accuracy: 0.9322
-Epoch 611/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9994 - val_loss: 0.3722 - val_accuracy: 0.9323
-Epoch 612/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9994 - val_loss: 0.3727 - val_accuracy: 0.9321
-Epoch 613/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9994 - val_loss: 0.3727 - val_accuracy: 0.9318
-Epoch 614/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9994 - val_loss: 0.3725 - val_accuracy: 0.9319
-Epoch 615/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9994 - val_loss: 0.3734 - val_accuracy: 0.9319
-Epoch 616/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9993 - val_loss: 0.3733 - val_accuracy: 0.9326
-Epoch 617/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9994 - val_loss: 0.3737 - val_accuracy: 0.9318
-Epoch 618/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9994 - val_loss: 0.3733 - val_accuracy: 0.9318
-Epoch 619/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9994 - val_loss: 0.3739 - val_accuracy: 0.9325
-Epoch 620/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9994 - val_loss: 0.3739 - val_accuracy: 0.9321
-Epoch 621/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9994 - val_loss: 0.3735 - val_accuracy: 0.9320
-Epoch 622/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9993 - val_loss: 0.3745 - val_accuracy: 0.9323
-Epoch 623/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9994 - val_loss: 0.3744 - val_accuracy: 0.9323
-Epoch 624/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9994 - val_loss: 0.3746 - val_accuracy: 0.9323
-Epoch 625/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9994 - val_loss: 0.3752 - val_accuracy: 0.9318
-Epoch 626/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9994 - val_loss: 0.3758 - val_accuracy: 0.9315
-Epoch 627/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9994 - val_loss: 0.3754 - val_accuracy: 0.9318
-Epoch 628/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9994 - val_loss: 0.3758 - val_accuracy: 0.9321
-Epoch 629/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9994 - val_loss: 0.3764 - val_accuracy: 0.9323
-Epoch 630/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9994 - val_loss: 0.3772 - val_accuracy: 0.9320
-Epoch 631/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9994 - val_loss: 0.3768 - val_accuracy: 0.9316
-Epoch 632/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9994 - val_loss: 0.3764 - val_accuracy: 0.9320
-Epoch 633/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9994 - val_loss: 0.3767 - val_accuracy: 0.9321
-Epoch 634/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9994 - val_loss: 0.3776 - val_accuracy: 0.9322
-Epoch 635/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9994 - val_loss: 0.3772 - val_accuracy: 0.9322
-Epoch 636/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9994 - val_loss: 0.3783 - val_accuracy: 0.9315
-Epoch 637/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9994 - val_loss: 0.3774 - val_accuracy: 0.9323
-Epoch 638/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9994 - val_loss: 0.3776 - val_accuracy: 0.9319
-Epoch 639/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9994 - val_loss: 0.3779 - val_accuracy: 0.9326
-Epoch 640/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9994 - val_loss: 0.3775 - val_accuracy: 0.9319
-Epoch 641/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9994 - val_loss: 0.3790 - val_accuracy: 0.9319
-Epoch 642/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3790 - val_accuracy: 0.9315
-Epoch 643/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3803 - val_accuracy: 0.9315
-Epoch 644/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3785 - val_accuracy: 0.9319
-Epoch 645/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3797 - val_accuracy: 0.9325
-Epoch 646/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3803 - val_accuracy: 0.9315
-Epoch 647/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3795 - val_accuracy: 0.9320
-Epoch 648/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9994 - val_loss: 0.3800 - val_accuracy: 0.9324
-Epoch 649/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3800 - val_accuracy: 0.9318
-Epoch 650/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3810 - val_accuracy: 0.9312
-Epoch 651/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3803 - val_accuracy: 0.9320
-Epoch 652/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3813 - val_accuracy: 0.9315
-Epoch 653/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3811 - val_accuracy: 0.9321
-Epoch 654/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3811 - val_accuracy: 0.9316
-Epoch 655/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9994 - val_loss: 0.3815 - val_accuracy: 0.9321
-Epoch 656/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3808 - val_accuracy: 0.9316
-Epoch 657/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3817 - val_accuracy: 0.9320
-Epoch 658/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3823 - val_accuracy: 0.9318
-Epoch 659/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3823 - val_accuracy: 0.9324
-Epoch 660/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3824 - val_accuracy: 0.9319
-Epoch 661/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3821 - val_accuracy: 0.9319
-Epoch 662/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9994 - val_loss: 0.3824 - val_accuracy: 0.9318
-Epoch 663/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9994 - val_loss: 0.3819 - val_accuracy: 0.9319
-Epoch 664/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9994 - val_loss: 0.3832 - val_accuracy: 0.9319
-Epoch 665/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9994 - val_loss: 0.3838 - val_accuracy: 0.9314
-Epoch 666/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9994 - val_loss: 0.3839 - val_accuracy: 0.9315
-Epoch 667/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9994 - val_loss: 0.3844 - val_accuracy: 0.9313
-Epoch 668/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9994 - val_loss: 0.3852 - val_accuracy: 0.9313
-Epoch 669/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9994 - val_loss: 0.3845 - val_accuracy: 0.9322
-Epoch 670/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9994 - val_loss: 0.3847 - val_accuracy: 0.9313
-Epoch 671/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9994 - val_loss: 0.3846 - val_accuracy: 0.9317
-Epoch 672/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9995 - val_loss: 0.3844 - val_accuracy: 0.9314
-Epoch 673/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9994 - val_loss: 0.3850 - val_accuracy: 0.9316
-Epoch 674/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9995 - val_loss: 0.3848 - val_accuracy: 0.9316
-Epoch 675/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9994 - val_loss: 0.3854 - val_accuracy: 0.9317
-Epoch 676/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9994 - val_loss: 0.3859 - val_accuracy: 0.9314
-Epoch 677/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9995 - val_loss: 0.3857 - val_accuracy: 0.9319
-Epoch 678/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9995 - val_loss: 0.3860 - val_accuracy: 0.9319
-Epoch 679/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9994 - val_loss: 0.3859 - val_accuracy: 0.9314
-Epoch 680/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9994 - val_loss: 0.3857 - val_accuracy: 0.9317
-Epoch 681/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9994 - val_loss: 0.3865 - val_accuracy: 0.9318
-Epoch 682/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9994 - val_loss: 0.3870 - val_accuracy: 0.9317
-Epoch 683/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9995 - val_loss: 0.3874 - val_accuracy: 0.9315
-Epoch 684/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9994 - val_loss: 0.3872 - val_accuracy: 0.9327
-Epoch 685/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9995 - val_loss: 0.3873 - val_accuracy: 0.9319
-Epoch 686/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9995 - val_loss: 0.3872 - val_accuracy: 0.9318
-Epoch 687/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9995 - val_loss: 0.3877 - val_accuracy: 0.9318
-Epoch 688/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9994 - val_loss: 0.3877 - val_accuracy: 0.9318
-Epoch 689/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9995 - val_loss: 0.3882 - val_accuracy: 0.9314
-Epoch 690/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9995 - val_loss: 0.3887 - val_accuracy: 0.9314
-Epoch 691/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9995 - val_loss: 0.3885 - val_accuracy: 0.9323
-Epoch 692/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9995 - val_loss: 0.3899 - val_accuracy: 0.9311
-Epoch 693/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9995 - val_loss: 0.3896 - val_accuracy: 0.9316
-Epoch 694/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9994 - val_loss: 0.3896 - val_accuracy: 0.9313
-Epoch 695/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9995 - val_loss: 0.3898 - val_accuracy: 0.9315
-Epoch 696/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9995 - val_loss: 0.3897 - val_accuracy: 0.9315
-Epoch 697/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9994 - val_loss: 0.3897 - val_accuracy: 0.9313
-Epoch 698/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9995 - val_loss: 0.3903 - val_accuracy: 0.9317
-Epoch 699/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9994 - val_loss: 0.3897 - val_accuracy: 0.9316
-Epoch 700/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9995 - val_loss: 0.3902 - val_accuracy: 0.9315
-Epoch 701/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9995 - val_loss: 0.3903 - val_accuracy: 0.9314
-Epoch 702/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9995 - val_loss: 0.3904 - val_accuracy: 0.9322
-Epoch 703/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9995 - val_loss: 0.3908 - val_accuracy: 0.9317
-Epoch 704/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9994 - val_loss: 0.3909 - val_accuracy: 0.9309
-Epoch 705/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9995 - val_loss: 0.3911 - val_accuracy: 0.9312
-Epoch 706/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9994 - val_loss: 0.3914 - val_accuracy: 0.9315
-Epoch 707/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9995 - val_loss: 0.3916 - val_accuracy: 0.9309
-Epoch 708/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9994 - val_loss: 0.3921 - val_accuracy: 0.9315
-Epoch 709/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9995 - val_loss: 0.3919 - val_accuracy: 0.9311
-Epoch 710/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9995 - val_loss: 0.3926 - val_accuracy: 0.9311
-Epoch 711/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9995 - val_loss: 0.3936 - val_accuracy: 0.9311
-Epoch 712/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9995 - val_loss: 0.3924 - val_accuracy: 0.9313
-Epoch 713/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9994 - val_loss: 0.3931 - val_accuracy: 0.9317
-Epoch 714/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3931 - val_accuracy: 0.9315
-Epoch 715/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3942 - val_accuracy: 0.9311
-Epoch 716/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3937 - val_accuracy: 0.9311
-Epoch 717/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3939 - val_accuracy: 0.9310
-Epoch 718/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3944 - val_accuracy: 0.9315
-Epoch 719/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3946 - val_accuracy: 0.9316
-Epoch 720/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9995 - val_loss: 0.3938 - val_accuracy: 0.9312
-Epoch 721/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3947 - val_accuracy: 0.9313
-Epoch 722/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3945 - val_accuracy: 0.9309
-Epoch 723/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3948 - val_accuracy: 0.9316
-Epoch 724/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3941 - val_accuracy: 0.9318
-Epoch 725/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3958 - val_accuracy: 0.9312
-Epoch 726/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3957 - val_accuracy: 0.9310
-Epoch 727/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3957 - val_accuracy: 0.9313
-Epoch 728/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9995 - val_loss: 0.3958 - val_accuracy: 0.9318
-Epoch 729/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3959 - val_accuracy: 0.9319
-Epoch 730/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3966 - val_accuracy: 0.9311
-Epoch 731/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3960 - val_accuracy: 0.9315
-Epoch 732/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3968 - val_accuracy: 0.9312
-Epoch 733/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3973 - val_accuracy: 0.9312
-Epoch 734/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3973 - val_accuracy: 0.9314
-Epoch 735/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3975 - val_accuracy: 0.9311
-Epoch 736/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3977 - val_accuracy: 0.9313
-Epoch 737/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9995 - val_loss: 0.3972 - val_accuracy: 0.9312
-Epoch 738/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3976 - val_accuracy: 0.9313
-Epoch 739/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3970 - val_accuracy: 0.9312
-Epoch 740/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3982 - val_accuracy: 0.9312
-Epoch 741/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3981 - val_accuracy: 0.9314
-Epoch 742/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3994 - val_accuracy: 0.9314
-Epoch 743/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3986 - val_accuracy: 0.9317
-Epoch 744/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3990 - val_accuracy: 0.9310
-Epoch 745/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3991 - val_accuracy: 0.9309
-Epoch 746/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9995 - val_loss: 0.3998 - val_accuracy: 0.9309
-Epoch 747/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.3989 - val_accuracy: 0.9309
-Epoch 748/1000
-60000/60000 - 7s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.3995 - val_accuracy: 0.9315
-Epoch 749/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.3990 - val_accuracy: 0.9312
-Epoch 750/1000
-60000/60000 - 7s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.4002 - val_accuracy: 0.9308
-Epoch 751/1000
-60000/60000 - 7s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.3997 - val_accuracy: 0.9314
-Epoch 752/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.4006 - val_accuracy: 0.9308
-Epoch 753/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.4008 - val_accuracy: 0.9317
-Epoch 754/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.4004 - val_accuracy: 0.9315
-Epoch 755/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9995 - val_loss: 0.4005 - val_accuracy: 0.9315
-Epoch 756/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4003 - val_accuracy: 0.9313
-Epoch 757/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4017 - val_accuracy: 0.9313
-Epoch 758/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4019 - val_accuracy: 0.9312
-Epoch 759/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4011 - val_accuracy: 0.9313
-Epoch 760/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4019 - val_accuracy: 0.9312
-Epoch 761/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4013 - val_accuracy: 0.9316
-Epoch 762/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4022 - val_accuracy: 0.9314
-Epoch 763/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4014 - val_accuracy: 0.9321
-Epoch 764/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9995 - val_loss: 0.4031 - val_accuracy: 0.9311
-Epoch 765/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4022 - val_accuracy: 0.9315
-Epoch 766/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4029 - val_accuracy: 0.9314
-Epoch 767/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4022 - val_accuracy: 0.9313
-Epoch 768/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4029 - val_accuracy: 0.9313
-Epoch 769/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4025 - val_accuracy: 0.9315
-Epoch 770/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4043 - val_accuracy: 0.9313
-Epoch 771/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4030 - val_accuracy: 0.9318
-Epoch 772/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4037 - val_accuracy: 0.9315
-Epoch 773/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4042 - val_accuracy: 0.9317
-Epoch 774/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9995 - val_loss: 0.4049 - val_accuracy: 0.9311
-Epoch 775/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4039 - val_accuracy: 0.9319
-Epoch 776/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4048 - val_accuracy: 0.9312
-Epoch 777/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4039 - val_accuracy: 0.9316
-Epoch 778/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4046 - val_accuracy: 0.9306
-Epoch 779/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4049 - val_accuracy: 0.9314
-Epoch 780/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4049 - val_accuracy: 0.9308
-Epoch 781/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4055 - val_accuracy: 0.9309
-Epoch 782/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4056 - val_accuracy: 0.9316
-Epoch 783/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4051 - val_accuracy: 0.9311
-Epoch 784/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9995 - val_loss: 0.4061 - val_accuracy: 0.9308
-Epoch 785/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4064 - val_accuracy: 0.9309
-Epoch 786/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4068 - val_accuracy: 0.9314
-Epoch 787/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4073 - val_accuracy: 0.9316
-Epoch 788/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4072 - val_accuracy: 0.9311
-Epoch 789/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4060 - val_accuracy: 0.9311
-Epoch 790/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4063 - val_accuracy: 0.9312
-Epoch 791/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4068 - val_accuracy: 0.9320
-Epoch 792/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4070 - val_accuracy: 0.9313
-Epoch 793/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4072 - val_accuracy: 0.9312
-Epoch 794/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9995 - val_loss: 0.4073 - val_accuracy: 0.9311
-Epoch 795/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4071 - val_accuracy: 0.9312
-Epoch 796/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4079 - val_accuracy: 0.9310
-Epoch 797/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4084 - val_accuracy: 0.9308
-Epoch 798/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4077 - val_accuracy: 0.9311
-Epoch 799/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4074 - val_accuracy: 0.9316
-Epoch 800/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4084 - val_accuracy: 0.9309
-Epoch 801/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4087 - val_accuracy: 0.9309
-Epoch 802/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4091 - val_accuracy: 0.9311
-Epoch 803/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4090 - val_accuracy: 0.9308
-Epoch 804/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9995 - val_loss: 0.4090 - val_accuracy: 0.9307
-Epoch 805/1000
-60000/60000 - 7s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4099 - val_accuracy: 0.9316
-Epoch 806/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4095 - val_accuracy: 0.9311
-Epoch 807/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4097 - val_accuracy: 0.9313
-Epoch 808/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4100 - val_accuracy: 0.9313
-Epoch 809/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4104 - val_accuracy: 0.9308
-Epoch 810/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4103 - val_accuracy: 0.9310
-Epoch 811/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4112 - val_accuracy: 0.9308
-Epoch 812/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4105 - val_accuracy: 0.9313
-Epoch 813/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4107 - val_accuracy: 0.9307
-Epoch 814/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4114 - val_accuracy: 0.9313
-Epoch 815/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4115 - val_accuracy: 0.9308
-Epoch 816/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9995 - val_loss: 0.4113 - val_accuracy: 0.9311
-Epoch 817/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4113 - val_accuracy: 0.9316
-Epoch 818/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9996 - val_loss: 0.4119 - val_accuracy: 0.9311
-Epoch 819/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4127 - val_accuracy: 0.9307
-Epoch 820/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4123 - val_accuracy: 0.9311
-Epoch 821/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4125 - val_accuracy: 0.9310
-Epoch 822/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9996 - val_loss: 0.4122 - val_accuracy: 0.9313
-Epoch 823/1000
-60000/60000 - 7s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4124 - val_accuracy: 0.9310
-Epoch 824/1000
-60000/60000 - 7s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4127 - val_accuracy: 0.9311
-Epoch 825/1000
-60000/60000 - 7s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4126 - val_accuracy: 0.9311
-Epoch 826/1000
-60000/60000 - 7s - loss: 0.0072 - accuracy: 0.9995 - val_loss: 0.4129 - val_accuracy: 0.9310
-Epoch 827/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4124 - val_accuracy: 0.9314
-Epoch 828/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4134 - val_accuracy: 0.9313
-Epoch 829/1000
-60000/60000 - 7s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4137 - val_accuracy: 0.9314
-Epoch 830/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9996 - val_loss: 0.4134 - val_accuracy: 0.9306
-Epoch 831/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4135 - val_accuracy: 0.9313
-Epoch 832/1000
-60000/60000 - 7s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4143 - val_accuracy: 0.9305
-Epoch 833/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4143 - val_accuracy: 0.9311
-Epoch 834/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4146 - val_accuracy: 0.9310
-Epoch 835/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4148 - val_accuracy: 0.9314
-Epoch 836/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4152 - val_accuracy: 0.9310
-Epoch 837/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4146 - val_accuracy: 0.9308
-Epoch 838/1000
-60000/60000 - 7s - loss: 0.0071 - accuracy: 0.9995 - val_loss: 0.4150 - val_accuracy: 0.9311
-Epoch 839/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4150 - val_accuracy: 0.9307
-Epoch 840/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4152 - val_accuracy: 0.9313
-Epoch 841/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9996 - val_loss: 0.4160 - val_accuracy: 0.9313
-Epoch 842/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9996 - val_loss: 0.4158 - val_accuracy: 0.9312
-Epoch 843/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4160 - val_accuracy: 0.9307
-Epoch 844/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4157 - val_accuracy: 0.9313
-Epoch 845/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4166 - val_accuracy: 0.9310
-Epoch 846/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9996 - val_loss: 0.4166 - val_accuracy: 0.9310
-Epoch 847/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4161 - val_accuracy: 0.9314
-Epoch 848/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4169 - val_accuracy: 0.9308
-Epoch 849/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9995 - val_loss: 0.4168 - val_accuracy: 0.9311
-Epoch 850/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4168 - val_accuracy: 0.9306
-Epoch 851/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4171 - val_accuracy: 0.9309
-Epoch 852/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9996 - val_loss: 0.4173 - val_accuracy: 0.9314
-Epoch 853/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4171 - val_accuracy: 0.9313
-Epoch 854/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4177 - val_accuracy: 0.9307
-Epoch 855/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4179 - val_accuracy: 0.9309
-Epoch 856/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4178 - val_accuracy: 0.9310
-Epoch 857/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9996 - val_loss: 0.4179 - val_accuracy: 0.9312
-Epoch 858/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9996 - val_loss: 0.4185 - val_accuracy: 0.9311
-Epoch 859/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9995 - val_loss: 0.4179 - val_accuracy: 0.9309
-Epoch 860/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9996 - val_loss: 0.4185 - val_accuracy: 0.9312
-Epoch 861/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4195 - val_accuracy: 0.9308
-Epoch 862/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4201 - val_accuracy: 0.9312
-Epoch 863/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9995 - val_loss: 0.4184 - val_accuracy: 0.9312
-Epoch 864/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4193 - val_accuracy: 0.9313
-Epoch 865/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4191 - val_accuracy: 0.9310
-Epoch 866/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9995 - val_loss: 0.4196 - val_accuracy: 0.9312
-Epoch 867/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4205 - val_accuracy: 0.9309
-Epoch 868/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9995 - val_loss: 0.4197 - val_accuracy: 0.9312
-Epoch 869/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4203 - val_accuracy: 0.9313
-Epoch 870/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9995 - val_loss: 0.4201 - val_accuracy: 0.9311
-Epoch 871/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4210 - val_accuracy: 0.9315
-Epoch 872/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4199 - val_accuracy: 0.9314
-Epoch 873/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4208 - val_accuracy: 0.9310
-Epoch 874/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4209 - val_accuracy: 0.9310
-Epoch 875/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9996 - val_loss: 0.4218 - val_accuracy: 0.9315
-Epoch 876/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4216 - val_accuracy: 0.9311
-Epoch 877/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9995 - val_loss: 0.4209 - val_accuracy: 0.9308
-Epoch 878/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4218 - val_accuracy: 0.9312
-Epoch 879/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4216 - val_accuracy: 0.9308
-Epoch 880/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4220 - val_accuracy: 0.9310
-Epoch 881/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4219 - val_accuracy: 0.9310
-Epoch 882/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9995 - val_loss: 0.4226 - val_accuracy: 0.9313
-Epoch 883/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4222 - val_accuracy: 0.9313
-Epoch 884/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4224 - val_accuracy: 0.9312
-Epoch 885/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9995 - val_loss: 0.4227 - val_accuracy: 0.9311
-Epoch 886/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4227 - val_accuracy: 0.9309
-Epoch 887/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9996 - val_loss: 0.4234 - val_accuracy: 0.9306
-Epoch 888/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4226 - val_accuracy: 0.9310
-Epoch 889/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4235 - val_accuracy: 0.9312
-Epoch 890/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4229 - val_accuracy: 0.9312
-Epoch 891/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4233 - val_accuracy: 0.9310
-Epoch 892/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4242 - val_accuracy: 0.9307
-Epoch 893/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4234 - val_accuracy: 0.9315
-Epoch 894/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4248 - val_accuracy: 0.9311
-Epoch 895/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4238 - val_accuracy: 0.9314
-Epoch 896/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4246 - val_accuracy: 0.9308
-Epoch 897/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4249 - val_accuracy: 0.9304
-Epoch 898/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4244 - val_accuracy: 0.9310
-Epoch 899/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4253 - val_accuracy: 0.9308
-Epoch 900/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4248 - val_accuracy: 0.9308
-Epoch 901/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9996 - val_loss: 0.4247 - val_accuracy: 0.9307
-Epoch 902/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4252 - val_accuracy: 0.9303
-Epoch 903/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4254 - val_accuracy: 0.9306
-Epoch 904/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4254 - val_accuracy: 0.9312
-Epoch 905/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4259 - val_accuracy: 0.9310
-Epoch 906/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4260 - val_accuracy: 0.9310
-Epoch 907/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4258 - val_accuracy: 0.9307
-Epoch 908/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4266 - val_accuracy: 0.9310
-Epoch 909/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4262 - val_accuracy: 0.9307
-Epoch 910/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4264 - val_accuracy: 0.9310
-Epoch 911/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4268 - val_accuracy: 0.9311
-Epoch 912/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4267 - val_accuracy: 0.9312
-Epoch 913/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4278 - val_accuracy: 0.9309
-Epoch 914/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9996 - val_loss: 0.4268 - val_accuracy: 0.9311
-Epoch 915/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4271 - val_accuracy: 0.9314
-Epoch 916/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4272 - val_accuracy: 0.9308
-Epoch 917/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4278 - val_accuracy: 0.9308
-Epoch 918/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4277 - val_accuracy: 0.9310
-Epoch 919/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4282 - val_accuracy: 0.9309
-Epoch 920/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4283 - val_accuracy: 0.9313
-Epoch 921/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4280 - val_accuracy: 0.9310
-Epoch 922/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4284 - val_accuracy: 0.9310
-Epoch 923/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4292 - val_accuracy: 0.9308
-Epoch 924/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4291 - val_accuracy: 0.9308
-Epoch 925/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4291 - val_accuracy: 0.9312
-Epoch 926/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4288 - val_accuracy: 0.9309
-Epoch 927/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4292 - val_accuracy: 0.9309
-Epoch 928/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4291 - val_accuracy: 0.9311
-Epoch 929/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9996 - val_loss: 0.4296 - val_accuracy: 0.9312
-Epoch 930/1000
-60000/60000 - 7s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4302 - val_accuracy: 0.9308
-Epoch 931/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4299 - val_accuracy: 0.9309
-Epoch 932/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4298 - val_accuracy: 0.9306
-Epoch 933/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4299 - val_accuracy: 0.9307
-Epoch 934/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4294 - val_accuracy: 0.9312
-Epoch 935/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4305 - val_accuracy: 0.9306
-Epoch 936/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4311 - val_accuracy: 0.9306
-Epoch 937/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4311 - val_accuracy: 0.9306
-Epoch 938/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4312 - val_accuracy: 0.9309
-Epoch 939/1000
-60000/60000 - 7s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4309 - val_accuracy: 0.9310
-Epoch 940/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4313 - val_accuracy: 0.9307
-Epoch 941/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4314 - val_accuracy: 0.9310
-Epoch 942/1000
-60000/60000 - 7s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4311 - val_accuracy: 0.9308
-Epoch 943/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4314 - val_accuracy: 0.9309
-Epoch 944/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4319 - val_accuracy: 0.9310
-Epoch 945/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9996 - val_loss: 0.4319 - val_accuracy: 0.9310
-Epoch 946/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4319 - val_accuracy: 0.9311
-Epoch 947/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4327 - val_accuracy: 0.9313
-Epoch 948/1000
-60000/60000 - 7s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4322 - val_accuracy: 0.9308
-Epoch 949/1000
-60000/60000 - 7s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4325 - val_accuracy: 0.9309
-Epoch 950/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4320 - val_accuracy: 0.9310
-Epoch 951/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4327 - val_accuracy: 0.9313
-Epoch 952/1000
-60000/60000 - 7s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4331 - val_accuracy: 0.9310
-Epoch 953/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4325 - val_accuracy: 0.9310
-Epoch 954/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4328 - val_accuracy: 0.9314
-Epoch 955/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4335 - val_accuracy: 0.9307
-Epoch 956/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4337 - val_accuracy: 0.9309
-Epoch 957/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4339 - val_accuracy: 0.9304
-Epoch 958/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4332 - val_accuracy: 0.9307
-Epoch 959/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9996 - val_loss: 0.4342 - val_accuracy: 0.9310
-Epoch 960/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4340 - val_accuracy: 0.9307
-Epoch 961/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4344 - val_accuracy: 0.9302
-Epoch 962/1000
-60000/60000 - 7s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4344 - val_accuracy: 0.9313
-Epoch 963/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4347 - val_accuracy: 0.9309
-Epoch 964/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4349 - val_accuracy: 0.9311
-Epoch 965/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4348 - val_accuracy: 0.9312
-Epoch 966/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4356 - val_accuracy: 0.9308
-Epoch 967/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4348 - val_accuracy: 0.9311
-Epoch 968/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4354 - val_accuracy: 0.9315
-Epoch 969/1000
-60000/60000 - 7s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4359 - val_accuracy: 0.9311
-Epoch 970/1000
-60000/60000 - 7s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4349 - val_accuracy: 0.9308
-Epoch 971/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4357 - val_accuracy: 0.9312
-Epoch 972/1000
-60000/60000 - 7s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4362 - val_accuracy: 0.9310
-Epoch 973/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4361 - val_accuracy: 0.9312
-Epoch 974/1000
-60000/60000 - 7s - loss: 0.0061 - accuracy: 0.9996 - val_loss: 0.4359 - val_accuracy: 0.9309
-Epoch 975/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4367 - val_accuracy: 0.9307
-Epoch 976/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4363 - val_accuracy: 0.9317
-Epoch 977/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4363 - val_accuracy: 0.9312
-Epoch 978/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4368 - val_accuracy: 0.9309
-Epoch 979/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4371 - val_accuracy: 0.9307
-Epoch 980/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4368 - val_accuracy: 0.9310
-Epoch 981/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4376 - val_accuracy: 0.9307
-Epoch 982/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4374 - val_accuracy: 0.9310
-Epoch 983/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4370 - val_accuracy: 0.9314
-Epoch 984/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4373 - val_accuracy: 0.9309
-Epoch 985/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4377 - val_accuracy: 0.9306
-Epoch 986/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4382 - val_accuracy: 0.9313
-Epoch 987/1000
-60000/60000 - 7s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4383 - val_accuracy: 0.9313
-Epoch 988/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9996 - val_loss: 0.4376 - val_accuracy: 0.9309
-Epoch 989/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4380 - val_accuracy: 0.9305
-Epoch 990/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4380 - val_accuracy: 0.9312
-Epoch 991/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4384 - val_accuracy: 0.9313
-Epoch 992/1000
-60000/60000 - 7s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4389 - val_accuracy: 0.9304
-Epoch 993/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4381 - val_accuracy: 0.9315
-Epoch 994/1000
-60000/60000 - 7s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4390 - val_accuracy: 0.9310
-Epoch 995/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4388 - val_accuracy: 0.9311
-Epoch 996/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4392 - val_accuracy: 0.9311
-Epoch 997/1000
-60000/60000 - 7s - loss: 0.0059 - accuracy: 0.9996 - val_loss: 0.4393 - val_accuracy: 0.9309
-Epoch 998/1000
-60000/60000 - 7s - loss: 0.0058 - accuracy: 0.9996 - val_loss: 0.4391 - val_accuracy: 0.9312
-Epoch 999/1000
-60000/60000 - 7s - loss: 0.0058 - accuracy: 0.9996 - val_loss: 0.4398 - val_accuracy: 0.9309
-Epoch 1000/1000
-60000/60000 - 7s - loss: 0.0058 - accuracy: 0.9996 - val_loss: 0.4402 - val_accuracy: 0.9313
-Test loss was 0.1972, test accuracy was 0.9472
diff --git a/MNIST/nonlinear_withsigmoid_pnweights.txt b/MNIST/nonlinear_withsigmoid_pnweights.txt
deleted file mode 100644
index 73a54ee..0000000
--- a/MNIST/nonlinear_withsigmoid_pnweights.txt
+++ /dev/null
@@ -1,2005 +0,0 @@
-X_train shape: (60000, 784)
-60000 train samples
-10000 test samples
-Train on 60000 samples, validate on 10000 samples
-Epoch 1/1000
-60000/60000 - 9s - loss: 0.5901 - accuracy: 0.8227 - val_loss: 0.3636 - val_accuracy: 0.8920
-Epoch 2/1000
-60000/60000 - 6s - loss: 0.3389 - accuracy: 0.9004 - val_loss: 0.3002 - val_accuracy: 0.9129
-Epoch 3/1000
-60000/60000 - 6s - loss: 0.2826 - accuracy: 0.9175 - val_loss: 0.2706 - val_accuracy: 0.9213
-Epoch 4/1000
-60000/60000 - 6s - loss: 0.2505 - accuracy: 0.9276 - val_loss: 0.2507 - val_accuracy: 0.9272
-Epoch 5/1000
-60000/60000 - 6s - loss: 0.2284 - accuracy: 0.9330 - val_loss: 0.2375 - val_accuracy: 0.9328
-Epoch 6/1000
-60000/60000 - 6s - loss: 0.2117 - accuracy: 0.9392 - val_loss: 0.2283 - val_accuracy: 0.9327
-Epoch 7/1000
-60000/60000 - 6s - loss: 0.1986 - accuracy: 0.9422 - val_loss: 0.2207 - val_accuracy: 0.9369
-Epoch 8/1000
-60000/60000 - 6s - loss: 0.1876 - accuracy: 0.9456 - val_loss: 0.2144 - val_accuracy: 0.9371
-Epoch 9/1000
-60000/60000 - 6s - loss: 0.1786 - accuracy: 0.9487 - val_loss: 0.2112 - val_accuracy: 0.9377
-Epoch 10/1000
-60000/60000 - 6s - loss: 0.1705 - accuracy: 0.9511 - val_loss: 0.2048 - val_accuracy: 0.9399
-Epoch 11/1000
-60000/60000 - 6s - loss: 0.1637 - accuracy: 0.9538 - val_loss: 0.2026 - val_accuracy: 0.9420
-Epoch 12/1000
-60000/60000 - 6s - loss: 0.1578 - accuracy: 0.9554 - val_loss: 0.2030 - val_accuracy: 0.9399
-Epoch 13/1000
-60000/60000 - 6s - loss: 0.1518 - accuracy: 0.9566 - val_loss: 0.1977 - val_accuracy: 0.9409
-Epoch 14/1000
-60000/60000 - 6s - loss: 0.1468 - accuracy: 0.9578 - val_loss: 0.1965 - val_accuracy: 0.9417
-Epoch 15/1000
-60000/60000 - 6s - loss: 0.1424 - accuracy: 0.9596 - val_loss: 0.1965 - val_accuracy: 0.9424
-Epoch 16/1000
-60000/60000 - 6s - loss: 0.1381 - accuracy: 0.9609 - val_loss: 0.1953 - val_accuracy: 0.9416
-Epoch 17/1000
-60000/60000 - 6s - loss: 0.1340 - accuracy: 0.9620 - val_loss: 0.1927 - val_accuracy: 0.9429
-Epoch 18/1000
-60000/60000 - 6s - loss: 0.1305 - accuracy: 0.9632 - val_loss: 0.1912 - val_accuracy: 0.9436
-Epoch 19/1000
-60000/60000 - 6s - loss: 0.1269 - accuracy: 0.9636 - val_loss: 0.1908 - val_accuracy: 0.9447
-Epoch 20/1000
-60000/60000 - 6s - loss: 0.1236 - accuracy: 0.9654 - val_loss: 0.1898 - val_accuracy: 0.9444
-Epoch 21/1000
-60000/60000 - 6s - loss: 0.1207 - accuracy: 0.9658 - val_loss: 0.1908 - val_accuracy: 0.9454
-Epoch 22/1000
-60000/60000 - 6s - loss: 0.1179 - accuracy: 0.9664 - val_loss: 0.1896 - val_accuracy: 0.9452
-Epoch 23/1000
-60000/60000 - 6s - loss: 0.1152 - accuracy: 0.9678 - val_loss: 0.1914 - val_accuracy: 0.9450
-Epoch 24/1000
-60000/60000 - 6s - loss: 0.1125 - accuracy: 0.9684 - val_loss: 0.1908 - val_accuracy: 0.9436
-Epoch 25/1000
-60000/60000 - 6s - loss: 0.1103 - accuracy: 0.9685 - val_loss: 0.1903 - val_accuracy: 0.9462
-Epoch 26/1000
-60000/60000 - 6s - loss: 0.1079 - accuracy: 0.9697 - val_loss: 0.1893 - val_accuracy: 0.9461
-Epoch 27/1000
-60000/60000 - 6s - loss: 0.1053 - accuracy: 0.9710 - val_loss: 0.1942 - val_accuracy: 0.9446
-Epoch 28/1000
-60000/60000 - 6s - loss: 0.1036 - accuracy: 0.9709 - val_loss: 0.1919 - val_accuracy: 0.9436
-Epoch 29/1000
-60000/60000 - 6s - loss: 0.1011 - accuracy: 0.9718 - val_loss: 0.1925 - val_accuracy: 0.9439
-Epoch 30/1000
-60000/60000 - 6s - loss: 0.0998 - accuracy: 0.9722 - val_loss: 0.1942 - val_accuracy: 0.9442
-Epoch 31/1000
-60000/60000 - 6s - loss: 0.0976 - accuracy: 0.9731 - val_loss: 0.1924 - val_accuracy: 0.9453
-Epoch 32/1000
-60000/60000 - 6s - loss: 0.0959 - accuracy: 0.9735 - val_loss: 0.1951 - val_accuracy: 0.9439
-Epoch 33/1000
-60000/60000 - 6s - loss: 0.0942 - accuracy: 0.9742 - val_loss: 0.1923 - val_accuracy: 0.9450
-Epoch 34/1000
-60000/60000 - 6s - loss: 0.0926 - accuracy: 0.9745 - val_loss: 0.1937 - val_accuracy: 0.9453
-Epoch 35/1000
-60000/60000 - 6s - loss: 0.0912 - accuracy: 0.9748 - val_loss: 0.1901 - val_accuracy: 0.9459
-Epoch 36/1000
-60000/60000 - 6s - loss: 0.0896 - accuracy: 0.9755 - val_loss: 0.1929 - val_accuracy: 0.9449
-Epoch 37/1000
-60000/60000 - 6s - loss: 0.0882 - accuracy: 0.9761 - val_loss: 0.1937 - val_accuracy: 0.9457
-Epoch 38/1000
-60000/60000 - 6s - loss: 0.0870 - accuracy: 0.9762 - val_loss: 0.1930 - val_accuracy: 0.9450
-Epoch 39/1000
-60000/60000 - 6s - loss: 0.0856 - accuracy: 0.9766 - val_loss: 0.1956 - val_accuracy: 0.9450
-Epoch 40/1000
-60000/60000 - 6s - loss: 0.0843 - accuracy: 0.9771 - val_loss: 0.1949 - val_accuracy: 0.9446
-Epoch 41/1000
-60000/60000 - 6s - loss: 0.0831 - accuracy: 0.9773 - val_loss: 0.2003 - val_accuracy: 0.9432
-Epoch 42/1000
-60000/60000 - 6s - loss: 0.0818 - accuracy: 0.9783 - val_loss: 0.1943 - val_accuracy: 0.9447
-Epoch 43/1000
-60000/60000 - 6s - loss: 0.0804 - accuracy: 0.9786 - val_loss: 0.1968 - val_accuracy: 0.9443
-Epoch 44/1000
-60000/60000 - 6s - loss: 0.0794 - accuracy: 0.9791 - val_loss: 0.1964 - val_accuracy: 0.9450
-Epoch 45/1000
-60000/60000 - 6s - loss: 0.0782 - accuracy: 0.9790 - val_loss: 0.1993 - val_accuracy: 0.9440
-Epoch 46/1000
-60000/60000 - 6s - loss: 0.0769 - accuracy: 0.9794 - val_loss: 0.2010 - val_accuracy: 0.9434
-Epoch 47/1000
-60000/60000 - 6s - loss: 0.0759 - accuracy: 0.9801 - val_loss: 0.1977 - val_accuracy: 0.9438
-Epoch 48/1000
-60000/60000 - 6s - loss: 0.0751 - accuracy: 0.9801 - val_loss: 0.1995 - val_accuracy: 0.9435
-Epoch 49/1000
-60000/60000 - 6s - loss: 0.0739 - accuracy: 0.9808 - val_loss: 0.2001 - val_accuracy: 0.9433
-Epoch 50/1000
-60000/60000 - 6s - loss: 0.0732 - accuracy: 0.9804 - val_loss: 0.2018 - val_accuracy: 0.9427
-Epoch 51/1000
-60000/60000 - 6s - loss: 0.0718 - accuracy: 0.9815 - val_loss: 0.2047 - val_accuracy: 0.9427
-Epoch 52/1000
-60000/60000 - 6s - loss: 0.0711 - accuracy: 0.9814 - val_loss: 0.2048 - val_accuracy: 0.9422
-Epoch 53/1000
-60000/60000 - 6s - loss: 0.0702 - accuracy: 0.9820 - val_loss: 0.2052 - val_accuracy: 0.9415
-Epoch 54/1000
-60000/60000 - 6s - loss: 0.0693 - accuracy: 0.9817 - val_loss: 0.2054 - val_accuracy: 0.9427
-Epoch 55/1000
-60000/60000 - 6s - loss: 0.0683 - accuracy: 0.9823 - val_loss: 0.2058 - val_accuracy: 0.9430
-Epoch 56/1000
-60000/60000 - 6s - loss: 0.0677 - accuracy: 0.9827 - val_loss: 0.2076 - val_accuracy: 0.9434
-Epoch 57/1000
-60000/60000 - 6s - loss: 0.0667 - accuracy: 0.9827 - val_loss: 0.2078 - val_accuracy: 0.9428
-Epoch 58/1000
-60000/60000 - 6s - loss: 0.0660 - accuracy: 0.9830 - val_loss: 0.2092 - val_accuracy: 0.9415
-Epoch 59/1000
-60000/60000 - 6s - loss: 0.0652 - accuracy: 0.9834 - val_loss: 0.2085 - val_accuracy: 0.9424
-Epoch 60/1000
-60000/60000 - 6s - loss: 0.0645 - accuracy: 0.9837 - val_loss: 0.2099 - val_accuracy: 0.9428
-Epoch 61/1000
-60000/60000 - 6s - loss: 0.0636 - accuracy: 0.9840 - val_loss: 0.2121 - val_accuracy: 0.9420
-Epoch 62/1000
-60000/60000 - 6s - loss: 0.0629 - accuracy: 0.9844 - val_loss: 0.2128 - val_accuracy: 0.9419
-Epoch 63/1000
-60000/60000 - 6s - loss: 0.0618 - accuracy: 0.9847 - val_loss: 0.2100 - val_accuracy: 0.9427
-Epoch 64/1000
-60000/60000 - 6s - loss: 0.0616 - accuracy: 0.9847 - val_loss: 0.2118 - val_accuracy: 0.9422
-Epoch 65/1000
-60000/60000 - 6s - loss: 0.0608 - accuracy: 0.9853 - val_loss: 0.2161 - val_accuracy: 0.9415
-Epoch 66/1000
-60000/60000 - 6s - loss: 0.0602 - accuracy: 0.9850 - val_loss: 0.2135 - val_accuracy: 0.9414
-Epoch 67/1000
-60000/60000 - 6s - loss: 0.0593 - accuracy: 0.9856 - val_loss: 0.2153 - val_accuracy: 0.9411
-Epoch 68/1000
-60000/60000 - 6s - loss: 0.0590 - accuracy: 0.9854 - val_loss: 0.2167 - val_accuracy: 0.9422
-Epoch 69/1000
-60000/60000 - 6s - loss: 0.0584 - accuracy: 0.9855 - val_loss: 0.2169 - val_accuracy: 0.9417
-Epoch 70/1000
-60000/60000 - 6s - loss: 0.0575 - accuracy: 0.9860 - val_loss: 0.2182 - val_accuracy: 0.9415
-Epoch 71/1000
-60000/60000 - 6s - loss: 0.0569 - accuracy: 0.9863 - val_loss: 0.2172 - val_accuracy: 0.9422
-Epoch 72/1000
-60000/60000 - 6s - loss: 0.0561 - accuracy: 0.9871 - val_loss: 0.2200 - val_accuracy: 0.9419
-Epoch 73/1000
-60000/60000 - 6s - loss: 0.0555 - accuracy: 0.9871 - val_loss: 0.2217 - val_accuracy: 0.9412
-Epoch 74/1000
-60000/60000 - 6s - loss: 0.0549 - accuracy: 0.9868 - val_loss: 0.2209 - val_accuracy: 0.9427
-Epoch 75/1000
-60000/60000 - 6s - loss: 0.0547 - accuracy: 0.9873 - val_loss: 0.2214 - val_accuracy: 0.9402
-Epoch 76/1000
-60000/60000 - 6s - loss: 0.0537 - accuracy: 0.9872 - val_loss: 0.2229 - val_accuracy: 0.9405
-Epoch 77/1000
-60000/60000 - 6s - loss: 0.0531 - accuracy: 0.9878 - val_loss: 0.2218 - val_accuracy: 0.9414
-Epoch 78/1000
-60000/60000 - 6s - loss: 0.0527 - accuracy: 0.9877 - val_loss: 0.2253 - val_accuracy: 0.9412
-Epoch 79/1000
-60000/60000 - 6s - loss: 0.0521 - accuracy: 0.9879 - val_loss: 0.2252 - val_accuracy: 0.9404
-Epoch 80/1000
-60000/60000 - 6s - loss: 0.0520 - accuracy: 0.9878 - val_loss: 0.2267 - val_accuracy: 0.9409
-Epoch 81/1000
-60000/60000 - 6s - loss: 0.0513 - accuracy: 0.9880 - val_loss: 0.2256 - val_accuracy: 0.9409
-Epoch 82/1000
-60000/60000 - 6s - loss: 0.0506 - accuracy: 0.9885 - val_loss: 0.2259 - val_accuracy: 0.9414
-Epoch 83/1000
-60000/60000 - 6s - loss: 0.0503 - accuracy: 0.9887 - val_loss: 0.2289 - val_accuracy: 0.9406
-Epoch 84/1000
-60000/60000 - 6s - loss: 0.0498 - accuracy: 0.9891 - val_loss: 0.2266 - val_accuracy: 0.9408
-Epoch 85/1000
-60000/60000 - 6s - loss: 0.0493 - accuracy: 0.9886 - val_loss: 0.2293 - val_accuracy: 0.9410
-Epoch 86/1000
-60000/60000 - 6s - loss: 0.0487 - accuracy: 0.9889 - val_loss: 0.2309 - val_accuracy: 0.9408
-Epoch 87/1000
-60000/60000 - 6s - loss: 0.0483 - accuracy: 0.9889 - val_loss: 0.2366 - val_accuracy: 0.9390
-Epoch 88/1000
-60000/60000 - 6s - loss: 0.0478 - accuracy: 0.9894 - val_loss: 0.2317 - val_accuracy: 0.9399
-Epoch 89/1000
-60000/60000 - 6s - loss: 0.0475 - accuracy: 0.9894 - val_loss: 0.2347 - val_accuracy: 0.9402
-Epoch 90/1000
-60000/60000 - 6s - loss: 0.0471 - accuracy: 0.9898 - val_loss: 0.2342 - val_accuracy: 0.9401
-Epoch 91/1000
-60000/60000 - 6s - loss: 0.0465 - accuracy: 0.9899 - val_loss: 0.2350 - val_accuracy: 0.9405
-Epoch 92/1000
-60000/60000 - 6s - loss: 0.0460 - accuracy: 0.9901 - val_loss: 0.2354 - val_accuracy: 0.9411
-Epoch 93/1000
-60000/60000 - 6s - loss: 0.0456 - accuracy: 0.9901 - val_loss: 0.2370 - val_accuracy: 0.9404
-Epoch 94/1000
-60000/60000 - 6s - loss: 0.0452 - accuracy: 0.9901 - val_loss: 0.2368 - val_accuracy: 0.9392
-Epoch 95/1000
-60000/60000 - 6s - loss: 0.0448 - accuracy: 0.9905 - val_loss: 0.2369 - val_accuracy: 0.9403
-Epoch 96/1000
-60000/60000 - 6s - loss: 0.0444 - accuracy: 0.9906 - val_loss: 0.2380 - val_accuracy: 0.9404
-Epoch 97/1000
-60000/60000 - 6s - loss: 0.0442 - accuracy: 0.9904 - val_loss: 0.2397 - val_accuracy: 0.9402
-Epoch 98/1000
-60000/60000 - 6s - loss: 0.0438 - accuracy: 0.9907 - val_loss: 0.2405 - val_accuracy: 0.9397
-Epoch 99/1000
-60000/60000 - 6s - loss: 0.0434 - accuracy: 0.9907 - val_loss: 0.2444 - val_accuracy: 0.9391
-Epoch 100/1000
-60000/60000 - 6s - loss: 0.0429 - accuracy: 0.9913 - val_loss: 0.2402 - val_accuracy: 0.9394
-Epoch 101/1000
-60000/60000 - 6s - loss: 0.0429 - accuracy: 0.9908 - val_loss: 0.2425 - val_accuracy: 0.9401
-Epoch 102/1000
-60000/60000 - 6s - loss: 0.0421 - accuracy: 0.9915 - val_loss: 0.2423 - val_accuracy: 0.9399
-Epoch 103/1000
-60000/60000 - 6s - loss: 0.0419 - accuracy: 0.9915 - val_loss: 0.2439 - val_accuracy: 0.9389
-Epoch 104/1000
-60000/60000 - 6s - loss: 0.0415 - accuracy: 0.9916 - val_loss: 0.2448 - val_accuracy: 0.9391
-Epoch 105/1000
-60000/60000 - 6s - loss: 0.0415 - accuracy: 0.9915 - val_loss: 0.2441 - val_accuracy: 0.9402
-Epoch 106/1000
-60000/60000 - 6s - loss: 0.0407 - accuracy: 0.9916 - val_loss: 0.2464 - val_accuracy: 0.9396
-Epoch 107/1000
-60000/60000 - 6s - loss: 0.0408 - accuracy: 0.9916 - val_loss: 0.2467 - val_accuracy: 0.9387
-Epoch 108/1000
-60000/60000 - 6s - loss: 0.0402 - accuracy: 0.9919 - val_loss: 0.2480 - val_accuracy: 0.9385
-Epoch 109/1000
-60000/60000 - 6s - loss: 0.0400 - accuracy: 0.9919 - val_loss: 0.2487 - val_accuracy: 0.9388
-Epoch 110/1000
-60000/60000 - 6s - loss: 0.0395 - accuracy: 0.9918 - val_loss: 0.2500 - val_accuracy: 0.9390
-Epoch 111/1000
-60000/60000 - 6s - loss: 0.0394 - accuracy: 0.9923 - val_loss: 0.2514 - val_accuracy: 0.9383
-Epoch 112/1000
-60000/60000 - 6s - loss: 0.0390 - accuracy: 0.9922 - val_loss: 0.2501 - val_accuracy: 0.9381
-Epoch 113/1000
-60000/60000 - 6s - loss: 0.0387 - accuracy: 0.9922 - val_loss: 0.2520 - val_accuracy: 0.9390
-Epoch 114/1000
-60000/60000 - 6s - loss: 0.0384 - accuracy: 0.9926 - val_loss: 0.2533 - val_accuracy: 0.9376
-Epoch 115/1000
-60000/60000 - 6s - loss: 0.0383 - accuracy: 0.9927 - val_loss: 0.2559 - val_accuracy: 0.9384
-Epoch 116/1000
-60000/60000 - 6s - loss: 0.0378 - accuracy: 0.9926 - val_loss: 0.2552 - val_accuracy: 0.9384
-Epoch 117/1000
-60000/60000 - 6s - loss: 0.0374 - accuracy: 0.9928 - val_loss: 0.2533 - val_accuracy: 0.9393
-Epoch 118/1000
-60000/60000 - 6s - loss: 0.0374 - accuracy: 0.9928 - val_loss: 0.2552 - val_accuracy: 0.9388
-Epoch 119/1000
-60000/60000 - 6s - loss: 0.0370 - accuracy: 0.9929 - val_loss: 0.2540 - val_accuracy: 0.9398
-Epoch 120/1000
-60000/60000 - 6s - loss: 0.0368 - accuracy: 0.9932 - val_loss: 0.2558 - val_accuracy: 0.9400
-Epoch 121/1000
-60000/60000 - 6s - loss: 0.0364 - accuracy: 0.9931 - val_loss: 0.2589 - val_accuracy: 0.9387
-Epoch 122/1000
-60000/60000 - 6s - loss: 0.0362 - accuracy: 0.9931 - val_loss: 0.2586 - val_accuracy: 0.9391
-Epoch 123/1000
-60000/60000 - 6s - loss: 0.0360 - accuracy: 0.9933 - val_loss: 0.2587 - val_accuracy: 0.9386
-Epoch 124/1000
-60000/60000 - 6s - loss: 0.0357 - accuracy: 0.9932 - val_loss: 0.2592 - val_accuracy: 0.9389
-Epoch 125/1000
-60000/60000 - 6s - loss: 0.0356 - accuracy: 0.9932 - val_loss: 0.2592 - val_accuracy: 0.9388
-Epoch 126/1000
-60000/60000 - 6s - loss: 0.0353 - accuracy: 0.9935 - val_loss: 0.2628 - val_accuracy: 0.9384
-Epoch 127/1000
-60000/60000 - 6s - loss: 0.0350 - accuracy: 0.9937 - val_loss: 0.2617 - val_accuracy: 0.9394
-Epoch 128/1000
-60000/60000 - 6s - loss: 0.0347 - accuracy: 0.9936 - val_loss: 0.2625 - val_accuracy: 0.9381
-Epoch 129/1000
-60000/60000 - 6s - loss: 0.0346 - accuracy: 0.9935 - val_loss: 0.2637 - val_accuracy: 0.9391
-Epoch 130/1000
-60000/60000 - 6s - loss: 0.0343 - accuracy: 0.9936 - val_loss: 0.2609 - val_accuracy: 0.9389
-Epoch 131/1000
-60000/60000 - 6s - loss: 0.0340 - accuracy: 0.9937 - val_loss: 0.2656 - val_accuracy: 0.9376
-Epoch 132/1000
-60000/60000 - 6s - loss: 0.0336 - accuracy: 0.9941 - val_loss: 0.2646 - val_accuracy: 0.9394
-Epoch 133/1000
-60000/60000 - 6s - loss: 0.0336 - accuracy: 0.9937 - val_loss: 0.2636 - val_accuracy: 0.9388
-Epoch 134/1000
-60000/60000 - 6s - loss: 0.0333 - accuracy: 0.9941 - val_loss: 0.2642 - val_accuracy: 0.9386
-Epoch 135/1000
-60000/60000 - 6s - loss: 0.0331 - accuracy: 0.9943 - val_loss: 0.2670 - val_accuracy: 0.9395
-Epoch 136/1000
-60000/60000 - 6s - loss: 0.0331 - accuracy: 0.9939 - val_loss: 0.2665 - val_accuracy: 0.9373
-Epoch 137/1000
-60000/60000 - 6s - loss: 0.0328 - accuracy: 0.9941 - val_loss: 0.2678 - val_accuracy: 0.9393
-Epoch 138/1000
-60000/60000 - 6s - loss: 0.0326 - accuracy: 0.9941 - val_loss: 0.2718 - val_accuracy: 0.9385
-Epoch 139/1000
-60000/60000 - 6s - loss: 0.0322 - accuracy: 0.9943 - val_loss: 0.2700 - val_accuracy: 0.9382
-Epoch 140/1000
-60000/60000 - 6s - loss: 0.0320 - accuracy: 0.9944 - val_loss: 0.2687 - val_accuracy: 0.9394
-Epoch 141/1000
-60000/60000 - 6s - loss: 0.0318 - accuracy: 0.9946 - val_loss: 0.2698 - val_accuracy: 0.9379
-Epoch 142/1000
-60000/60000 - 6s - loss: 0.0317 - accuracy: 0.9944 - val_loss: 0.2700 - val_accuracy: 0.9381
-Epoch 143/1000
-60000/60000 - 6s - loss: 0.0315 - accuracy: 0.9942 - val_loss: 0.2721 - val_accuracy: 0.9389
-Epoch 144/1000
-60000/60000 - 6s - loss: 0.0312 - accuracy: 0.9947 - val_loss: 0.2725 - val_accuracy: 0.9387
-Epoch 145/1000
-60000/60000 - 6s - loss: 0.0310 - accuracy: 0.9945 - val_loss: 0.2714 - val_accuracy: 0.9388
-Epoch 146/1000
-60000/60000 - 6s - loss: 0.0309 - accuracy: 0.9949 - val_loss: 0.2743 - val_accuracy: 0.9388
-Epoch 147/1000
-60000/60000 - 6s - loss: 0.0306 - accuracy: 0.9947 - val_loss: 0.2725 - val_accuracy: 0.9389
-Epoch 148/1000
-60000/60000 - 6s - loss: 0.0303 - accuracy: 0.9947 - val_loss: 0.2758 - val_accuracy: 0.9376
-Epoch 149/1000
-60000/60000 - 6s - loss: 0.0303 - accuracy: 0.9950 - val_loss: 0.2759 - val_accuracy: 0.9392
-Epoch 150/1000
-60000/60000 - 6s - loss: 0.0301 - accuracy: 0.9949 - val_loss: 0.2765 - val_accuracy: 0.9395
-Epoch 151/1000
-60000/60000 - 6s - loss: 0.0299 - accuracy: 0.9948 - val_loss: 0.2763 - val_accuracy: 0.9384
-Epoch 152/1000
-60000/60000 - 6s - loss: 0.0297 - accuracy: 0.9948 - val_loss: 0.2774 - val_accuracy: 0.9382
-Epoch 153/1000
-60000/60000 - 6s - loss: 0.0295 - accuracy: 0.9949 - val_loss: 0.2785 - val_accuracy: 0.9375
-Epoch 154/1000
-60000/60000 - 6s - loss: 0.0294 - accuracy: 0.9950 - val_loss: 0.2789 - val_accuracy: 0.9378
-Epoch 155/1000
-60000/60000 - 6s - loss: 0.0292 - accuracy: 0.9950 - val_loss: 0.2786 - val_accuracy: 0.9382
-Epoch 156/1000
-60000/60000 - 6s - loss: 0.0291 - accuracy: 0.9952 - val_loss: 0.2796 - val_accuracy: 0.9368
-Epoch 157/1000
-60000/60000 - 6s - loss: 0.0288 - accuracy: 0.9953 - val_loss: 0.2806 - val_accuracy: 0.9382
-Epoch 158/1000
-60000/60000 - 6s - loss: 0.0287 - accuracy: 0.9952 - val_loss: 0.2800 - val_accuracy: 0.9384
-Epoch 159/1000
-60000/60000 - 6s - loss: 0.0285 - accuracy: 0.9954 - val_loss: 0.2808 - val_accuracy: 0.9370
-Epoch 160/1000
-60000/60000 - 6s - loss: 0.0283 - accuracy: 0.9954 - val_loss: 0.2815 - val_accuracy: 0.9375
-Epoch 161/1000
-60000/60000 - 6s - loss: 0.0281 - accuracy: 0.9953 - val_loss: 0.2843 - val_accuracy: 0.9386
-Epoch 162/1000
-60000/60000 - 6s - loss: 0.0281 - accuracy: 0.9955 - val_loss: 0.2831 - val_accuracy: 0.9380
-Epoch 163/1000
-60000/60000 - 6s - loss: 0.0278 - accuracy: 0.9958 - val_loss: 0.2829 - val_accuracy: 0.9382
-Epoch 164/1000
-60000/60000 - 6s - loss: 0.0276 - accuracy: 0.9957 - val_loss: 0.2836 - val_accuracy: 0.9379
-Epoch 165/1000
-60000/60000 - 6s - loss: 0.0276 - accuracy: 0.9955 - val_loss: 0.2861 - val_accuracy: 0.9364
-Epoch 166/1000
-60000/60000 - 6s - loss: 0.0274 - accuracy: 0.9957 - val_loss: 0.2867 - val_accuracy: 0.9369
-Epoch 167/1000
-60000/60000 - 6s - loss: 0.0272 - accuracy: 0.9957 - val_loss: 0.2851 - val_accuracy: 0.9369
-Epoch 168/1000
-60000/60000 - 6s - loss: 0.0271 - accuracy: 0.9959 - val_loss: 0.2860 - val_accuracy: 0.9372
-Epoch 169/1000
-60000/60000 - 6s - loss: 0.0269 - accuracy: 0.9959 - val_loss: 0.2887 - val_accuracy: 0.9380
-Epoch 170/1000
-60000/60000 - 6s - loss: 0.0269 - accuracy: 0.9959 - val_loss: 0.2884 - val_accuracy: 0.9380
-Epoch 171/1000
-60000/60000 - 6s - loss: 0.0266 - accuracy: 0.9958 - val_loss: 0.2874 - val_accuracy: 0.9376
-Epoch 172/1000
-60000/60000 - 6s - loss: 0.0265 - accuracy: 0.9959 - val_loss: 0.2881 - val_accuracy: 0.9372
-Epoch 173/1000
-60000/60000 - 6s - loss: 0.0263 - accuracy: 0.9961 - val_loss: 0.2893 - val_accuracy: 0.9372
-Epoch 174/1000
-60000/60000 - 6s - loss: 0.0262 - accuracy: 0.9959 - val_loss: 0.2872 - val_accuracy: 0.9378
-Epoch 175/1000
-60000/60000 - 6s - loss: 0.0261 - accuracy: 0.9960 - val_loss: 0.2893 - val_accuracy: 0.9379
-Epoch 176/1000
-60000/60000 - 6s - loss: 0.0259 - accuracy: 0.9959 - val_loss: 0.2905 - val_accuracy: 0.9381
-Epoch 177/1000
-60000/60000 - 6s - loss: 0.0259 - accuracy: 0.9961 - val_loss: 0.2909 - val_accuracy: 0.9381
-Epoch 178/1000
-60000/60000 - 6s - loss: 0.0257 - accuracy: 0.9961 - val_loss: 0.2899 - val_accuracy: 0.9371
-Epoch 179/1000
-60000/60000 - 6s - loss: 0.0255 - accuracy: 0.9962 - val_loss: 0.2916 - val_accuracy: 0.9373
-Epoch 180/1000
-60000/60000 - 6s - loss: 0.0255 - accuracy: 0.9962 - val_loss: 0.2919 - val_accuracy: 0.9373
-Epoch 181/1000
-60000/60000 - 6s - loss: 0.0253 - accuracy: 0.9962 - val_loss: 0.2930 - val_accuracy: 0.9383
-Epoch 182/1000
-60000/60000 - 6s - loss: 0.0251 - accuracy: 0.9964 - val_loss: 0.2937 - val_accuracy: 0.9374
-Epoch 183/1000
-60000/60000 - 6s - loss: 0.0250 - accuracy: 0.9962 - val_loss: 0.2930 - val_accuracy: 0.9382
-Epoch 184/1000
-60000/60000 - 6s - loss: 0.0249 - accuracy: 0.9963 - val_loss: 0.2949 - val_accuracy: 0.9369
-Epoch 185/1000
-60000/60000 - 6s - loss: 0.0247 - accuracy: 0.9962 - val_loss: 0.2941 - val_accuracy: 0.9373
-Epoch 186/1000
-60000/60000 - 6s - loss: 0.0246 - accuracy: 0.9965 - val_loss: 0.2963 - val_accuracy: 0.9370
-Epoch 187/1000
-60000/60000 - 6s - loss: 0.0245 - accuracy: 0.9965 - val_loss: 0.2955 - val_accuracy: 0.9377
-Epoch 188/1000
-60000/60000 - 6s - loss: 0.0245 - accuracy: 0.9962 - val_loss: 0.2960 - val_accuracy: 0.9377
-Epoch 189/1000
-60000/60000 - 6s - loss: 0.0243 - accuracy: 0.9968 - val_loss: 0.2968 - val_accuracy: 0.9371
-Epoch 190/1000
-60000/60000 - 6s - loss: 0.0242 - accuracy: 0.9965 - val_loss: 0.2967 - val_accuracy: 0.9368
-Epoch 191/1000
-60000/60000 - 6s - loss: 0.0240 - accuracy: 0.9966 - val_loss: 0.2998 - val_accuracy: 0.9374
-Epoch 192/1000
-60000/60000 - 6s - loss: 0.0239 - accuracy: 0.9966 - val_loss: 0.2985 - val_accuracy: 0.9381
-Epoch 193/1000
-60000/60000 - 6s - loss: 0.0238 - accuracy: 0.9965 - val_loss: 0.2977 - val_accuracy: 0.9389
-Epoch 194/1000
-60000/60000 - 6s - loss: 0.0237 - accuracy: 0.9968 - val_loss: 0.2991 - val_accuracy: 0.9378
-Epoch 195/1000
-60000/60000 - 6s - loss: 0.0235 - accuracy: 0.9967 - val_loss: 0.3000 - val_accuracy: 0.9370
-Epoch 196/1000
-60000/60000 - 6s - loss: 0.0234 - accuracy: 0.9967 - val_loss: 0.2990 - val_accuracy: 0.9389
-Epoch 197/1000
-60000/60000 - 6s - loss: 0.0234 - accuracy: 0.9968 - val_loss: 0.3009 - val_accuracy: 0.9383
-Epoch 198/1000
-60000/60000 - 6s - loss: 0.0234 - accuracy: 0.9968 - val_loss: 0.3015 - val_accuracy: 0.9382
-Epoch 199/1000
-60000/60000 - 6s - loss: 0.0232 - accuracy: 0.9968 - val_loss: 0.3029 - val_accuracy: 0.9370
-Epoch 200/1000
-60000/60000 - 6s - loss: 0.0231 - accuracy: 0.9969 - val_loss: 0.3033 - val_accuracy: 0.9374
-Epoch 201/1000
-60000/60000 - 6s - loss: 0.0229 - accuracy: 0.9968 - val_loss: 0.3019 - val_accuracy: 0.9365
-Epoch 202/1000
-60000/60000 - 6s - loss: 0.0228 - accuracy: 0.9969 - val_loss: 0.3023 - val_accuracy: 0.9374
-Epoch 203/1000
-60000/60000 - 6s - loss: 0.0226 - accuracy: 0.9969 - val_loss: 0.3045 - val_accuracy: 0.9370
-Epoch 204/1000
-60000/60000 - 6s - loss: 0.0227 - accuracy: 0.9968 - val_loss: 0.3046 - val_accuracy: 0.9372
-Epoch 205/1000
-60000/60000 - 6s - loss: 0.0225 - accuracy: 0.9970 - val_loss: 0.3041 - val_accuracy: 0.9375
-Epoch 206/1000
-60000/60000 - 6s - loss: 0.0224 - accuracy: 0.9968 - val_loss: 0.3038 - val_accuracy: 0.9386
-Epoch 207/1000
-60000/60000 - 6s - loss: 0.0223 - accuracy: 0.9969 - val_loss: 0.3059 - val_accuracy: 0.9369
-Epoch 208/1000
-60000/60000 - 6s - loss: 0.0221 - accuracy: 0.9970 - val_loss: 0.3057 - val_accuracy: 0.9375
-Epoch 209/1000
-60000/60000 - 6s - loss: 0.0220 - accuracy: 0.9970 - val_loss: 0.3062 - val_accuracy: 0.9366
-Epoch 210/1000
-60000/60000 - 6s - loss: 0.0220 - accuracy: 0.9969 - val_loss: 0.3069 - val_accuracy: 0.9378
-Epoch 211/1000
-60000/60000 - 6s - loss: 0.0219 - accuracy: 0.9970 - val_loss: 0.3075 - val_accuracy: 0.9378
-Epoch 212/1000
-60000/60000 - 6s - loss: 0.0219 - accuracy: 0.9970 - val_loss: 0.3071 - val_accuracy: 0.9378
-Epoch 213/1000
-60000/60000 - 6s - loss: 0.0217 - accuracy: 0.9972 - val_loss: 0.3071 - val_accuracy: 0.9364
-Epoch 214/1000
-60000/60000 - 6s - loss: 0.0216 - accuracy: 0.9971 - val_loss: 0.3077 - val_accuracy: 0.9366
-Epoch 215/1000
-60000/60000 - 6s - loss: 0.0215 - accuracy: 0.9971 - val_loss: 0.3081 - val_accuracy: 0.9372
-Epoch 216/1000
-60000/60000 - 6s - loss: 0.0214 - accuracy: 0.9973 - val_loss: 0.3118 - val_accuracy: 0.9361
-Epoch 217/1000
-60000/60000 - 6s - loss: 0.0212 - accuracy: 0.9972 - val_loss: 0.3095 - val_accuracy: 0.9371
-Epoch 218/1000
-60000/60000 - 6s - loss: 0.0212 - accuracy: 0.9973 - val_loss: 0.3098 - val_accuracy: 0.9372
-Epoch 219/1000
-60000/60000 - 6s - loss: 0.0212 - accuracy: 0.9972 - val_loss: 0.3096 - val_accuracy: 0.9385
-Epoch 220/1000
-60000/60000 - 6s - loss: 0.0210 - accuracy: 0.9972 - val_loss: 0.3107 - val_accuracy: 0.9366
-Epoch 221/1000
-60000/60000 - 6s - loss: 0.0209 - accuracy: 0.9973 - val_loss: 0.3112 - val_accuracy: 0.9369
-Epoch 222/1000
-60000/60000 - 6s - loss: 0.0209 - accuracy: 0.9972 - val_loss: 0.3111 - val_accuracy: 0.9371
-Epoch 223/1000
-60000/60000 - 6s - loss: 0.0208 - accuracy: 0.9973 - val_loss: 0.3125 - val_accuracy: 0.9372
-Epoch 224/1000
-60000/60000 - 6s - loss: 0.0207 - accuracy: 0.9973 - val_loss: 0.3126 - val_accuracy: 0.9367
-Epoch 225/1000
-60000/60000 - 6s - loss: 0.0206 - accuracy: 0.9974 - val_loss: 0.3122 - val_accuracy: 0.9376
-Epoch 226/1000
-60000/60000 - 6s - loss: 0.0205 - accuracy: 0.9974 - val_loss: 0.3136 - val_accuracy: 0.9370
-Epoch 227/1000
-60000/60000 - 6s - loss: 0.0205 - accuracy: 0.9973 - val_loss: 0.3138 - val_accuracy: 0.9368
-Epoch 228/1000
-60000/60000 - 6s - loss: 0.0203 - accuracy: 0.9974 - val_loss: 0.3150 - val_accuracy: 0.9362
-Epoch 229/1000
-60000/60000 - 6s - loss: 0.0203 - accuracy: 0.9975 - val_loss: 0.3160 - val_accuracy: 0.9368
-Epoch 230/1000
-60000/60000 - 6s - loss: 0.0202 - accuracy: 0.9975 - val_loss: 0.3138 - val_accuracy: 0.9372
-Epoch 231/1000
-60000/60000 - 6s - loss: 0.0201 - accuracy: 0.9975 - val_loss: 0.3140 - val_accuracy: 0.9366
-Epoch 232/1000
-60000/60000 - 6s - loss: 0.0201 - accuracy: 0.9973 - val_loss: 0.3156 - val_accuracy: 0.9371
-Epoch 233/1000
-60000/60000 - 6s - loss: 0.0199 - accuracy: 0.9976 - val_loss: 0.3159 - val_accuracy: 0.9370
-Epoch 234/1000
-60000/60000 - 6s - loss: 0.0198 - accuracy: 0.9975 - val_loss: 0.3163 - val_accuracy: 0.9363
-Epoch 235/1000
-60000/60000 - 6s - loss: 0.0198 - accuracy: 0.9975 - val_loss: 0.3168 - val_accuracy: 0.9366
-Epoch 236/1000
-60000/60000 - 6s - loss: 0.0197 - accuracy: 0.9975 - val_loss: 0.3161 - val_accuracy: 0.9370
-Epoch 237/1000
-60000/60000 - 6s - loss: 0.0196 - accuracy: 0.9975 - val_loss: 0.3183 - val_accuracy: 0.9366
-Epoch 238/1000
-60000/60000 - 6s - loss: 0.0196 - accuracy: 0.9975 - val_loss: 0.3179 - val_accuracy: 0.9361
-Epoch 239/1000
-60000/60000 - 6s - loss: 0.0195 - accuracy: 0.9976 - val_loss: 0.3196 - val_accuracy: 0.9359
-Epoch 240/1000
-60000/60000 - 6s - loss: 0.0193 - accuracy: 0.9976 - val_loss: 0.3190 - val_accuracy: 0.9362
-Epoch 241/1000
-60000/60000 - 6s - loss: 0.0194 - accuracy: 0.9976 - val_loss: 0.3193 - val_accuracy: 0.9366
-Epoch 242/1000
-60000/60000 - 6s - loss: 0.0192 - accuracy: 0.9976 - val_loss: 0.3202 - val_accuracy: 0.9363
-Epoch 243/1000
-60000/60000 - 6s - loss: 0.0191 - accuracy: 0.9976 - val_loss: 0.3196 - val_accuracy: 0.9360
-Epoch 244/1000
-60000/60000 - 6s - loss: 0.0191 - accuracy: 0.9977 - val_loss: 0.3217 - val_accuracy: 0.9363
-Epoch 245/1000
-60000/60000 - 6s - loss: 0.0190 - accuracy: 0.9977 - val_loss: 0.3217 - val_accuracy: 0.9359
-Epoch 246/1000
-60000/60000 - 6s - loss: 0.0190 - accuracy: 0.9976 - val_loss: 0.3218 - val_accuracy: 0.9365
-Epoch 247/1000
-60000/60000 - 6s - loss: 0.0189 - accuracy: 0.9977 - val_loss: 0.3210 - val_accuracy: 0.9362
-Epoch 248/1000
-60000/60000 - 6s - loss: 0.0188 - accuracy: 0.9977 - val_loss: 0.3223 - val_accuracy: 0.9357
-Epoch 249/1000
-60000/60000 - 6s - loss: 0.0187 - accuracy: 0.9977 - val_loss: 0.3233 - val_accuracy: 0.9360
-Epoch 250/1000
-60000/60000 - 6s - loss: 0.0186 - accuracy: 0.9977 - val_loss: 0.3227 - val_accuracy: 0.9354
-Epoch 251/1000
-60000/60000 - 6s - loss: 0.0185 - accuracy: 0.9978 - val_loss: 0.3238 - val_accuracy: 0.9359
-Epoch 252/1000
-60000/60000 - 6s - loss: 0.0185 - accuracy: 0.9978 - val_loss: 0.3231 - val_accuracy: 0.9365
-Epoch 253/1000
-60000/60000 - 6s - loss: 0.0185 - accuracy: 0.9977 - val_loss: 0.3225 - val_accuracy: 0.9354
-Epoch 254/1000
-60000/60000 - 6s - loss: 0.0184 - accuracy: 0.9977 - val_loss: 0.3249 - val_accuracy: 0.9359
-Epoch 255/1000
-60000/60000 - 6s - loss: 0.0183 - accuracy: 0.9978 - val_loss: 0.3256 - val_accuracy: 0.9362
-Epoch 256/1000
-60000/60000 - 6s - loss: 0.0182 - accuracy: 0.9977 - val_loss: 0.3257 - val_accuracy: 0.9358
-Epoch 257/1000
-60000/60000 - 6s - loss: 0.0181 - accuracy: 0.9979 - val_loss: 0.3261 - val_accuracy: 0.9351
-Epoch 258/1000
-60000/60000 - 6s - loss: 0.0181 - accuracy: 0.9977 - val_loss: 0.3256 - val_accuracy: 0.9359
-Epoch 259/1000
-60000/60000 - 6s - loss: 0.0180 - accuracy: 0.9979 - val_loss: 0.3255 - val_accuracy: 0.9360
-Epoch 260/1000
-60000/60000 - 6s - loss: 0.0180 - accuracy: 0.9979 - val_loss: 0.3263 - val_accuracy: 0.9357
-Epoch 261/1000
-60000/60000 - 6s - loss: 0.0179 - accuracy: 0.9977 - val_loss: 0.3272 - val_accuracy: 0.9357
-Epoch 262/1000
-60000/60000 - 6s - loss: 0.0178 - accuracy: 0.9980 - val_loss: 0.3269 - val_accuracy: 0.9356
-Epoch 263/1000
-60000/60000 - 6s - loss: 0.0177 - accuracy: 0.9980 - val_loss: 0.3272 - val_accuracy: 0.9361
-Epoch 264/1000
-60000/60000 - 6s - loss: 0.0177 - accuracy: 0.9979 - val_loss: 0.3276 - val_accuracy: 0.9352
-Epoch 265/1000
-60000/60000 - 6s - loss: 0.0176 - accuracy: 0.9979 - val_loss: 0.3275 - val_accuracy: 0.9354
-Epoch 266/1000
-60000/60000 - 6s - loss: 0.0176 - accuracy: 0.9978 - val_loss: 0.3288 - val_accuracy: 0.9363
-Epoch 267/1000
-60000/60000 - 6s - loss: 0.0176 - accuracy: 0.9978 - val_loss: 0.3300 - val_accuracy: 0.9349
-Epoch 268/1000
-60000/60000 - 6s - loss: 0.0174 - accuracy: 0.9980 - val_loss: 0.3294 - val_accuracy: 0.9356
-Epoch 269/1000
-60000/60000 - 6s - loss: 0.0174 - accuracy: 0.9980 - val_loss: 0.3292 - val_accuracy: 0.9354
-Epoch 270/1000
-60000/60000 - 6s - loss: 0.0173 - accuracy: 0.9979 - val_loss: 0.3301 - val_accuracy: 0.9354
-Epoch 271/1000
-60000/60000 - 6s - loss: 0.0173 - accuracy: 0.9980 - val_loss: 0.3306 - val_accuracy: 0.9344
-Epoch 272/1000
-60000/60000 - 6s - loss: 0.0172 - accuracy: 0.9980 - val_loss: 0.3324 - val_accuracy: 0.9344
-Epoch 273/1000
-60000/60000 - 6s - loss: 0.0171 - accuracy: 0.9980 - val_loss: 0.3321 - val_accuracy: 0.9348
-Epoch 274/1000
-60000/60000 - 6s - loss: 0.0170 - accuracy: 0.9979 - val_loss: 0.3321 - val_accuracy: 0.9346
-Epoch 275/1000
-60000/60000 - 6s - loss: 0.0171 - accuracy: 0.9980 - val_loss: 0.3309 - val_accuracy: 0.9354
-Epoch 276/1000
-60000/60000 - 6s - loss: 0.0169 - accuracy: 0.9980 - val_loss: 0.3308 - val_accuracy: 0.9352
-Epoch 277/1000
-60000/60000 - 6s - loss: 0.0168 - accuracy: 0.9980 - val_loss: 0.3310 - val_accuracy: 0.9351
-Epoch 278/1000
-60000/60000 - 6s - loss: 0.0168 - accuracy: 0.9980 - val_loss: 0.3324 - val_accuracy: 0.9349
-Epoch 279/1000
-60000/60000 - 6s - loss: 0.0167 - accuracy: 0.9981 - val_loss: 0.3346 - val_accuracy: 0.9343
-Epoch 280/1000
-60000/60000 - 6s - loss: 0.0167 - accuracy: 0.9981 - val_loss: 0.3331 - val_accuracy: 0.9349
-Epoch 281/1000
-60000/60000 - 6s - loss: 0.0166 - accuracy: 0.9980 - val_loss: 0.3339 - val_accuracy: 0.9344
-Epoch 282/1000
-60000/60000 - 6s - loss: 0.0166 - accuracy: 0.9981 - val_loss: 0.3351 - val_accuracy: 0.9347
-Epoch 283/1000
-60000/60000 - 6s - loss: 0.0166 - accuracy: 0.9981 - val_loss: 0.3362 - val_accuracy: 0.9348
-Epoch 284/1000
-60000/60000 - 6s - loss: 0.0165 - accuracy: 0.9981 - val_loss: 0.3354 - val_accuracy: 0.9338
-Epoch 285/1000
-60000/60000 - 6s - loss: 0.0164 - accuracy: 0.9981 - val_loss: 0.3350 - val_accuracy: 0.9356
-Epoch 286/1000
-60000/60000 - 6s - loss: 0.0164 - accuracy: 0.9981 - val_loss: 0.3355 - val_accuracy: 0.9349
-Epoch 287/1000
-60000/60000 - 6s - loss: 0.0163 - accuracy: 0.9981 - val_loss: 0.3356 - val_accuracy: 0.9346
-Epoch 288/1000
-60000/60000 - 6s - loss: 0.0163 - accuracy: 0.9981 - val_loss: 0.3365 - val_accuracy: 0.9339
-Epoch 289/1000
-60000/60000 - 6s - loss: 0.0162 - accuracy: 0.9981 - val_loss: 0.3357 - val_accuracy: 0.9351
-Epoch 290/1000
-60000/60000 - 6s - loss: 0.0161 - accuracy: 0.9981 - val_loss: 0.3377 - val_accuracy: 0.9348
-Epoch 291/1000
-60000/60000 - 6s - loss: 0.0161 - accuracy: 0.9982 - val_loss: 0.3377 - val_accuracy: 0.9345
-Epoch 292/1000
-60000/60000 - 6s - loss: 0.0160 - accuracy: 0.9981 - val_loss: 0.3370 - val_accuracy: 0.9350
-Epoch 293/1000
-60000/60000 - 6s - loss: 0.0160 - accuracy: 0.9983 - val_loss: 0.3380 - val_accuracy: 0.9340
-Epoch 294/1000
-60000/60000 - 6s - loss: 0.0160 - accuracy: 0.9981 - val_loss: 0.3387 - val_accuracy: 0.9345
-Epoch 295/1000
-60000/60000 - 6s - loss: 0.0159 - accuracy: 0.9982 - val_loss: 0.3378 - val_accuracy: 0.9343
-Epoch 296/1000
-60000/60000 - 6s - loss: 0.0158 - accuracy: 0.9982 - val_loss: 0.3379 - val_accuracy: 0.9348
-Epoch 297/1000
-60000/60000 - 6s - loss: 0.0157 - accuracy: 0.9983 - val_loss: 0.3384 - val_accuracy: 0.9342
-Epoch 298/1000
-60000/60000 - 6s - loss: 0.0157 - accuracy: 0.9983 - val_loss: 0.3389 - val_accuracy: 0.9340
-Epoch 299/1000
-60000/60000 - 6s - loss: 0.0157 - accuracy: 0.9983 - val_loss: 0.3400 - val_accuracy: 0.9334
-Epoch 300/1000
-60000/60000 - 6s - loss: 0.0156 - accuracy: 0.9983 - val_loss: 0.3401 - val_accuracy: 0.9339
-Epoch 301/1000
-60000/60000 - 6s - loss: 0.0156 - accuracy: 0.9984 - val_loss: 0.3393 - val_accuracy: 0.9344
-Epoch 302/1000
-60000/60000 - 6s - loss: 0.0155 - accuracy: 0.9984 - val_loss: 0.3403 - val_accuracy: 0.9345
-Epoch 303/1000
-60000/60000 - 6s - loss: 0.0154 - accuracy: 0.9982 - val_loss: 0.3408 - val_accuracy: 0.9343
-Epoch 304/1000
-60000/60000 - 6s - loss: 0.0154 - accuracy: 0.9983 - val_loss: 0.3415 - val_accuracy: 0.9347
-Epoch 305/1000
-60000/60000 - 6s - loss: 0.0154 - accuracy: 0.9983 - val_loss: 0.3417 - val_accuracy: 0.9340
-Epoch 306/1000
-60000/60000 - 6s - loss: 0.0153 - accuracy: 0.9983 - val_loss: 0.3407 - val_accuracy: 0.9341
-Epoch 307/1000
-60000/60000 - 6s - loss: 0.0153 - accuracy: 0.9984 - val_loss: 0.3414 - val_accuracy: 0.9348
-Epoch 308/1000
-60000/60000 - 6s - loss: 0.0152 - accuracy: 0.9984 - val_loss: 0.3433 - val_accuracy: 0.9341
-Epoch 309/1000
-60000/60000 - 6s - loss: 0.0152 - accuracy: 0.9984 - val_loss: 0.3423 - val_accuracy: 0.9339
-Epoch 310/1000
-60000/60000 - 6s - loss: 0.0152 - accuracy: 0.9984 - val_loss: 0.3435 - val_accuracy: 0.9344
-Epoch 311/1000
-60000/60000 - 6s - loss: 0.0151 - accuracy: 0.9984 - val_loss: 0.3447 - val_accuracy: 0.9341
-Epoch 312/1000
-60000/60000 - 6s - loss: 0.0150 - accuracy: 0.9984 - val_loss: 0.3436 - val_accuracy: 0.9346
-Epoch 313/1000
-60000/60000 - 6s - loss: 0.0150 - accuracy: 0.9984 - val_loss: 0.3441 - val_accuracy: 0.9339
-Epoch 314/1000
-60000/60000 - 6s - loss: 0.0150 - accuracy: 0.9983 - val_loss: 0.3444 - val_accuracy: 0.9341
-Epoch 315/1000
-60000/60000 - 6s - loss: 0.0149 - accuracy: 0.9983 - val_loss: 0.3442 - val_accuracy: 0.9347
-Epoch 316/1000
-60000/60000 - 6s - loss: 0.0149 - accuracy: 0.9984 - val_loss: 0.3450 - val_accuracy: 0.9336
-Epoch 317/1000
-60000/60000 - 6s - loss: 0.0147 - accuracy: 0.9984 - val_loss: 0.3455 - val_accuracy: 0.9343
-Epoch 318/1000
-60000/60000 - 6s - loss: 0.0148 - accuracy: 0.9984 - val_loss: 0.3459 - val_accuracy: 0.9341
-Epoch 319/1000
-60000/60000 - 6s - loss: 0.0147 - accuracy: 0.9985 - val_loss: 0.3447 - val_accuracy: 0.9341
-Epoch 320/1000
-60000/60000 - 6s - loss: 0.0146 - accuracy: 0.9985 - val_loss: 0.3459 - val_accuracy: 0.9345
-Epoch 321/1000
-60000/60000 - 6s - loss: 0.0146 - accuracy: 0.9984 - val_loss: 0.3456 - val_accuracy: 0.9344
-Epoch 322/1000
-60000/60000 - 6s - loss: 0.0146 - accuracy: 0.9984 - val_loss: 0.3456 - val_accuracy: 0.9350
-Epoch 323/1000
-60000/60000 - 6s - loss: 0.0145 - accuracy: 0.9985 - val_loss: 0.3476 - val_accuracy: 0.9339
-Epoch 324/1000
-60000/60000 - 6s - loss: 0.0145 - accuracy: 0.9985 - val_loss: 0.3475 - val_accuracy: 0.9333
-Epoch 325/1000
-60000/60000 - 6s - loss: 0.0144 - accuracy: 0.9984 - val_loss: 0.3477 - val_accuracy: 0.9339
-Epoch 326/1000
-60000/60000 - 6s - loss: 0.0144 - accuracy: 0.9985 - val_loss: 0.3483 - val_accuracy: 0.9334
-Epoch 327/1000
-60000/60000 - 6s - loss: 0.0144 - accuracy: 0.9985 - val_loss: 0.3471 - val_accuracy: 0.9343
-Epoch 328/1000
-60000/60000 - 6s - loss: 0.0143 - accuracy: 0.9985 - val_loss: 0.3484 - val_accuracy: 0.9340
-Epoch 329/1000
-60000/60000 - 6s - loss: 0.0143 - accuracy: 0.9986 - val_loss: 0.3491 - val_accuracy: 0.9329
-Epoch 330/1000
-60000/60000 - 6s - loss: 0.0142 - accuracy: 0.9984 - val_loss: 0.3490 - val_accuracy: 0.9336
-Epoch 331/1000
-60000/60000 - 6s - loss: 0.0142 - accuracy: 0.9985 - val_loss: 0.3498 - val_accuracy: 0.9337
-Epoch 332/1000
-60000/60000 - 6s - loss: 0.0141 - accuracy: 0.9985 - val_loss: 0.3487 - val_accuracy: 0.9338
-Epoch 333/1000
-60000/60000 - 6s - loss: 0.0141 - accuracy: 0.9985 - val_loss: 0.3491 - val_accuracy: 0.9340
-Epoch 334/1000
-60000/60000 - 6s - loss: 0.0140 - accuracy: 0.9986 - val_loss: 0.3501 - val_accuracy: 0.9333
-Epoch 335/1000
-60000/60000 - 6s - loss: 0.0140 - accuracy: 0.9985 - val_loss: 0.3506 - val_accuracy: 0.9334
-Epoch 336/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9986 - val_loss: 0.3503 - val_accuracy: 0.9342
-Epoch 337/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9986 - val_loss: 0.3513 - val_accuracy: 0.9326
-Epoch 338/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9986 - val_loss: 0.3494 - val_accuracy: 0.9328
-Epoch 339/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9985 - val_loss: 0.3510 - val_accuracy: 0.9326
-Epoch 340/1000
-60000/60000 - 6s - loss: 0.0139 - accuracy: 0.9986 - val_loss: 0.3515 - val_accuracy: 0.9329
-Epoch 341/1000
-60000/60000 - 6s - loss: 0.0137 - accuracy: 0.9986 - val_loss: 0.3518 - val_accuracy: 0.9341
-Epoch 342/1000
-60000/60000 - 6s - loss: 0.0138 - accuracy: 0.9986 - val_loss: 0.3513 - val_accuracy: 0.9334
-Epoch 343/1000
-60000/60000 - 6s - loss: 0.0137 - accuracy: 0.9986 - val_loss: 0.3512 - val_accuracy: 0.9333
-Epoch 344/1000
-60000/60000 - 6s - loss: 0.0136 - accuracy: 0.9985 - val_loss: 0.3522 - val_accuracy: 0.9334
-Epoch 345/1000
-60000/60000 - 6s - loss: 0.0137 - accuracy: 0.9987 - val_loss: 0.3522 - val_accuracy: 0.9341
-Epoch 346/1000
-60000/60000 - 6s - loss: 0.0136 - accuracy: 0.9986 - val_loss: 0.3535 - val_accuracy: 0.9331
-Epoch 347/1000
-60000/60000 - 6s - loss: 0.0136 - accuracy: 0.9986 - val_loss: 0.3541 - val_accuracy: 0.9328
-Epoch 348/1000
-60000/60000 - 6s - loss: 0.0135 - accuracy: 0.9987 - val_loss: 0.3531 - val_accuracy: 0.9326
-Epoch 349/1000
-60000/60000 - 6s - loss: 0.0135 - accuracy: 0.9987 - val_loss: 0.3538 - val_accuracy: 0.9336
-Epoch 350/1000
-60000/60000 - 6s - loss: 0.0134 - accuracy: 0.9987 - val_loss: 0.3531 - val_accuracy: 0.9338
-Epoch 351/1000
-60000/60000 - 6s - loss: 0.0134 - accuracy: 0.9987 - val_loss: 0.3530 - val_accuracy: 0.9337
-Epoch 352/1000
-60000/60000 - 6s - loss: 0.0134 - accuracy: 0.9986 - val_loss: 0.3536 - val_accuracy: 0.9334
-Epoch 353/1000
-60000/60000 - 6s - loss: 0.0133 - accuracy: 0.9987 - val_loss: 0.3563 - val_accuracy: 0.9334
-Epoch 354/1000
-60000/60000 - 6s - loss: 0.0133 - accuracy: 0.9986 - val_loss: 0.3573 - val_accuracy: 0.9330
-Epoch 355/1000
-60000/60000 - 6s - loss: 0.0132 - accuracy: 0.9987 - val_loss: 0.3537 - val_accuracy: 0.9335
-Epoch 356/1000
-60000/60000 - 6s - loss: 0.0132 - accuracy: 0.9987 - val_loss: 0.3554 - val_accuracy: 0.9327
-Epoch 357/1000
-60000/60000 - 6s - loss: 0.0132 - accuracy: 0.9987 - val_loss: 0.3557 - val_accuracy: 0.9334
-Epoch 358/1000
-60000/60000 - 6s - loss: 0.0131 - accuracy: 0.9987 - val_loss: 0.3562 - val_accuracy: 0.9326
-Epoch 359/1000
-60000/60000 - 6s - loss: 0.0131 - accuracy: 0.9987 - val_loss: 0.3581 - val_accuracy: 0.9339
-Epoch 360/1000
-60000/60000 - 6s - loss: 0.0130 - accuracy: 0.9988 - val_loss: 0.3567 - val_accuracy: 0.9337
-Epoch 361/1000
-60000/60000 - 6s - loss: 0.0130 - accuracy: 0.9987 - val_loss: 0.3567 - val_accuracy: 0.9328
-Epoch 362/1000
-60000/60000 - 6s - loss: 0.0130 - accuracy: 0.9987 - val_loss: 0.3588 - val_accuracy: 0.9334
-Epoch 363/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9987 - val_loss: 0.3576 - val_accuracy: 0.9333
-Epoch 364/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9987 - val_loss: 0.3575 - val_accuracy: 0.9334
-Epoch 365/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9987 - val_loss: 0.3570 - val_accuracy: 0.9328
-Epoch 366/1000
-60000/60000 - 6s - loss: 0.0129 - accuracy: 0.9988 - val_loss: 0.3580 - val_accuracy: 0.9339
-Epoch 367/1000
-60000/60000 - 6s - loss: 0.0128 - accuracy: 0.9987 - val_loss: 0.3589 - val_accuracy: 0.9328
-Epoch 368/1000
-60000/60000 - 6s - loss: 0.0128 - accuracy: 0.9987 - val_loss: 0.3587 - val_accuracy: 0.9325
-Epoch 369/1000
-60000/60000 - 6s - loss: 0.0127 - accuracy: 0.9988 - val_loss: 0.3586 - val_accuracy: 0.9329
-Epoch 370/1000
-60000/60000 - 6s - loss: 0.0127 - accuracy: 0.9987 - val_loss: 0.3587 - val_accuracy: 0.9331
-Epoch 371/1000
-60000/60000 - 6s - loss: 0.0127 - accuracy: 0.9988 - val_loss: 0.3589 - val_accuracy: 0.9332
-Epoch 372/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9988 - val_loss: 0.3605 - val_accuracy: 0.9321
-Epoch 373/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9988 - val_loss: 0.3602 - val_accuracy: 0.9330
-Epoch 374/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9988 - val_loss: 0.3598 - val_accuracy: 0.9331
-Epoch 375/1000
-60000/60000 - 6s - loss: 0.0126 - accuracy: 0.9988 - val_loss: 0.3604 - val_accuracy: 0.9331
-Epoch 376/1000
-60000/60000 - 6s - loss: 0.0125 - accuracy: 0.9988 - val_loss: 0.3602 - val_accuracy: 0.9333
-Epoch 377/1000
-60000/60000 - 6s - loss: 0.0125 - accuracy: 0.9987 - val_loss: 0.3622 - val_accuracy: 0.9328
-Epoch 378/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9988 - val_loss: 0.3599 - val_accuracy: 0.9336
-Epoch 379/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9987 - val_loss: 0.3621 - val_accuracy: 0.9322
-Epoch 380/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9988 - val_loss: 0.3616 - val_accuracy: 0.9337
-Epoch 381/1000
-60000/60000 - 6s - loss: 0.0124 - accuracy: 0.9988 - val_loss: 0.3621 - val_accuracy: 0.9334
-Epoch 382/1000
-60000/60000 - 6s - loss: 0.0123 - accuracy: 0.9988 - val_loss: 0.3629 - val_accuracy: 0.9342
-Epoch 383/1000
-60000/60000 - 6s - loss: 0.0123 - accuracy: 0.9988 - val_loss: 0.3632 - val_accuracy: 0.9333
-Epoch 384/1000
-60000/60000 - 6s - loss: 0.0123 - accuracy: 0.9988 - val_loss: 0.3630 - val_accuracy: 0.9334
-Epoch 385/1000
-60000/60000 - 6s - loss: 0.0122 - accuracy: 0.9988 - val_loss: 0.3624 - val_accuracy: 0.9332
-Epoch 386/1000
-60000/60000 - 6s - loss: 0.0122 - accuracy: 0.9989 - val_loss: 0.3631 - val_accuracy: 0.9323
-Epoch 387/1000
-60000/60000 - 6s - loss: 0.0121 - accuracy: 0.9988 - val_loss: 0.3641 - val_accuracy: 0.9340
-Epoch 388/1000
-60000/60000 - 6s - loss: 0.0121 - accuracy: 0.9989 - val_loss: 0.3636 - val_accuracy: 0.9325
-Epoch 389/1000
-60000/60000 - 6s - loss: 0.0121 - accuracy: 0.9988 - val_loss: 0.3646 - val_accuracy: 0.9330
-Epoch 390/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9988 - val_loss: 0.3654 - val_accuracy: 0.9327
-Epoch 391/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9988 - val_loss: 0.3643 - val_accuracy: 0.9330
-Epoch 392/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9988 - val_loss: 0.3647 - val_accuracy: 0.9326
-Epoch 393/1000
-60000/60000 - 6s - loss: 0.0120 - accuracy: 0.9988 - val_loss: 0.3656 - val_accuracy: 0.9330
-Epoch 394/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9988 - val_loss: 0.3640 - val_accuracy: 0.9324
-Epoch 395/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9989 - val_loss: 0.3649 - val_accuracy: 0.9334
-Epoch 396/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9989 - val_loss: 0.3658 - val_accuracy: 0.9337
-Epoch 397/1000
-60000/60000 - 6s - loss: 0.0119 - accuracy: 0.9989 - val_loss: 0.3661 - val_accuracy: 0.9338
-Epoch 398/1000
-60000/60000 - 6s - loss: 0.0118 - accuracy: 0.9989 - val_loss: 0.3658 - val_accuracy: 0.9330
-Epoch 399/1000
-60000/60000 - 6s - loss: 0.0118 - accuracy: 0.9989 - val_loss: 0.3672 - val_accuracy: 0.9336
-Epoch 400/1000
-60000/60000 - 6s - loss: 0.0118 - accuracy: 0.9989 - val_loss: 0.3670 - val_accuracy: 0.9329
-Epoch 401/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9989 - val_loss: 0.3665 - val_accuracy: 0.9333
-Epoch 402/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9989 - val_loss: 0.3671 - val_accuracy: 0.9331
-Epoch 403/1000
-60000/60000 - 6s - loss: 0.0117 - accuracy: 0.9989 - val_loss: 0.3667 - val_accuracy: 0.9332
-Epoch 404/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9989 - val_loss: 0.3667 - val_accuracy: 0.9333
-Epoch 405/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9989 - val_loss: 0.3681 - val_accuracy: 0.9337
-Epoch 406/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9989 - val_loss: 0.3677 - val_accuracy: 0.9327
-Epoch 407/1000
-60000/60000 - 6s - loss: 0.0116 - accuracy: 0.9989 - val_loss: 0.3686 - val_accuracy: 0.9329
-Epoch 408/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9989 - val_loss: 0.3681 - val_accuracy: 0.9328
-Epoch 409/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9989 - val_loss: 0.3682 - val_accuracy: 0.9340
-Epoch 410/1000
-60000/60000 - 6s - loss: 0.0115 - accuracy: 0.9989 - val_loss: 0.3681 - val_accuracy: 0.9334
-Epoch 411/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9989 - val_loss: 0.3698 - val_accuracy: 0.9334
-Epoch 412/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9989 - val_loss: 0.3692 - val_accuracy: 0.9328
-Epoch 413/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9989 - val_loss: 0.3698 - val_accuracy: 0.9334
-Epoch 414/1000
-60000/60000 - 6s - loss: 0.0114 - accuracy: 0.9989 - val_loss: 0.3702 - val_accuracy: 0.9327
-Epoch 415/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9989 - val_loss: 0.3701 - val_accuracy: 0.9327
-Epoch 416/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9989 - val_loss: 0.3705 - val_accuracy: 0.9326
-Epoch 417/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9990 - val_loss: 0.3714 - val_accuracy: 0.9329
-Epoch 418/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9990 - val_loss: 0.3710 - val_accuracy: 0.9330
-Epoch 419/1000
-60000/60000 - 6s - loss: 0.0113 - accuracy: 0.9989 - val_loss: 0.3706 - val_accuracy: 0.9325
-Epoch 420/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9989 - val_loss: 0.3718 - val_accuracy: 0.9336
-Epoch 421/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9990 - val_loss: 0.3730 - val_accuracy: 0.9327
-Epoch 422/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9989 - val_loss: 0.3720 - val_accuracy: 0.9330
-Epoch 423/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9989 - val_loss: 0.3724 - val_accuracy: 0.9335
-Epoch 424/1000
-60000/60000 - 6s - loss: 0.0112 - accuracy: 0.9989 - val_loss: 0.3721 - val_accuracy: 0.9323
-Epoch 425/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9990 - val_loss: 0.3720 - val_accuracy: 0.9323
-Epoch 426/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9990 - val_loss: 0.3725 - val_accuracy: 0.9324
-Epoch 427/1000
-60000/60000 - 6s - loss: 0.0111 - accuracy: 0.9989 - val_loss: 0.3730 - val_accuracy: 0.9331
-Epoch 428/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9989 - val_loss: 0.3734 - val_accuracy: 0.9321
-Epoch 429/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9990 - val_loss: 0.3731 - val_accuracy: 0.9327
-Epoch 430/1000
-60000/60000 - 6s - loss: 0.0110 - accuracy: 0.9990 - val_loss: 0.3734 - val_accuracy: 0.9328
-Epoch 431/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9990 - val_loss: 0.3742 - val_accuracy: 0.9324
-Epoch 432/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9990 - val_loss: 0.3739 - val_accuracy: 0.9333
-Epoch 433/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9990 - val_loss: 0.3736 - val_accuracy: 0.9332
-Epoch 434/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9991 - val_loss: 0.3755 - val_accuracy: 0.9315
-Epoch 435/1000
-60000/60000 - 6s - loss: 0.0109 - accuracy: 0.9990 - val_loss: 0.3752 - val_accuracy: 0.9324
-Epoch 436/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9990 - val_loss: 0.3747 - val_accuracy: 0.9328
-Epoch 437/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9991 - val_loss: 0.3762 - val_accuracy: 0.9330
-Epoch 438/1000
-60000/60000 - 6s - loss: 0.0108 - accuracy: 0.9990 - val_loss: 0.3766 - val_accuracy: 0.9327
-Epoch 439/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9990 - val_loss: 0.3754 - val_accuracy: 0.9326
-Epoch 440/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9990 - val_loss: 0.3756 - val_accuracy: 0.9326
-Epoch 441/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9990 - val_loss: 0.3766 - val_accuracy: 0.9324
-Epoch 442/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9990 - val_loss: 0.3760 - val_accuracy: 0.9323
-Epoch 443/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9990 - val_loss: 0.3774 - val_accuracy: 0.9331
-Epoch 444/1000
-60000/60000 - 6s - loss: 0.0107 - accuracy: 0.9991 - val_loss: 0.3775 - val_accuracy: 0.9325
-Epoch 445/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9990 - val_loss: 0.3766 - val_accuracy: 0.9324
-Epoch 446/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9990 - val_loss: 0.3776 - val_accuracy: 0.9327
-Epoch 447/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9990 - val_loss: 0.3773 - val_accuracy: 0.9324
-Epoch 448/1000
-60000/60000 - 6s - loss: 0.0106 - accuracy: 0.9991 - val_loss: 0.3778 - val_accuracy: 0.9322
-Epoch 449/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9991 - val_loss: 0.3788 - val_accuracy: 0.9325
-Epoch 450/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9991 - val_loss: 0.3794 - val_accuracy: 0.9325
-Epoch 451/1000
-60000/60000 - 6s - loss: 0.0105 - accuracy: 0.9990 - val_loss: 0.3779 - val_accuracy: 0.9319
-Epoch 452/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9990 - val_loss: 0.3793 - val_accuracy: 0.9315
-Epoch 453/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9991 - val_loss: 0.3787 - val_accuracy: 0.9326
-Epoch 454/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9990 - val_loss: 0.3787 - val_accuracy: 0.9322
-Epoch 455/1000
-60000/60000 - 6s - loss: 0.0104 - accuracy: 0.9990 - val_loss: 0.3794 - val_accuracy: 0.9318
-Epoch 456/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9991 - val_loss: 0.3808 - val_accuracy: 0.9328
-Epoch 457/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9991 - val_loss: 0.3800 - val_accuracy: 0.9326
-Epoch 458/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9990 - val_loss: 0.3795 - val_accuracy: 0.9325
-Epoch 459/1000
-60000/60000 - 6s - loss: 0.0103 - accuracy: 0.9991 - val_loss: 0.3812 - val_accuracy: 0.9321
-Epoch 460/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9990 - val_loss: 0.3811 - val_accuracy: 0.9323
-Epoch 461/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9991 - val_loss: 0.3806 - val_accuracy: 0.9322
-Epoch 462/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9991 - val_loss: 0.3812 - val_accuracy: 0.9331
-Epoch 463/1000
-60000/60000 - 6s - loss: 0.0102 - accuracy: 0.9991 - val_loss: 0.3813 - val_accuracy: 0.9328
-Epoch 464/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9991 - val_loss: 0.3812 - val_accuracy: 0.9322
-Epoch 465/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9991 - val_loss: 0.3817 - val_accuracy: 0.9319
-Epoch 466/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9991 - val_loss: 0.3817 - val_accuracy: 0.9324
-Epoch 467/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9991 - val_loss: 0.3833 - val_accuracy: 0.9326
-Epoch 468/1000
-60000/60000 - 6s - loss: 0.0101 - accuracy: 0.9991 - val_loss: 0.3824 - val_accuracy: 0.9324
-Epoch 469/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9991 - val_loss: 0.3825 - val_accuracy: 0.9315
-Epoch 470/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9991 - val_loss: 0.3830 - val_accuracy: 0.9321
-Epoch 471/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9990 - val_loss: 0.3822 - val_accuracy: 0.9326
-Epoch 472/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9991 - val_loss: 0.3839 - val_accuracy: 0.9320
-Epoch 473/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9991 - val_loss: 0.3827 - val_accuracy: 0.9320
-Epoch 474/1000
-60000/60000 - 6s - loss: 0.0100 - accuracy: 0.9991 - val_loss: 0.3839 - val_accuracy: 0.9328
-Epoch 475/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9991 - val_loss: 0.3835 - val_accuracy: 0.9327
-Epoch 476/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9991 - val_loss: 0.3838 - val_accuracy: 0.9326
-Epoch 477/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9991 - val_loss: 0.3841 - val_accuracy: 0.9322
-Epoch 478/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9991 - val_loss: 0.3845 - val_accuracy: 0.9328
-Epoch 479/1000
-60000/60000 - 6s - loss: 0.0099 - accuracy: 0.9991 - val_loss: 0.3845 - val_accuracy: 0.9323
-Epoch 480/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9991 - val_loss: 0.3844 - val_accuracy: 0.9327
-Epoch 481/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9991 - val_loss: 0.3853 - val_accuracy: 0.9322
-Epoch 482/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9991 - val_loss: 0.3853 - val_accuracy: 0.9322
-Epoch 483/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9991 - val_loss: 0.3857 - val_accuracy: 0.9321
-Epoch 484/1000
-60000/60000 - 6s - loss: 0.0098 - accuracy: 0.9991 - val_loss: 0.3866 - val_accuracy: 0.9321
-Epoch 485/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9991 - val_loss: 0.3857 - val_accuracy: 0.9319
-Epoch 486/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9991 - val_loss: 0.3859 - val_accuracy: 0.9321
-Epoch 487/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9991 - val_loss: 0.3858 - val_accuracy: 0.9317
-Epoch 488/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9991 - val_loss: 0.3867 - val_accuracy: 0.9325
-Epoch 489/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9991 - val_loss: 0.3867 - val_accuracy: 0.9323
-Epoch 490/1000
-60000/60000 - 6s - loss: 0.0097 - accuracy: 0.9991 - val_loss: 0.3872 - val_accuracy: 0.9323
-Epoch 491/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9991 - val_loss: 0.3876 - val_accuracy: 0.9323
-Epoch 492/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9991 - val_loss: 0.3878 - val_accuracy: 0.9326
-Epoch 493/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9991 - val_loss: 0.3871 - val_accuracy: 0.9319
-Epoch 494/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9992 - val_loss: 0.3881 - val_accuracy: 0.9322
-Epoch 495/1000
-60000/60000 - 6s - loss: 0.0096 - accuracy: 0.9991 - val_loss: 0.3887 - val_accuracy: 0.9328
-Epoch 496/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9992 - val_loss: 0.3881 - val_accuracy: 0.9322
-Epoch 497/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9991 - val_loss: 0.3884 - val_accuracy: 0.9317
-Epoch 498/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9991 - val_loss: 0.3888 - val_accuracy: 0.9322
-Epoch 499/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9991 - val_loss: 0.3890 - val_accuracy: 0.9326
-Epoch 500/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9991 - val_loss: 0.3885 - val_accuracy: 0.9324
-Epoch 501/1000
-60000/60000 - 6s - loss: 0.0095 - accuracy: 0.9991 - val_loss: 0.3890 - val_accuracy: 0.9321
-Epoch 502/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9991 - val_loss: 0.3891 - val_accuracy: 0.9317
-Epoch 503/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9992 - val_loss: 0.3896 - val_accuracy: 0.9322
-Epoch 504/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9992 - val_loss: 0.3905 - val_accuracy: 0.9319
-Epoch 505/1000
-60000/60000 - 6s - loss: 0.0094 - accuracy: 0.9992 - val_loss: 0.3898 - val_accuracy: 0.9317
-Epoch 506/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9991 - val_loss: 0.3911 - val_accuracy: 0.9317
-Epoch 507/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9991 - val_loss: 0.3900 - val_accuracy: 0.9326
-Epoch 508/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9992 - val_loss: 0.3906 - val_accuracy: 0.9317
-Epoch 509/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9991 - val_loss: 0.3901 - val_accuracy: 0.9321
-Epoch 510/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9992 - val_loss: 0.3919 - val_accuracy: 0.9320
-Epoch 511/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9992 - val_loss: 0.3904 - val_accuracy: 0.9319
-Epoch 512/1000
-60000/60000 - 6s - loss: 0.0093 - accuracy: 0.9991 - val_loss: 0.3908 - val_accuracy: 0.9320
-Epoch 513/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9992 - val_loss: 0.3916 - val_accuracy: 0.9322
-Epoch 514/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9991 - val_loss: 0.3918 - val_accuracy: 0.9319
-Epoch 515/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9992 - val_loss: 0.3920 - val_accuracy: 0.9321
-Epoch 516/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9992 - val_loss: 0.3919 - val_accuracy: 0.9321
-Epoch 517/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9992 - val_loss: 0.3921 - val_accuracy: 0.9316
-Epoch 518/1000
-60000/60000 - 6s - loss: 0.0092 - accuracy: 0.9992 - val_loss: 0.3930 - val_accuracy: 0.9326
-Epoch 519/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9992 - val_loss: 0.3940 - val_accuracy: 0.9316
-Epoch 520/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9991 - val_loss: 0.3930 - val_accuracy: 0.9311
-Epoch 521/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9992 - val_loss: 0.3936 - val_accuracy: 0.9323
-Epoch 522/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9992 - val_loss: 0.3937 - val_accuracy: 0.9324
-Epoch 523/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9992 - val_loss: 0.3943 - val_accuracy: 0.9316
-Epoch 524/1000
-60000/60000 - 6s - loss: 0.0091 - accuracy: 0.9992 - val_loss: 0.3942 - val_accuracy: 0.9317
-Epoch 525/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9992 - val_loss: 0.3937 - val_accuracy: 0.9316
-Epoch 526/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9992 - val_loss: 0.3936 - val_accuracy: 0.9318
-Epoch 527/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9992 - val_loss: 0.3947 - val_accuracy: 0.9319
-Epoch 528/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9992 - val_loss: 0.3944 - val_accuracy: 0.9309
-Epoch 529/1000
-60000/60000 - 6s - loss: 0.0090 - accuracy: 0.9992 - val_loss: 0.3956 - val_accuracy: 0.9315
-Epoch 530/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9992 - val_loss: 0.3950 - val_accuracy: 0.9314
-Epoch 531/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9992 - val_loss: 0.3966 - val_accuracy: 0.9319
-Epoch 532/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9992 - val_loss: 0.3951 - val_accuracy: 0.9318
-Epoch 533/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9992 - val_loss: 0.3952 - val_accuracy: 0.9319
-Epoch 534/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9992 - val_loss: 0.3952 - val_accuracy: 0.9320
-Epoch 535/1000
-60000/60000 - 6s - loss: 0.0089 - accuracy: 0.9992 - val_loss: 0.3952 - val_accuracy: 0.9323
-Epoch 536/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3964 - val_accuracy: 0.9313
-Epoch 537/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3971 - val_accuracy: 0.9318
-Epoch 538/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3972 - val_accuracy: 0.9322
-Epoch 539/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3969 - val_accuracy: 0.9321
-Epoch 540/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3979 - val_accuracy: 0.9317
-Epoch 541/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3975 - val_accuracy: 0.9317
-Epoch 542/1000
-60000/60000 - 6s - loss: 0.0088 - accuracy: 0.9992 - val_loss: 0.3976 - val_accuracy: 0.9313
-Epoch 543/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9992 - val_loss: 0.3990 - val_accuracy: 0.9317
-Epoch 544/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9992 - val_loss: 0.3980 - val_accuracy: 0.9320
-Epoch 545/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9992 - val_loss: 0.3973 - val_accuracy: 0.9320
-Epoch 546/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9992 - val_loss: 0.3977 - val_accuracy: 0.9322
-Epoch 547/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9993 - val_loss: 0.3986 - val_accuracy: 0.9323
-Epoch 548/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9992 - val_loss: 0.3991 - val_accuracy: 0.9319
-Epoch 549/1000
-60000/60000 - 6s - loss: 0.0087 - accuracy: 0.9992 - val_loss: 0.3980 - val_accuracy: 0.9311
-Epoch 550/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9993 - val_loss: 0.3999 - val_accuracy: 0.9316
-Epoch 551/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9992 - val_loss: 0.3991 - val_accuracy: 0.9318
-Epoch 552/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9993 - val_loss: 0.3989 - val_accuracy: 0.9321
-Epoch 553/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9992 - val_loss: 0.3995 - val_accuracy: 0.9317
-Epoch 554/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9992 - val_loss: 0.3997 - val_accuracy: 0.9315
-Epoch 555/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9992 - val_loss: 0.3995 - val_accuracy: 0.9318
-Epoch 556/1000
-60000/60000 - 6s - loss: 0.0086 - accuracy: 0.9993 - val_loss: 0.3998 - val_accuracy: 0.9314
-Epoch 557/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9992 - val_loss: 0.3997 - val_accuracy: 0.9318
-Epoch 558/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9992 - val_loss: 0.4005 - val_accuracy: 0.9318
-Epoch 559/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9992 - val_loss: 0.4007 - val_accuracy: 0.9315
-Epoch 560/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9992 - val_loss: 0.4003 - val_accuracy: 0.9309
-Epoch 561/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9993 - val_loss: 0.4014 - val_accuracy: 0.9316
-Epoch 562/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9993 - val_loss: 0.4018 - val_accuracy: 0.9315
-Epoch 563/1000
-60000/60000 - 6s - loss: 0.0085 - accuracy: 0.9993 - val_loss: 0.4022 - val_accuracy: 0.9323
-Epoch 564/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9992 - val_loss: 0.4019 - val_accuracy: 0.9313
-Epoch 565/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9992 - val_loss: 0.4020 - val_accuracy: 0.9323
-Epoch 566/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9993 - val_loss: 0.4022 - val_accuracy: 0.9311
-Epoch 567/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9992 - val_loss: 0.4017 - val_accuracy: 0.9313
-Epoch 568/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9993 - val_loss: 0.4027 - val_accuracy: 0.9315
-Epoch 569/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9993 - val_loss: 0.4021 - val_accuracy: 0.9312
-Epoch 570/1000
-60000/60000 - 6s - loss: 0.0084 - accuracy: 0.9993 - val_loss: 0.4028 - val_accuracy: 0.9309
-Epoch 571/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9993 - val_loss: 0.4034 - val_accuracy: 0.9316
-Epoch 572/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9993 - val_loss: 0.4035 - val_accuracy: 0.9316
-Epoch 573/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9992 - val_loss: 0.4022 - val_accuracy: 0.9314
-Epoch 574/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9993 - val_loss: 0.4034 - val_accuracy: 0.9317
-Epoch 575/1000
-60000/60000 - 6s - loss: 0.0083 - accuracy: 0.9993 - val_loss: 0.4044 - val_accuracy: 0.9316
-Epoch 576/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4034 - val_accuracy: 0.9317
-Epoch 577/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4042 - val_accuracy: 0.9315
-Epoch 578/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4040 - val_accuracy: 0.9317
-Epoch 579/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4042 - val_accuracy: 0.9317
-Epoch 580/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4046 - val_accuracy: 0.9312
-Epoch 581/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9992 - val_loss: 0.4055 - val_accuracy: 0.9317
-Epoch 582/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4054 - val_accuracy: 0.9317
-Epoch 583/1000
-60000/60000 - 6s - loss: 0.0082 - accuracy: 0.9993 - val_loss: 0.4050 - val_accuracy: 0.9318
-Epoch 584/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9993 - val_loss: 0.4053 - val_accuracy: 0.9315
-Epoch 585/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9993 - val_loss: 0.4067 - val_accuracy: 0.9321
-Epoch 586/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9993 - val_loss: 0.4066 - val_accuracy: 0.9317
-Epoch 587/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9993 - val_loss: 0.4065 - val_accuracy: 0.9310
-Epoch 588/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4061 - val_accuracy: 0.9313
-Epoch 589/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9993 - val_loss: 0.4065 - val_accuracy: 0.9313
-Epoch 590/1000
-60000/60000 - 6s - loss: 0.0081 - accuracy: 0.9993 - val_loss: 0.4068 - val_accuracy: 0.9311
-Epoch 591/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4066 - val_accuracy: 0.9318
-Epoch 592/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4078 - val_accuracy: 0.9311
-Epoch 593/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4068 - val_accuracy: 0.9308
-Epoch 594/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4073 - val_accuracy: 0.9313
-Epoch 595/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4069 - val_accuracy: 0.9317
-Epoch 596/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4070 - val_accuracy: 0.9314
-Epoch 597/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4067 - val_accuracy: 0.9313
-Epoch 598/1000
-60000/60000 - 6s - loss: 0.0080 - accuracy: 0.9993 - val_loss: 0.4088 - val_accuracy: 0.9312
-Epoch 599/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4086 - val_accuracy: 0.9311
-Epoch 600/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4081 - val_accuracy: 0.9314
-Epoch 601/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4072 - val_accuracy: 0.9310
-Epoch 602/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4090 - val_accuracy: 0.9313
-Epoch 603/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4086 - val_accuracy: 0.9308
-Epoch 604/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4077 - val_accuracy: 0.9319
-Epoch 605/1000
-60000/60000 - 6s - loss: 0.0079 - accuracy: 0.9993 - val_loss: 0.4094 - val_accuracy: 0.9312
-Epoch 606/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4092 - val_accuracy: 0.9315
-Epoch 607/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4099 - val_accuracy: 0.9316
-Epoch 608/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4090 - val_accuracy: 0.9316
-Epoch 609/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4105 - val_accuracy: 0.9306
-Epoch 610/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4094 - val_accuracy: 0.9317
-Epoch 611/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4100 - val_accuracy: 0.9309
-Epoch 612/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4104 - val_accuracy: 0.9310
-Epoch 613/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4102 - val_accuracy: 0.9312
-Epoch 614/1000
-60000/60000 - 6s - loss: 0.0078 - accuracy: 0.9993 - val_loss: 0.4109 - val_accuracy: 0.9316
-Epoch 615/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4108 - val_accuracy: 0.9314
-Epoch 616/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4111 - val_accuracy: 0.9316
-Epoch 617/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4111 - val_accuracy: 0.9312
-Epoch 618/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4112 - val_accuracy: 0.9321
-Epoch 619/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4120 - val_accuracy: 0.9312
-Epoch 620/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4116 - val_accuracy: 0.9307
-Epoch 621/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4119 - val_accuracy: 0.9315
-Epoch 622/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4125 - val_accuracy: 0.9318
-Epoch 623/1000
-60000/60000 - 6s - loss: 0.0077 - accuracy: 0.9993 - val_loss: 0.4124 - val_accuracy: 0.9321
-Epoch 624/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4117 - val_accuracy: 0.9310
-Epoch 625/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4124 - val_accuracy: 0.9318
-Epoch 626/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4131 - val_accuracy: 0.9310
-Epoch 627/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4144 - val_accuracy: 0.9316
-Epoch 628/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4135 - val_accuracy: 0.9312
-Epoch 629/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4128 - val_accuracy: 0.9320
-Epoch 630/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4138 - val_accuracy: 0.9316
-Epoch 631/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4134 - val_accuracy: 0.9315
-Epoch 632/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4142 - val_accuracy: 0.9316
-Epoch 633/1000
-60000/60000 - 6s - loss: 0.0076 - accuracy: 0.9993 - val_loss: 0.4136 - val_accuracy: 0.9308
-Epoch 634/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4141 - val_accuracy: 0.9315
-Epoch 635/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4138 - val_accuracy: 0.9312
-Epoch 636/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4143 - val_accuracy: 0.9318
-Epoch 637/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4157 - val_accuracy: 0.9309
-Epoch 638/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4143 - val_accuracy: 0.9317
-Epoch 639/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4147 - val_accuracy: 0.9315
-Epoch 640/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4152 - val_accuracy: 0.9316
-Epoch 641/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4151 - val_accuracy: 0.9311
-Epoch 642/1000
-60000/60000 - 6s - loss: 0.0075 - accuracy: 0.9993 - val_loss: 0.4162 - val_accuracy: 0.9313
-Epoch 643/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4159 - val_accuracy: 0.9314
-Epoch 644/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4159 - val_accuracy: 0.9316
-Epoch 645/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4158 - val_accuracy: 0.9312
-Epoch 646/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4155 - val_accuracy: 0.9313
-Epoch 647/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4168 - val_accuracy: 0.9320
-Epoch 648/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4163 - val_accuracy: 0.9318
-Epoch 649/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9993 - val_loss: 0.4163 - val_accuracy: 0.9315
-Epoch 650/1000
-60000/60000 - 6s - loss: 0.0074 - accuracy: 0.9994 - val_loss: 0.4171 - val_accuracy: 0.9313
-Epoch 651/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9993 - val_loss: 0.4168 - val_accuracy: 0.9321
-Epoch 652/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9994 - val_loss: 0.4172 - val_accuracy: 0.9319
-Epoch 653/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9993 - val_loss: 0.4177 - val_accuracy: 0.9318
-Epoch 654/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9994 - val_loss: 0.4174 - val_accuracy: 0.9313
-Epoch 655/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9993 - val_loss: 0.4177 - val_accuracy: 0.9309
-Epoch 656/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9993 - val_loss: 0.4176 - val_accuracy: 0.9309
-Epoch 657/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9994 - val_loss: 0.4174 - val_accuracy: 0.9315
-Epoch 658/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9993 - val_loss: 0.4184 - val_accuracy: 0.9316
-Epoch 659/1000
-60000/60000 - 6s - loss: 0.0073 - accuracy: 0.9994 - val_loss: 0.4182 - val_accuracy: 0.9310
-Epoch 660/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9993 - val_loss: 0.4179 - val_accuracy: 0.9313
-Epoch 661/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9993 - val_loss: 0.4174 - val_accuracy: 0.9316
-Epoch 662/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9994 - val_loss: 0.4191 - val_accuracy: 0.9314
-Epoch 663/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9993 - val_loss: 0.4177 - val_accuracy: 0.9313
-Epoch 664/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9994 - val_loss: 0.4189 - val_accuracy: 0.9315
-Epoch 665/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9994 - val_loss: 0.4184 - val_accuracy: 0.9317
-Epoch 666/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9994 - val_loss: 0.4197 - val_accuracy: 0.9313
-Epoch 667/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9994 - val_loss: 0.4194 - val_accuracy: 0.9320
-Epoch 668/1000
-60000/60000 - 6s - loss: 0.0072 - accuracy: 0.9994 - val_loss: 0.4191 - val_accuracy: 0.9311
-Epoch 669/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4188 - val_accuracy: 0.9310
-Epoch 670/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4197 - val_accuracy: 0.9315
-Epoch 671/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9993 - val_loss: 0.4198 - val_accuracy: 0.9314
-Epoch 672/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4201 - val_accuracy: 0.9315
-Epoch 673/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4198 - val_accuracy: 0.9309
-Epoch 674/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4210 - val_accuracy: 0.9320
-Epoch 675/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4218 - val_accuracy: 0.9311
-Epoch 676/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4204 - val_accuracy: 0.9310
-Epoch 677/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4206 - val_accuracy: 0.9306
-Epoch 678/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4205 - val_accuracy: 0.9318
-Epoch 679/1000
-60000/60000 - 6s - loss: 0.0071 - accuracy: 0.9994 - val_loss: 0.4218 - val_accuracy: 0.9312
-Epoch 680/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4215 - val_accuracy: 0.9307
-Epoch 681/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4217 - val_accuracy: 0.9314
-Epoch 682/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4222 - val_accuracy: 0.9311
-Epoch 683/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4219 - val_accuracy: 0.9308
-Epoch 684/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4228 - val_accuracy: 0.9313
-Epoch 685/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4233 - val_accuracy: 0.9310
-Epoch 686/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4221 - val_accuracy: 0.9310
-Epoch 687/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4229 - val_accuracy: 0.9316
-Epoch 688/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4224 - val_accuracy: 0.9310
-Epoch 689/1000
-60000/60000 - 6s - loss: 0.0070 - accuracy: 0.9994 - val_loss: 0.4233 - val_accuracy: 0.9312
-Epoch 690/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4232 - val_accuracy: 0.9314
-Epoch 691/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4229 - val_accuracy: 0.9312
-Epoch 692/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4240 - val_accuracy: 0.9303
-Epoch 693/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4239 - val_accuracy: 0.9302
-Epoch 694/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4240 - val_accuracy: 0.9314
-Epoch 695/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4246 - val_accuracy: 0.9314
-Epoch 696/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4248 - val_accuracy: 0.9314
-Epoch 697/1000
-60000/60000 - 6s - loss: 0.0069 - accuracy: 0.9994 - val_loss: 0.4244 - val_accuracy: 0.9310
-Epoch 698/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4239 - val_accuracy: 0.9311
-Epoch 699/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4252 - val_accuracy: 0.9308
-Epoch 700/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4254 - val_accuracy: 0.9309
-Epoch 701/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4252 - val_accuracy: 0.9310
-Epoch 702/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4252 - val_accuracy: 0.9313
-Epoch 703/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4247 - val_accuracy: 0.9311
-Epoch 704/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4248 - val_accuracy: 0.9315
-Epoch 705/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4258 - val_accuracy: 0.9309
-Epoch 706/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4253 - val_accuracy: 0.9311
-Epoch 707/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4265 - val_accuracy: 0.9301
-Epoch 708/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4261 - val_accuracy: 0.9306
-Epoch 709/1000
-60000/60000 - 6s - loss: 0.0068 - accuracy: 0.9994 - val_loss: 0.4264 - val_accuracy: 0.9300
-Epoch 710/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4269 - val_accuracy: 0.9310
-Epoch 711/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4270 - val_accuracy: 0.9301
-Epoch 712/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4270 - val_accuracy: 0.9306
-Epoch 713/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4263 - val_accuracy: 0.9308
-Epoch 714/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4287 - val_accuracy: 0.9308
-Epoch 715/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4269 - val_accuracy: 0.9305
-Epoch 716/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4274 - val_accuracy: 0.9303
-Epoch 717/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4278 - val_accuracy: 0.9308
-Epoch 718/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4279 - val_accuracy: 0.9308
-Epoch 719/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4285 - val_accuracy: 0.9305
-Epoch 720/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4283 - val_accuracy: 0.9307
-Epoch 721/1000
-60000/60000 - 6s - loss: 0.0067 - accuracy: 0.9994 - val_loss: 0.4293 - val_accuracy: 0.9300
-Epoch 722/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4279 - val_accuracy: 0.9306
-Epoch 723/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4292 - val_accuracy: 0.9307
-Epoch 724/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4297 - val_accuracy: 0.9305
-Epoch 725/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4292 - val_accuracy: 0.9297
-Epoch 726/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4298 - val_accuracy: 0.9305
-Epoch 727/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4291 - val_accuracy: 0.9304
-Epoch 728/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4293 - val_accuracy: 0.9305
-Epoch 729/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4298 - val_accuracy: 0.9304
-Epoch 730/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4302 - val_accuracy: 0.9301
-Epoch 731/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4309 - val_accuracy: 0.9304
-Epoch 732/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9994 - val_loss: 0.4304 - val_accuracy: 0.9303
-Epoch 733/1000
-60000/60000 - 6s - loss: 0.0066 - accuracy: 0.9995 - val_loss: 0.4297 - val_accuracy: 0.9303
-Epoch 734/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4310 - val_accuracy: 0.9303
-Epoch 735/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4308 - val_accuracy: 0.9300
-Epoch 736/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9995 - val_loss: 0.4309 - val_accuracy: 0.9303
-Epoch 737/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9995 - val_loss: 0.4316 - val_accuracy: 0.9308
-Epoch 738/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4316 - val_accuracy: 0.9305
-Epoch 739/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4313 - val_accuracy: 0.9301
-Epoch 740/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4307 - val_accuracy: 0.9304
-Epoch 741/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9995 - val_loss: 0.4311 - val_accuracy: 0.9303
-Epoch 742/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4316 - val_accuracy: 0.9304
-Epoch 743/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4329 - val_accuracy: 0.9301
-Epoch 744/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9994 - val_loss: 0.4316 - val_accuracy: 0.9301
-Epoch 745/1000
-60000/60000 - 6s - loss: 0.0065 - accuracy: 0.9995 - val_loss: 0.4328 - val_accuracy: 0.9305
-Epoch 746/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.4314 - val_accuracy: 0.9305
-Epoch 747/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4322 - val_accuracy: 0.9303
-Epoch 748/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.4327 - val_accuracy: 0.9300
-Epoch 749/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4322 - val_accuracy: 0.9298
-Epoch 750/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.4328 - val_accuracy: 0.9304
-Epoch 751/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4325 - val_accuracy: 0.9301
-Epoch 752/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4330 - val_accuracy: 0.9298
-Epoch 753/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.4336 - val_accuracy: 0.9297
-Epoch 754/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.4334 - val_accuracy: 0.9298
-Epoch 755/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4332 - val_accuracy: 0.9303
-Epoch 756/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9994 - val_loss: 0.4341 - val_accuracy: 0.9308
-Epoch 757/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4334 - val_accuracy: 0.9303
-Epoch 758/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4332 - val_accuracy: 0.9300
-Epoch 759/1000
-60000/60000 - 6s - loss: 0.0064 - accuracy: 0.9995 - val_loss: 0.4337 - val_accuracy: 0.9298
-Epoch 760/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4346 - val_accuracy: 0.9300
-Epoch 761/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4345 - val_accuracy: 0.9303
-Epoch 762/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4353 - val_accuracy: 0.9299
-Epoch 763/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4356 - val_accuracy: 0.9299
-Epoch 764/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4346 - val_accuracy: 0.9302
-Epoch 765/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4356 - val_accuracy: 0.9296
-Epoch 766/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4363 - val_accuracy: 0.9304
-Epoch 767/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4352 - val_accuracy: 0.9302
-Epoch 768/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9994 - val_loss: 0.4362 - val_accuracy: 0.9299
-Epoch 769/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4350 - val_accuracy: 0.9302
-Epoch 770/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4365 - val_accuracy: 0.9303
-Epoch 771/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9994 - val_loss: 0.4366 - val_accuracy: 0.9295
-Epoch 772/1000
-60000/60000 - 6s - loss: 0.0063 - accuracy: 0.9995 - val_loss: 0.4368 - val_accuracy: 0.9300
-Epoch 773/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9994 - val_loss: 0.4359 - val_accuracy: 0.9301
-Epoch 774/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4374 - val_accuracy: 0.9299
-Epoch 775/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4364 - val_accuracy: 0.9303
-Epoch 776/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4371 - val_accuracy: 0.9298
-Epoch 777/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4375 - val_accuracy: 0.9304
-Epoch 778/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4372 - val_accuracy: 0.9304
-Epoch 779/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4376 - val_accuracy: 0.9302
-Epoch 780/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4374 - val_accuracy: 0.9303
-Epoch 781/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4376 - val_accuracy: 0.9299
-Epoch 782/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4376 - val_accuracy: 0.9300
-Epoch 783/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4378 - val_accuracy: 0.9306
-Epoch 784/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4383 - val_accuracy: 0.9300
-Epoch 785/1000
-60000/60000 - 6s - loss: 0.0062 - accuracy: 0.9995 - val_loss: 0.4381 - val_accuracy: 0.9296
-Epoch 786/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9994 - val_loss: 0.4379 - val_accuracy: 0.9299
-Epoch 787/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4378 - val_accuracy: 0.9302
-Epoch 788/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4400 - val_accuracy: 0.9299
-Epoch 789/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9994 - val_loss: 0.4389 - val_accuracy: 0.9300
-Epoch 790/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4385 - val_accuracy: 0.9302
-Epoch 791/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4378 - val_accuracy: 0.9298
-Epoch 792/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4394 - val_accuracy: 0.9295
-Epoch 793/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4394 - val_accuracy: 0.9298
-Epoch 794/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4394 - val_accuracy: 0.9302
-Epoch 795/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4396 - val_accuracy: 0.9303
-Epoch 796/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4392 - val_accuracy: 0.9299
-Epoch 797/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4398 - val_accuracy: 0.9307
-Epoch 798/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4400 - val_accuracy: 0.9301
-Epoch 799/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4406 - val_accuracy: 0.9305
-Epoch 800/1000
-60000/60000 - 6s - loss: 0.0061 - accuracy: 0.9995 - val_loss: 0.4402 - val_accuracy: 0.9300
-Epoch 801/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4408 - val_accuracy: 0.9306
-Epoch 802/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4407 - val_accuracy: 0.9302
-Epoch 803/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4407 - val_accuracy: 0.9302
-Epoch 804/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4408 - val_accuracy: 0.9300
-Epoch 805/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4419 - val_accuracy: 0.9298
-Epoch 806/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4418 - val_accuracy: 0.9298
-Epoch 807/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4418 - val_accuracy: 0.9297
-Epoch 808/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4421 - val_accuracy: 0.9300
-Epoch 809/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4422 - val_accuracy: 0.9301
-Epoch 810/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4422 - val_accuracy: 0.9301
-Epoch 811/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4425 - val_accuracy: 0.9302
-Epoch 812/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4424 - val_accuracy: 0.9305
-Epoch 813/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4429 - val_accuracy: 0.9302
-Epoch 814/1000
-60000/60000 - 6s - loss: 0.0060 - accuracy: 0.9995 - val_loss: 0.4424 - val_accuracy: 0.9305
-Epoch 815/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4430 - val_accuracy: 0.9297
-Epoch 816/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4432 - val_accuracy: 0.9301
-Epoch 817/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4436 - val_accuracy: 0.9297
-Epoch 818/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4430 - val_accuracy: 0.9300
-Epoch 819/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4438 - val_accuracy: 0.9300
-Epoch 820/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4432 - val_accuracy: 0.9299
-Epoch 821/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4441 - val_accuracy: 0.9298
-Epoch 822/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4449 - val_accuracy: 0.9300
-Epoch 823/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4442 - val_accuracy: 0.9304
-Epoch 824/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4442 - val_accuracy: 0.9304
-Epoch 825/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4447 - val_accuracy: 0.9298
-Epoch 826/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4447 - val_accuracy: 0.9303
-Epoch 827/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4449 - val_accuracy: 0.9302
-Epoch 828/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4443 - val_accuracy: 0.9302
-Epoch 829/1000
-60000/60000 - 6s - loss: 0.0059 - accuracy: 0.9995 - val_loss: 0.4448 - val_accuracy: 0.9303
-Epoch 830/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4450 - val_accuracy: 0.9308
-Epoch 831/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4453 - val_accuracy: 0.9302
-Epoch 832/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4447 - val_accuracy: 0.9305
-Epoch 833/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4448 - val_accuracy: 0.9300
-Epoch 834/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4461 - val_accuracy: 0.9301
-Epoch 835/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4457 - val_accuracy: 0.9305
-Epoch 836/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4455 - val_accuracy: 0.9303
-Epoch 837/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4457 - val_accuracy: 0.9299
-Epoch 838/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4461 - val_accuracy: 0.9308
-Epoch 839/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4460 - val_accuracy: 0.9307
-Epoch 840/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4472 - val_accuracy: 0.9299
-Epoch 841/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4469 - val_accuracy: 0.9300
-Epoch 842/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4463 - val_accuracy: 0.9306
-Epoch 843/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4474 - val_accuracy: 0.9309
-Epoch 844/1000
-60000/60000 - 6s - loss: 0.0058 - accuracy: 0.9995 - val_loss: 0.4478 - val_accuracy: 0.9307
-Epoch 845/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4468 - val_accuracy: 0.9309
-Epoch 846/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4473 - val_accuracy: 0.9302
-Epoch 847/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4469 - val_accuracy: 0.9301
-Epoch 848/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4479 - val_accuracy: 0.9303
-Epoch 849/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4473 - val_accuracy: 0.9299
-Epoch 850/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4481 - val_accuracy: 0.9302
-Epoch 851/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4472 - val_accuracy: 0.9303
-Epoch 852/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4478 - val_accuracy: 0.9306
-Epoch 853/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4477 - val_accuracy: 0.9305
-Epoch 854/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4482 - val_accuracy: 0.9300
-Epoch 855/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4481 - val_accuracy: 0.9306
-Epoch 856/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4489 - val_accuracy: 0.9301
-Epoch 857/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4489 - val_accuracy: 0.9303
-Epoch 858/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4491 - val_accuracy: 0.9306
-Epoch 859/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4487 - val_accuracy: 0.9302
-Epoch 860/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4492 - val_accuracy: 0.9303
-Epoch 861/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4498 - val_accuracy: 0.9303
-Epoch 862/1000
-60000/60000 - 6s - loss: 0.0057 - accuracy: 0.9995 - val_loss: 0.4493 - val_accuracy: 0.9305
-Epoch 863/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4492 - val_accuracy: 0.9305
-Epoch 864/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4494 - val_accuracy: 0.9297
-Epoch 865/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4492 - val_accuracy: 0.9301
-Epoch 866/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4498 - val_accuracy: 0.9300
-Epoch 867/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4494 - val_accuracy: 0.9296
-Epoch 868/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4500 - val_accuracy: 0.9302
-Epoch 869/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4499 - val_accuracy: 0.9299
-Epoch 870/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4509 - val_accuracy: 0.9300
-Epoch 871/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4507 - val_accuracy: 0.9303
-Epoch 872/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4503 - val_accuracy: 0.9302
-Epoch 873/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4504 - val_accuracy: 0.9300
-Epoch 874/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4512 - val_accuracy: 0.9303
-Epoch 875/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4505 - val_accuracy: 0.9300
-Epoch 876/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4509 - val_accuracy: 0.9306
-Epoch 877/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4514 - val_accuracy: 0.9304
-Epoch 878/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4508 - val_accuracy: 0.9302
-Epoch 879/1000
-60000/60000 - 6s - loss: 0.0056 - accuracy: 0.9995 - val_loss: 0.4507 - val_accuracy: 0.9297
-Epoch 880/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4519 - val_accuracy: 0.9300
-Epoch 881/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4516 - val_accuracy: 0.9302
-Epoch 882/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4523 - val_accuracy: 0.9297
-Epoch 883/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4521 - val_accuracy: 0.9300
-Epoch 884/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4526 - val_accuracy: 0.9293
-Epoch 885/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4524 - val_accuracy: 0.9301
-Epoch 886/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4523 - val_accuracy: 0.9303
-Epoch 887/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4530 - val_accuracy: 0.9300
-Epoch 888/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4527 - val_accuracy: 0.9300
-Epoch 889/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4530 - val_accuracy: 0.9303
-Epoch 890/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4530 - val_accuracy: 0.9301
-Epoch 891/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4528 - val_accuracy: 0.9300
-Epoch 892/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4531 - val_accuracy: 0.9300
-Epoch 893/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4543 - val_accuracy: 0.9300
-Epoch 894/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4529 - val_accuracy: 0.9298
-Epoch 895/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4531 - val_accuracy: 0.9300
-Epoch 896/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4527 - val_accuracy: 0.9303
-Epoch 897/1000
-60000/60000 - 6s - loss: 0.0055 - accuracy: 0.9995 - val_loss: 0.4537 - val_accuracy: 0.9306
-Epoch 898/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4538 - val_accuracy: 0.9300
-Epoch 899/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4537 - val_accuracy: 0.9303
-Epoch 900/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4549 - val_accuracy: 0.9299
-Epoch 901/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4546 - val_accuracy: 0.9300
-Epoch 902/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4548 - val_accuracy: 0.9294
-Epoch 903/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4544 - val_accuracy: 0.9296
-Epoch 904/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4549 - val_accuracy: 0.9301
-Epoch 905/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4542 - val_accuracy: 0.9303
-Epoch 906/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4545 - val_accuracy: 0.9299
-Epoch 907/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4542 - val_accuracy: 0.9300
-Epoch 908/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4545 - val_accuracy: 0.9306
-Epoch 909/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4549 - val_accuracy: 0.9306
-Epoch 910/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4550 - val_accuracy: 0.9303
-Epoch 911/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4557 - val_accuracy: 0.9304
-Epoch 912/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4549 - val_accuracy: 0.9302
-Epoch 913/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4556 - val_accuracy: 0.9302
-Epoch 914/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4555 - val_accuracy: 0.9300
-Epoch 915/1000
-60000/60000 - 6s - loss: 0.0054 - accuracy: 0.9995 - val_loss: 0.4560 - val_accuracy: 0.9307
-Epoch 916/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4549 - val_accuracy: 0.9302
-Epoch 917/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4552 - val_accuracy: 0.9299
-Epoch 918/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4557 - val_accuracy: 0.9301
-Epoch 919/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4561 - val_accuracy: 0.9302
-Epoch 920/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4568 - val_accuracy: 0.9300
-Epoch 921/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4559 - val_accuracy: 0.9301
-Epoch 922/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4563 - val_accuracy: 0.9300
-Epoch 923/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4574 - val_accuracy: 0.9302
-Epoch 924/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4569 - val_accuracy: 0.9301
-Epoch 925/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4568 - val_accuracy: 0.9297
-Epoch 926/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4576 - val_accuracy: 0.9295
-Epoch 927/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4576 - val_accuracy: 0.9300
-Epoch 928/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4584 - val_accuracy: 0.9304
-Epoch 929/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4576 - val_accuracy: 0.9297
-Epoch 930/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4569 - val_accuracy: 0.9294
-Epoch 931/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4578 - val_accuracy: 0.9303
-Epoch 932/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4578 - val_accuracy: 0.9301
-Epoch 933/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4587 - val_accuracy: 0.9298
-Epoch 934/1000
-60000/60000 - 6s - loss: 0.0053 - accuracy: 0.9995 - val_loss: 0.4578 - val_accuracy: 0.9298
-Epoch 935/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4585 - val_accuracy: 0.9302
-Epoch 936/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4582 - val_accuracy: 0.9301
-Epoch 937/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4586 - val_accuracy: 0.9298
-Epoch 938/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4583 - val_accuracy: 0.9297
-Epoch 939/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4591 - val_accuracy: 0.9299
-Epoch 940/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4585 - val_accuracy: 0.9301
-Epoch 941/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4591 - val_accuracy: 0.9294
-Epoch 942/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4597 - val_accuracy: 0.9302
-Epoch 943/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4594 - val_accuracy: 0.9297
-Epoch 944/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4593 - val_accuracy: 0.9299
-Epoch 945/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4603 - val_accuracy: 0.9296
-Epoch 946/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4594 - val_accuracy: 0.9296
-Epoch 947/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4599 - val_accuracy: 0.9295
-Epoch 948/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4597 - val_accuracy: 0.9295
-Epoch 949/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4594 - val_accuracy: 0.9299
-Epoch 950/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4604 - val_accuracy: 0.9296
-Epoch 951/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4605 - val_accuracy: 0.9295
-Epoch 952/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4606 - val_accuracy: 0.9295
-Epoch 953/1000
-60000/60000 - 6s - loss: 0.0052 - accuracy: 0.9995 - val_loss: 0.4601 - val_accuracy: 0.9292
-Epoch 954/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4595 - val_accuracy: 0.9297
-Epoch 955/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4614 - val_accuracy: 0.9300
-Epoch 956/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4610 - val_accuracy: 0.9300
-Epoch 957/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4613 - val_accuracy: 0.9296
-Epoch 958/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4608 - val_accuracy: 0.9299
-Epoch 959/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4609 - val_accuracy: 0.9298
-Epoch 960/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4619 - val_accuracy: 0.9295
-Epoch 961/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4614 - val_accuracy: 0.9295
-Epoch 962/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4611 - val_accuracy: 0.9296
-Epoch 963/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4615 - val_accuracy: 0.9293
-Epoch 964/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4620 - val_accuracy: 0.9296
-Epoch 965/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4617 - val_accuracy: 0.9290
-Epoch 966/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4631 - val_accuracy: 0.9295
-Epoch 967/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4613 - val_accuracy: 0.9297
-Epoch 968/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4622 - val_accuracy: 0.9293
-Epoch 969/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4626 - val_accuracy: 0.9296
-Epoch 970/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4627 - val_accuracy: 0.9297
-Epoch 971/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4627 - val_accuracy: 0.9290
-Epoch 972/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4630 - val_accuracy: 0.9292
-Epoch 973/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4632 - val_accuracy: 0.9295
-Epoch 974/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4629 - val_accuracy: 0.9296
-Epoch 975/1000
-60000/60000 - 6s - loss: 0.0051 - accuracy: 0.9995 - val_loss: 0.4632 - val_accuracy: 0.9297
-Epoch 976/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4631 - val_accuracy: 0.9291
-Epoch 977/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4632 - val_accuracy: 0.9293
-Epoch 978/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4644 - val_accuracy: 0.9299
-Epoch 979/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4642 - val_accuracy: 0.9295
-Epoch 980/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4636 - val_accuracy: 0.9288
-Epoch 981/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4643 - val_accuracy: 0.9297
-Epoch 982/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4637 - val_accuracy: 0.9298
-Epoch 983/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4643 - val_accuracy: 0.9295
-Epoch 984/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4647 - val_accuracy: 0.9294
-Epoch 985/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4648 - val_accuracy: 0.9293
-Epoch 986/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4653 - val_accuracy: 0.9296
-Epoch 987/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4650 - val_accuracy: 0.9291
-Epoch 988/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4650 - val_accuracy: 0.9294
-Epoch 989/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4640 - val_accuracy: 0.9295
-Epoch 990/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4646 - val_accuracy: 0.9297
-Epoch 991/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4645 - val_accuracy: 0.9294
-Epoch 992/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4650 - val_accuracy: 0.9294
-Epoch 993/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4651 - val_accuracy: 0.9295
-Epoch 994/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4650 - val_accuracy: 0.9295
-Epoch 995/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4647 - val_accuracy: 0.9288
-Epoch 996/1000
-60000/60000 - 6s - loss: 0.0050 - accuracy: 0.9995 - val_loss: 0.4652 - val_accuracy: 0.9295
-Epoch 997/1000
-60000/60000 - 6s - loss: 0.0049 - accuracy: 0.9995 - val_loss: 0.4659 - val_accuracy: 0.9296
-Epoch 998/1000
-60000/60000 - 6s - loss: 0.0049 - accuracy: 0.9995 - val_loss: 0.4658 - val_accuracy: 0.9295
-Epoch 999/1000
-60000/60000 - 6s - loss: 0.0049 - accuracy: 0.9995 - val_loss: 0.4657 - val_accuracy: 0.9291
-Epoch 1000/1000
-60000/60000 - 6s - loss: 0.0049 - accuracy: 0.9995 - val_loss: 0.4661 - val_accuracy: 0.9291
-Test loss was 0.1903, test accuracy was 0.9462
diff --git a/MNIST/plot.py b/MNIST/plot.py
deleted file mode 100644
index 1f93fe3..0000000
--- a/MNIST/plot.py
+++ /dev/null
@@ -1,66 +0,0 @@
-import numpy as np
-import matplotlib.pyplot as plt
-
-
-def get_accuracy(log_dir_path):
-    filename = "{}/accuracy.csv".format(log_dir_path)
-    csv = np.genfromtxt(filename, delimiter=",")
-    accuracy = 100*csv
-    return accuracy
-
-def get_error(log_dir_path):
-    accuracy = get_accuracy(log_dir_path)
-    error = 100 - accuracy
-    return error
-
-def get_power(log_dir_path):
-    filename = "{}/power.csv".format(log_dir_path)
-    csv = np.genfromtxt(filename, delimiter=",")
-    power = csv[0::2] + csv[1::2]
-    return power
-
-
-dataset = "MNIST"
-network_types = ["regular", "non-regularized-aware", "regularized-aware"]
-group_idxs = [0, 2]
-colors = ["#e69f00", "#0072b2", "#009e73"]
-labels = ["Standard", "Nonideality-aware (non-regularised)", "Nonideality-aware (regularised)"]
-
-axis_label_font_size = 14
-legend_font_size = 12
-ticks_font_size = 10
-markersize = 7
-
-fig, ax = plt.subplots()
-boxplots = []
-
-for network_idx, network_type in enumerate(network_types):
-    for group_idx in group_idxs:
-        log_dir_path = "models/{}/{}/group-{}".format(
-                dataset, network_type, group_idx)
-        power = get_power(log_dir_path)
-        error = get_error(log_dir_path)
-        print("network_type:", network_type, "group_idx:", group_idx, "median error:", np.median(error), "median power:", np.median(power))
-        # plt.scatter(power, accuracy, marker="x", s=markersize, color=colors[network_idx], label=labels[network_idx])
-        w = 0.1
-        x_pos = np.median(power)
-        boxplots.append(plt.boxplot(error, positions=[x_pos], widths=[10**(np.log10(x_pos)+w/2.)-10**(np.log10(x_pos)-w/2.)], sym=colors[network_idx]))
-        bplot = boxplots[-1]
-        plt.setp(bplot['fliers'], marker='x', markersize=4)
-        for element in ['boxes', 'whiskers', 'fliers', 'means', 'medians', 'caps']:
-            plt.setp(bplot[element], color=colors[network_idx])
-
-
-ax.legend([boxplot["boxes"][0] for boxplot in boxplots[::len(group_idxs)]],
-        [label for label in labels], fontsize=legend_font_size)
-plt.xticks(fontsize=ticks_font_size)
-plt.yticks(fontsize=ticks_font_size)
-plt.xlabel("Ohmic power consumption (W)", fontsize=axis_label_font_size)
-plt.ylabel("Inference test error (%)", fontsize=axis_label_font_size)
-plt.semilogx()
-plt.semilogy()
-# x.ticklabel_format(style='plain')
-# ax.get_yaxis().get_major_formatter().set_scientific(False)
-# plt.show()
-plt.savefig("error-box-plot-two-groups.pdf", bbox_inches='tight')
-
diff --git a/README.md b/README.md
index 18ab05f..c6eb6a9 100644
--- a/README.md
+++ b/README.md
@@ -1,19 +1,40 @@
-# Adjusting Training to Enable Accurate Low-Power Memristive Neural Networks
+# Nonideality-Aware Training for Accurate and Robust Low-Power Memristive Neural Networks
 
 ## Requirements
 
-TensorFlow 2.0 or higher.
+Python ≥3.9 and the packages listed in [requirements.txt](/requirements.txt).
 
-## Test
+## Repository structure
 
-To train, go to `MNIST`, set `path_to_project` in `MNIST/Train.py` and then run `python Train.py`.
+`awarememristor/crossbar`: memristor nonidealities and mapping onto crossbar arrays.
 
-## Repo organisation
+`awarememristor/training`: network training.
 
-`model_architectures.py`: model topology.
+`awarememristor/simulations`: simulations presented in the manuscript.
 
-`memristor_utils.py`: custom layers including `memristor_dense`.
+`awarememristor/plotting`: figures presented in the manuscript.
 
-`crossbar`: mapping and nonidealities.
+## Reproducing results
 
-`MNIST/Train.py`: training setup.
+Script [reproduce_paper.py](/reproduce_paper.py) can be used to reproduce the simulations and plots presented in the manuscript.
+Please follow the instructions in the script to obtain any missing experimental data (or comment out the function calls that require these data).
+After that, execute
+```text
+python reproduce_paper.py
+```
+
+This might take a long time to finish, so you may want to split this file up in order to, for example, perform the simulations on multiple machines.
+
+## Testing
+
+To run unit tests, execute
+```text
+pytest tests
+```
+
+## Using this package
+
+**This package should not be used in production.**
+The code is extensible but was written mostly with specific [simulations](/awarememristor/simulations) in mind.
+Any new functionality (such as different nonidealities) should be incorporated carefully.
+For example, to handle combinations of *multiple* linearity-preserving nonidealities (which is not currently supported), it may *not* be sufficient to simply apply them one after another.
diff --git a/awarememristor/__init__.py b/awarememristor/__init__.py
new file mode 100644
index 0000000..5f986ab
--- /dev/null
+++ b/awarememristor/__init__.py
@@ -0,0 +1 @@
+from awarememristor import crossbar, plotting, simulations, training
diff --git a/awarememristor/crossbar/__init__.py b/awarememristor/crossbar/__init__.py
new file mode 100644
index 0000000..7731ce1
--- /dev/null
+++ b/awarememristor/crossbar/__init__.py
@@ -0,0 +1,3 @@
+import awarememristor.crossbar.ideal
+import awarememristor.crossbar.map
+import awarememristor.crossbar.nonidealities
diff --git a/awarememristor/crossbar/ideal.py b/awarememristor/crossbar/ideal.py
new file mode 100644
index 0000000..41e620c
--- /dev/null
+++ b/awarememristor/crossbar/ideal.py
@@ -0,0 +1,33 @@
+import tensorflow as tf
+
+from awarememristor.crossbar import utils
+
+
+def compute_I_all(V: tf.Tensor, G: tf.Tensor) -> tf.Tensor:
+    """Compute output and device currents of an ideal crossbar.
+
+    Args:
+        V: Voltages of shape `p x m`.
+        G: Conductances of shape `m x n`.
+
+    Returns:
+        I: Output currents of shape `p x n`.
+        I_ind: Currents of shape `p x m x n` produced by each of the conductances in the crossbar
+            array.
+    """
+    I_ind = tf.expand_dims(V, axis=-1) * tf.expand_dims(G, axis=0)
+    I = utils.add_I_BL(I_ind)
+    return I, I_ind
+
+
+def compute_I(V: tf.Tensor, G: tf.Tensor) -> tf.Tensor:
+    """Compute output currents of an ideal crossbar.
+
+    Args:
+        V: Voltages of shape `p x m`.
+        G: Conductances of shape `m x n`.
+
+    Returns:
+        Output currents of shape `p x n`.
+    """
+    return tf.tensordot(V, G, axes=1)
diff --git a/awarememristor/crossbar/map.py b/awarememristor/crossbar/map.py
new file mode 100644
index 0000000..261d1d2
--- /dev/null
+++ b/awarememristor/crossbar/map.py
@@ -0,0 +1,145 @@
+import tensorflow as tf
+
+
+def I_to_y(I: tf.Tensor, k_V: float, max_weight: float, G_on: float, G_off: float) -> tf.Tensor:
+    """Convert output currents of a dot-product engine onto synaptic layer inputs.
+
+    Args:
+        I: Output currents of shape `p x 2n`
+        k_V: Voltage scaling factor.
+        max_weight: Assumed maximum weight.
+        G_on: Memristor conductance in ON state.
+        G_off: Memristor conductance in OFF state.
+
+    Returns:
+        Outputs of shape `p x n` of a synaptic layer implemented using memristive crossbars.
+    """
+    I_total = I[:, 0::2] - I[:, 1::2]
+    y = I_total_to_y(I_total, k_V, max_weight, G_on, G_off)
+    return y
+
+
+def I_total_to_y(
+    I_total: tf.Tensor, k_V: float, max_weight: float, G_on: float, G_off: float
+) -> tf.Tensor:
+    """Convert total output currents of a dot-product engine onto synaptic layer inputs.
+
+    Args:
+        I_total: Total output currents of shape `p x n`
+        k_V: Voltage scaling factor.
+        max_weight: Assumed maximum weight.
+        G_on: Memristor conductance in ON state.
+        G_off: Memristor conductance in OFF state.
+
+    Returns:
+        Outputs of shape `p x n` of a synaptic layer implemented using memristive crossbars.
+    """
+    k_G = _compute_k_G(max_weight, G_on, G_off)
+    k_I = _compute_k_I(k_V, k_G)
+    y = I_total / k_I
+    return y
+
+
+def _compute_k_G(max_weight: float, G_on: float, G_off: float) -> float:
+    """Compute conductance scaling factor.
+
+    Args:
+        max_weight: Assumed maximum weight.
+        G_on: Memristor conductance in ON state.
+        G_off: Memristor conductance in OFF state.
+
+    Returns:
+        Conductance scaling factor.
+    """
+    k_G = (G_on - G_off) / max_weight
+
+    return k_G
+
+
+def _compute_k_I(k_V: float, k_G: float) -> float:
+    """Compute current scaling factor.
+
+    Args:
+        k_V: Voltage scaling factor.
+        k_G: Conductance scaling factor.
+
+    Returns:
+        Current scaling factor.
+    """
+    return k_V * k_G
+
+
+def x_to_V(x: tf.Tensor, k_V: float) -> tf.Tensor:
+    """Map inputs (to a synaptic layer) onto voltages.
+
+    Args:
+        x: Synaptic inputs.
+        k_V: Voltage scaling factor.
+
+    Returns:
+        Voltages.
+    """
+    return k_V * x
+
+
+def double_w_to_G(double_w: tf.Tensor, G_off: float, G_on: float) -> tf.Tensor:
+    """Map double weights onto conductances.
+
+    Args:
+        double_w: Double weights of shape `m x 2n`. These are used to train
+            each conductance (instead of pair of conductances) directly.
+        G_off: Memristor conductance in OFF state.
+        G_on: Memristor conductance in ON state.
+
+    Returns:
+        G: Conductances of shape `m x 2n`.
+        max_weight: Assumed maximum weight.
+    """
+    max_weight = tf.math.reduce_max(double_w)
+    k_G = _compute_k_G(max_weight, G_on, G_off)
+    G = k_G * double_w + G_off
+
+    return G, max_weight
+
+
+@tf.function
+def w_to_G(
+    weights: tf.Tensor, G_off: float, G_on: float, mapping_rule: str = "default"
+) -> tf.Tensor:
+    """Map weights onto conductances.
+
+    Args:
+        weights: Weights of shape `m x n`.
+        G_off: Memristor conductance in OFF state.
+        G_on: Memristor conductance in ON state.
+        mapping_rule: One of `("default", "avg")`.
+
+    Returns:
+        G: Conductances of shape `m x 2n`.
+        max_weight: Assumed maximum weight.
+    """
+    max_weight = tf.math.reduce_max(tf.math.abs(weights))
+
+    k_G = _compute_k_G(max_weight, G_on, G_off)
+    G_eff = k_G * weights
+
+    if mapping_rule == "default":
+        # We implement the pairs by choosing the lowest possible conductances.
+        G_pos = tf.math.maximum(G_eff, 0.0) + G_off
+        G_neg = -tf.math.minimum(G_eff, 0.0) + G_off
+    elif mapping_rule == "avg":
+        # We map the pairs symmetrically around `G_avg`.
+        G_avg = (G_off + G_on) / 2
+        G_pos = G_avg + 0.5 * G_eff
+        G_neg = G_avg - 0.5 * G_eff
+    else:
+        raise ValueError(f"Mapping rule {mapping_rule} is not recognized!")
+
+    # Odd columns dedicated to positive weights.
+    # Even columns dedicated to negative weights.
+    G = tf.reshape(
+        tf.concat([G_pos[..., tf.newaxis], G_neg[..., tf.newaxis]], axis=-1),
+        [tf.shape(G_pos)[0], -1],
+    )
+
+    return G, max_weight
diff --git a/awarememristor/crossbar/nonidealities.py b/awarememristor/crossbar/nonidealities.py
new file mode 100644
index 0000000..1172cd4
--- /dev/null
+++ b/awarememristor/crossbar/nonidealities.py
@@ -0,0 +1,266 @@
+from abc import ABC, abstractmethod
+
+import numpy as np
+import tensorflow as tf
+import tensorflow_probability as tfp
+from KDEpy import bw_selection
+from tensorflow.python.ops.numpy_ops import np_config
+from tensorflow_probability import distributions as tfd
+
+from awarememristor.crossbar import utils
+
+
+class Nonideality(ABC):
+    """Physical effect that influences the behavior of memristive devices."""
+
+    @abstractmethod
+    def label(self) -> str:
+        """Returns nonideality label used in directory names, for example."""
+
+    def __eq__(self, other):
+        if self is None or other is None:
+            if self is None and other is None:
+                return True
+            return False
+        return self.label() == other.label()
+
+
+class LinearityPreserving(ABC):
+    """Nonideality whose effect can be simulated by disturbing the conductances."""
+
+    @abstractmethod
+    def disturb_G(self, G: tf.Tensor) -> tf.Tensor:
+        """Disturb conductances."""
+
+
+class LinearityNonpreserving(ABC):
+    """Nonideality in which nonlinearity manifests itself in individual devices
+    and the output current of a device is a function of its conductance
+    parameter and the voltage applied across it."""
+
+    @abstractmethod
+    def compute_I(self, V: tf.Tensor, G: tf.Tensor) -> tuple[tf.Tensor, tf.Tensor]:
+        """Compute currents in a crossbar suffering from linearity-nonpreserving nonideality.
+
+        Args:
+            V: Voltages of shape `p x m`.
+            G: Conductances of shape `m x n`.
+
+        Returns:
+            I: Output currents of shape `p x n`.
+            I_ind: Currents of shape `p x m x n` produced by each of the
+                conductances in the crossbar array.
+        """
+
+    @abstractmethod
+    def k_V(self) -> float:
+        """Return voltage scaling factor."""
+
+
+class IVNonlinearity(Nonideality, LinearityNonpreserving):
+    """Uses nonlinearity parameter to model deviations from ohmic I-V behavior.
+
+    Nonlinearity parameter `n` is defined as the current generated by a
+    resistive device at voltage `2*V` divided by the current generated by the
+    device at voltage `V`. We introduce voltage `V_ref` at which the generated
+    amount of current equals the expected amount described by Ohm's law, i.e.
+    `I = V*G`.
+    """
+
+    def __init__(self, V_ref: float, n_avg: float, n_std: float) -> None:
+        self.V_ref = V_ref
+        self.n_avg = n_avg
+        self.n_std = n_std
+
+    def label(self):
+        return f"IVNL:{self.n_avg:.3g}_{self.n_std:.3g}"
+
+    def compute_I(self, V, G):
+        # n <= 1 would produce unrealistic behaviour, while 1 < n < 2 is not typical in I-V curves
+        distribution = tfd.TruncatedNormal(
+            loc=self.n_avg,
+            scale=self.n_std,
+            low=2.0,
+            high=np.inf,
+        )
+        n = distribution.sample(sample_shape=G.get_shape().as_list())
+
+        sign = tf.expand_dims(tf.sign(V), axis=-1)
+        ohmic_current = sign * self.V_ref * tf.expand_dims(G, axis=0)
+        # Take absolute value of V to prevent negative numbers from being raised to
+        # a negative power. We assume symmetrical behaviour with negative voltages.
+        ratio = tf.expand_dims(tf.abs(V) / self.V_ref, axis=-1)
+        exponent = utils.tf_log2(n)
+
+        I_ind = ohmic_current * ratio ** exponent
+
+        I = utils.add_I_BL(I_ind)
+
+        return I, I_ind
+
+    def k_V(self):
+        return 2 * self.V_ref
+
+
+class StuckAt(Nonideality, LinearityPreserving):
+    """Models a fraction of the devices as stuck in one conductance state."""
+
+    def __init__(self, value: float, probability: float) -> None:
+        """
+        Args:
+            value: Conductance value to set randomly selected devices to.
+            probability: Probability that a given device will be set to `val`.
+                Probability must be in the [0.0, 1.0] range.
+        """
+        self.value = value
+        self.probability = probability
+
+    def label(self):
+        return f"Stuck:{self.value:.3g}_{self.probability:.3g}"
+
+    def disturb_G(self, G):
+        mask = utils.random_bool_tensor(G.shape, self.probability)
+        G = tf.where(mask, self.value, G)
+        return G
+
+
+class StuckAtGOff(StuckAt):
+    """Models a fraction of the devices as stuck at `G_off`."""
+
+    def __init__(self, G_off: float, probability: float) -> None:
+        StuckAt.__init__(self, G_off, probability)
+
+    def label(self):
+        return f"StuckOff:{self.probability:.3g}"
+
+
+class StuckAtGOn(StuckAt):
+    """Models a fraction of the devices as stuck at `G_on`."""
+
+    def __init__(self, G_on: float, probability: float) -> None:
+        StuckAt.__init__(self, G_on, probability)
+
+    def label(self):
+        return f"StuckOn:{self.probability:.3g}"
+
+
+class StuckDistribution(Nonideality, LinearityPreserving):
+    """Models a fraction of the devices as stuck at conductance states drawn
+    from a random distribution.
+
+    Kernel density estimation (KDE) with truncated normal distributions is
+    constructed using a list of conductance values at which the devices got
+    stuck.
+    """
+
+    def __init__(
+        self, means: list[float], probability: float, bandwidth_def=bw_selection.scotts_rule
+    ):
+        """
+        Args:
+            means: Means of underlying normal distributions.
+            probability: Probability that a given device will get stuck.
+            bandwidth_def: Function used to determine the bandwidth parameter of KDE.
+        """
+        bandwidth = bandwidth_def(np.reshape(means, (len(means), 1)))
+        self.probability = probability
+        self.bandwidth = bandwidth
+        self.distribution = self._kde(means, bandwidth)
+
+    def label(self) -> str:
+        return f"StuckDistr:{self.probability:.3g}_{self.bandwidth:.3g}"
+
+    @staticmethod
+    def _kde(means: list[float], bandwidth: float) -> tfd.Distribution:
+        """Kernel density estimation.
+
+        Args:
+            means: Means of underlying normal distributions.
+            bandwidth: Standard deviation of underlying normal distributions.
+
+        Returns:
+            KDE distribution.
+        """
+        weights = []
+        distr_means = []
+        for mean in means:
+            distr_means.append(mean)
+            prob_neg = tfd.Normal(loc=mean, scale=bandwidth).cdf(0.0)
+            # To ensure numerical stability, only include reflection if it will
+            # have a non-negligible effect.
+            if prob_neg > 1e-8:
+                distr_means.append(-mean)
+                prob_pos = 1.0 - prob_neg
+                weights.extend([prob_pos, prob_neg])
+            else:
+                weights.append(1.0)
+
+        np_config.enable_numpy_behavior()
+        distr_means_32 = tf.constant(distr_means)
+        bandwidth_32 = tf.constant(bandwidth)
+
+        kde_distribution = tfd.MixtureSameFamily(
+            tfd.Categorical(probs=weights),
+            tfd.TruncatedNormal(
+                loc=distr_means_32.astype(tf.float32),
+                scale=bandwidth_32.astype(tf.float32),
+                low=0.0,
+                high=np.inf,
+            ),
+        )
+
+        return kde_distribution
+
+    def disturb_G(self, G):
+        mask = utils.random_bool_tensor(G.shape, self.probability)
+        idxs = tf.where(mask)
+        zeroed_G = tf.where(mask, 0.0, G)
+        stuck_G = self.distribution.sample(tf.math.count_nonzero(mask))
+        disturbed_G = zeroed_G + tf.scatter_nd(idxs, stuck_G, G.shape)
+        return disturbed_G
+
+
+class D2DLognormal(Nonideality, LinearityPreserving):
+    """Models D2D programming variability as lognormal deviations of resistances."""
+
+    def __init__(
+        self, G_off: float, G_on: float, R_on_log_std: float, R_off_log_std: float
+    ) -> None:
+        """
+        Args:
+            G_on: Memristor conductance in ON state.
+            G_off: Memristor conductance in OFF state.
+            R_on_log_std: Standard deviation of the (lognormal distribution's) underlying normal
+                distribution associated with R_on (i.e. 1/G_on).
+            R_off_log_std: Standard deviation of the (lognormal distribution's) underlying normal
+                distribution associated with R_off (i.e. 1/G_off).
+        """
+        self.G_off = G_off
+        self.G_on = G_on
+        self.R_on_log_std = R_on_log_std
+        self.R_off_log_std = R_off_log_std
+
+    def label(self):
+        return f"D2DLN:{self.R_on_log_std:.3g}_{self.R_off_log_std:.3g}"
+
+    def disturb_G(self, G):
+        R = 1 / G
+        R_on = 1 / self.G_on
+        R_off = 1 / self.G_off
+
+        # Piece-wise linear interpolation.
+        log_std_ref = [self.R_on_log_std, self.R_off_log_std]
+        log_std = tfp.math.interp_regular_1d_grid(R, R_on, R_off, log_std_ref)
+
+        # Lognormal modelling.
+        R_squared = tf.math.pow(R, 2)
+        log_var = tf.math.pow(log_std, 2)
+        # Because $\sigma = \ln ( 1 + \frac{\sigma_X^2}{\mu_X^2} )$,
+        # $\sigma_X^2 = \mu_X^2 (e^{\sigma^2} - 1)$.
+        R_var = R_squared * (tf.math.exp(log_var) - 1.0)
+        log_mu = tf.math.log(R_squared / tf.math.sqrt(R_squared + R_var))
+        R = tfd.LogNormal(log_mu, log_std, validate_args=True).sample()
+
+        G = 1 / R
+
+        return G
diff --git a/awarememristor/crossbar/utils.py b/awarememristor/crossbar/utils.py
new file mode 100644
index 0000000..4bef9a6
--- /dev/null
+++ b/awarememristor/crossbar/utils.py
@@ -0,0 +1,38 @@
+import tensorflow as tf
+
+
+def tf_log2(x: tf.Tensor) -> tf.Tensor:
+    """Compute logarithm of base 2 of `x`.
+
+    Args:
+        x: An array.
+    """
+    numerator = tf.math.log(x)
+    denominator = tf.math.log(tf.constant(2, dtype=numerator.dtype))
+    return numerator / denominator
+
+
+def add_I_BL(I_ind: tf.Tensor) -> tf.Tensor:
+    """Add currents along the bit lines.
+
+    Args:
+        I_ind: Currents of shape `p x m x n` produced by each of the conductances in the crossbar
+            array.
+
+    Returns:
+        Output currents of shape `p x n`.
+    """
+    I = tf.math.reduce_sum(I_ind, axis=1)
+    return I
+
+
+def random_bool_tensor(shape: list[int], prob_true: float) -> tf.Tensor:
+    """Return random boolean tensor.
+
+    Args:
+        shape: Tensor shape.
+        prob_true: Probability that a given entry is going to be True. Probability must be in the
+            [0.0, 1.0] range.
+    """
+    random_float_tensor = tf.random.uniform(shape, minval=0, maxval=1, dtype=tf.dtypes.float64)
+    return random_float_tensor < prob_true
diff --git a/awarememristor/plotting/__init__.py b/awarememristor/plotting/__init__.py
new file mode 100644
index 0000000..5723f87
--- /dev/null
+++ b/awarememristor/plotting/__init__.py
@@ -0,0 +1 @@
+from awarememristor.plotting import figures, supporting_figures
diff --git a/awarememristor/plotting/figures.py b/awarememristor/plotting/figures.py
new file mode 100644
index 0000000..1452e4e
--- /dev/null
+++ b/awarememristor/plotting/figures.py
@@ -0,0 +1,459 @@
+import matplotlib.pyplot as plt
+import numpy as np
+import pandas as pd
+from matplotlib import rc
+from matplotlib.lines import Line2D
+
+from awarememristor import crossbar, simulations
+from awarememristor.crossbar.nonidealities import StuckDistribution
+from awarememristor.plotting import utils
+from awarememristor.training import architecture
+
+
+def _SiO_x_panels(fig, axes):
+    data = simulations.data.load_SiO_x_multistate()
+
+    N = 1000
+    palette = plt.cm.inferno(np.linspace(0, 1, N))
+    min_voltage, max_voltage = 0.0, 0.5
+
+    curves = simulations.data.low_high_n_SiO_x_curves(data)
+    for axis, (voltages, currents) in zip(axes, curves):
+        for idx in range(voltages.shape[0]):
+            voltage_curve = voltages[idx, :]
+            current_curve = currents[idx, :]
+            n = simulations.data.nonlinearity_parameter(current_curve)
+            palette_idx = int(np.floor(N * (n - 2) / 2))
+            axis.plot(
+                voltage_curve,
+                current_curve,
+                color=palette[palette_idx],
+                linewidth=utils.Config.LINEWIDTH,
+            )
+
+        axis.set_xlim([min_voltage, max_voltage])
+        axis.set_ylim(bottom=0)
+        axis.set_xlabel(utils.axis_label("voltage"))
+        axis.ticklabel_format(axis="y", scilimits=(-1, 1))
+        axis.yaxis.get_offset_text().set_fontsize(utils.Config.TICKS_FONT_SIZE)
+
+    sm = plt.cm.ScalarMappable(cmap="inferno", norm=plt.Normalize(vmin=2, vmax=4))
+    cbar = fig.colorbar(sm, ax=axes)
+    cbar.set_label(
+        label=utils.axis_label("nonlinearity-parameter"),
+        fontsize=utils.Config.AXIS_LABEL_FONT_SIZE,
+        rotation=-90,
+        va="bottom",
+    )
+    cbar.ax.tick_params(axis="both", which="both", labelsize=utils.Config.TICKS_FONT_SIZE)
+
+    axes[0].set_ylabel(utils.axis_label("current"))
+
+
+def _HfO2_panels(fig, axes):
+    data = simulations.data.load_Ta_HfO2()
+    G_min, G_max = simulations.data.extract_G_off_and_G_on(data)
+    vals, p = simulations.data.extract_stuck(data, G_min, G_max)
+    median_range = G_max - G_min
+    colors = utils.color_dict()
+
+    axis = axes[0]
+    shape = data.shape
+    num_pulses = shape[0] * shape[1]
+    num_bl = shape[2]
+    num_wl = shape[3]
+    pulsing_step_size = 10
+    random_proportion = 0.01
+    x = [i + 1 for i in range(0, num_pulses, pulsing_step_size)]
+    data = np.reshape(data, (num_pulses, num_bl, num_wl))
+    num_devices = num_wl * num_bl
+    num_reduced_devices = int(np.around(random_proportion * num_devices))
+    np.random.seed(0)
+    random_idxs = np.random.choice(num_devices, num_reduced_devices)
+    bl_idxs, wl_idxs = np.unravel_index(random_idxs, (num_bl, num_wl))
+    for bl_idx, wl_idx in zip(bl_idxs, wl_idxs):
+        curve_data = data[:, bl_idx, wl_idx]
+        if np.max(curve_data) - np.min(curve_data) < simulations.data.stuck_device_threshold(
+            median_range
+        ):
+            color = colors["vermilion"]
+        else:
+            color = colors["bluish-green"]
+        y = curve_data[::pulsing_step_size]
+        axis.plot(x, 1000 * y, color=color, lw=utils.Config.LINEWIDTH / 2, alpha=1 / 2)
+
+    for G in [G_min, G_max]:
+        axis.axhline(
+            1000 * G,
+            0,
+            1,
+            color=colors["blue"],
+            lw=utils.Config.LINEWIDTH,
+            linestyle="dashed",
+            zorder=10,
+        )
+
+    axis.set_xlabel(utils.axis_label("pulse-number"))
+    axis.set_ylabel(utils.axis_label("conductance", unit_prefix="m"))
+
+    axis.set_xlim([0, x[-1]])
+    axis.set_ylim(bottom=0.0)
+
+    handles = [
+        Line2D([0], [0], color=colors["vermilion"], label="Stuck devices"),
+        Line2D([0], [0], color=colors["bluish-green"], label="Other devices"),
+        Line2D(
+            [0],
+            [0],
+            color=colors["blue"],
+            linestyle="dashed",
+            label=r"$G_\mathrm{off}, G_\mathrm{on}$",
+        ),
+    ]
+
+    # Distribution
+    axis = axes[1]
+    distribution = StuckDistribution(vals, p).distribution
+    x = np.linspace(0.0, 1.5e-3, int(1e4))
+    y = distribution.prob(x)
+    y = y / 1000
+    x = 1000 * x
+
+    axis.plot(y, x, lw=0.5, color=colors["vermilion"])
+    axis.scatter(
+        np.zeros_like(vals),
+        [1000 * val for val in vals],
+        marker="_",
+        alpha=0.1,
+        lw=utils.Config.LINEWIDTH / 2,
+        color=colors["vermilion"],
+    )
+    for G in [G_min, G_max]:
+        axis.axhline(
+            1000 * G,
+            0,
+            1,
+            color=colors["blue"],
+            lw=utils.Config.LINEWIDTH,
+            linestyle="dashed",
+            zorder=10,
+        )
+
+    axis.set_xlabel(r"Probability density ($\mathrm{mS}^{-1}$)")
+
+    axis.set_xlim(left=0.0)
+
+    utils.add_legend(
+        fig,
+        ncol=3,
+        bbox_to_anchor=(0.5, 0.515),
+        handles=handles,
+    )
+
+
+def experimental_data():
+    fig = plt.figure(constrained_layout=True)
+    gs = fig.add_gridspec(3, 1, height_ratios=[1.0, 0.08, 1.1])
+
+    gs_top = gs[0].subgridspec(1, 2, wspace=0.03)
+    gs_bottom = gs[2].subgridspec(1, 2, wspace=0.03)
+
+    subplots = list(gs_top) + list(gs_bottom)
+    for subplot in subplots:
+        fig.add_subplot(subplot)
+
+    fig, axes = utils.fig_init(2, 0.9, custom_fig=fig)
+
+    axes[1].sharex(axes[0])
+    axes[3].sharey(axes[2])
+    axes[3].label_outer()
+
+    _SiO_x_panels(fig, axes[[0, 1]])
+    _HfO2_panels(fig, axes[[2, 3]])
+
+    utils.save_fig(fig, "experimental-data")
+
+
+def iv_nonlinearity_training(metric="error"):
+    fig, axes = utils.fig_init(2, 0.55, fig_shape=(2, 3), sharex=True, sharey=True)
+
+    iterators = simulations.iv_nonlinearity.get_iterators()
+    # Same training, different inference.
+    iterators.insert(3, iterators[0])
+    inference_idxs = [0, 0, 0, 1, 0, 0]
+
+    for idx, (iterator, inference_idx) in enumerate(zip(iterators, inference_idxs)):
+        i, j = np.unravel_index(idx, axes.shape)
+        axis = axes[i, j]
+        utils.plot_training_curves(axis, iterator, metric=metric, inference_idx=inference_idx)
+        if i + 1 == axes.shape[0]:
+            axis.set_xlabel(utils.axis_label("epoch"))
+        if j == 0:
+            axis.set_ylabel(utils.axis_label(metric))
+
+    utils.add_legend(
+        fig,
+        labels=["Training", "Validation", "Test (nonideal)"],
+        ncol=axes.shape[1],
+        bbox_to_anchor=(0.5, 1.03),
+    )
+
+    utils.save_fig(fig, "iv-nonlinearity-training", metric=metric)
+
+
+def iv_nonlinearity_inference(metric="error"):
+    fig, axes = utils.fig_init(1, 0.8)
+
+    iterators = simulations.iv_nonlinearity.get_iterators()
+    # Same training, different inference.
+    iterators.insert(3, iterators[0])
+    inference_idxs = [0, 0, 0, 1, 0, 0]
+
+    colors = [utils.color_dict()[key] for key in ["vermilion", "blue", "bluish-green"]]
+
+    boxplots = []
+
+    for idx, (iterator, inference_idx) in enumerate(zip(iterators, inference_idxs)):
+        avg_power = iterator.test_metric("avg_power", inference_idx=inference_idx)
+        y = iterator.test_metric(metric, inference_idx=inference_idx)
+        color = colors[idx % 3]
+        boxplot = utils.plot_boxplot(
+            axes, y, color, metric=metric, x=avg_power, is_x_log=True, linear_width=0.2
+        )
+        boxplots.append(boxplot)
+
+    utils.add_boxplot_legend(
+        axes, boxplots, ["Standard", "Nonideality-aware", "Nonideality-aware (regularized)"]
+    )
+
+    plt.xlabel(utils.axis_label("power-consumption"))
+    plt.ylabel(utils.axis_label(metric))
+
+    utils.save_fig(fig, "iv-nonlinearity-inference", metric=metric)
+
+
+def iv_nonlinearity_cnn(metric="error"):
+    fig, axes = utils.fig_init(2, 1 / 3, fig_shape=(1, 3), sharey=True)
+
+    colors = utils.color_dict()
+
+    iterators = simulations.iv_nonlinearity_cnn.get_iterators()
+
+    axes[0].set_ylabel(utils.axis_label(metric))
+
+    # Error curves.
+    for axis, iterator in zip(axes, iterators):
+        utils.plot_training_curves(axis, iterator, metric=metric)
+        axis.set_xlabel(utils.axis_label("epoch"))
+
+    # Box plots.
+    axis = axes[2]
+    for idx, (iterator, color) in enumerate(zip(iterators, [colors["vermilion"], colors["blue"]])):
+        y = iterator.test_metric(metric)
+        _ = utils.plot_boxplot(axis, y, color, x=idx, metric=metric, linewidth_scaling=2 / 3)
+    axis.set_xticks([0, 1])
+    axis.set_xticklabels(["Standard", "Nonideality-aware"])
+
+    axis.set_xlabel("Training")
+    axis.set_ylim(top=95.0)
+
+    utils.add_legend(
+        fig,
+        labels=["Training", "Validation", "Test (nonideal)"],
+        ncol=len(axes),
+        bbox_to_anchor=(0.35, 1.05),
+    )
+
+    utils.save_fig(fig, "iv-nonlinearity-cnn", metric=metric)
+
+
+def weight_implementation(metric="error"):
+    fig = plt.figure(constrained_layout=True)
+    gs = fig.add_gridspec(2, 1, height_ratios=[1.0, 0.8])
+
+    gs_top = gs[0].subgridspec(2, 4)
+    gs_bottom = gs[1].subgridspec(1, 2)
+
+    subplots = list(gs_top) + list(gs_bottom)
+    for subplot in subplots:
+        fig.add_subplot(subplot)
+
+    fig, axes = utils.fig_init(2, 1.0, custom_fig=fig)
+
+    for axis in axes[:8]:
+        axis.sharex(axes[4])
+        axis.sharey(axes[0])
+        axis.label_outer()
+        axis.set_aspect("equal", adjustable="box")
+
+    iterators = simulations.weight_implementation.get_iterators()[1:]
+    colors = [
+        utils.color_dict()[key] for key in ["reddish-purple", "vermilion", "blue", "bluish-green"]
+    ]
+
+    temp_iterators = [iterators[idx] for idx in [0, 1, 4, 5, 2, 3, 6, 7]]
+    for idx, (axis, iterator, color) in enumerate(zip(axes, temp_iterators, colors + colors)):
+        iterator.is_training = False
+        iterator.inference_idx = 0
+        model = architecture.get_model(iterator, custom_weights_path=iterator.weights_path())
+        weights = model.layers[1].combined_weights()
+
+        inference = iterator.current_stage()
+        G_off = inference.G_off
+        G_on = inference.G_on
+
+        if iterator.training.uses_double_weights():
+            G, _ = crossbar.map.double_w_to_G(weights, G_off, G_on)
+        else:
+            G, _ = crossbar.map.w_to_G(weights, G_off, G_on, mapping_rule=inference.mapping_rule)
+
+        G = 1e6 * G
+        utils.plot_scatter(axis, G[:, ::2], G[:, 1::2], color, random_proportion=0.1)
+
+        axis.xaxis.set_ticks(np.arange(1.0, 3.0, 0.5))
+        axis.yaxis.set_ticks(np.arange(1.0, 3.0, 0.5))
+
+        if idx > 3:
+            axis.set_xlabel(utils.axis_label("g-plus", unit_prefix="μ"))
+        if idx in [0, 4]:
+            axis.set_ylabel(utils.axis_label("g-minus", unit_prefix="μ"))
+
+    for iterator_idxs, axis in zip([[0, 1, 4, 5], [2, 3, 6, 7]], axes[-2:]):
+        for iterator_idx, color in zip(iterator_idxs, colors):
+            iterator = iterators[iterator_idx]
+            avg_power = iterator.test_metric("avg_power")
+            y = iterator.test_metric(metric)
+            utils.plot_boxplot(axis, y, color, x=1000 * avg_power, metric=metric, linear_width=0.2)
+            axis.set_xlabel(utils.axis_label("power-consumption", unit_prefix="m"))
+
+    axes[-2].set_ylabel(utils.axis_label(metric))
+    axes[-1].sharey(axes[-2])
+    axes[-1].sharex(axes[-2])
+    axes[-1].label_outer()
+
+    utils.save_fig(fig, "weight-implementation", metric=metric)
+
+
+def memristive_validation(metric="error"):
+    fig, axes = utils.fig_init(2, 1 / 3, fig_shape=(1, 3), sharey=True)
+
+    iterator = simulations.memristive_validation.get_nonideal_iterators()[0]
+
+    axes[0].set_ylabel(utils.axis_label(metric))
+
+    # Curves
+    for idx, (standard_mode, axis) in enumerate(zip([True, False], axes)):
+        iterator.training.is_standard_validation_mode = standard_mode
+        utils.plot_training_curves(axis, iterator, metric=metric)
+        axis.set_xlabel(utils.axis_label("epoch"))
+
+    # Box plots
+    axis = axes[-1]
+    colors = [utils.color_dict()[key] for key in ["vermilion", "blue"]]
+
+    for idx, (standard_mode, color) in enumerate(zip([True, False], colors)):
+        iterator.training.is_standard_validation_mode = standard_mode
+        y = iterator.test_metric(metric)
+        _ = utils.plot_boxplot(axis, y, color, x=idx, metric=metric, linewidth_scaling=2 / 3)
+
+    axis.set_xticks([0, 1])
+    axis.set_xticklabels(["Standard", "Memristive"])
+    axis.set_xlabel(utils.axis_label("checkpoint"))
+
+    utils.add_legend(
+        fig,
+        labels=["Training", "Validation", "Test (nonideal)"],
+        ncol=len(axes),
+        bbox_to_anchor=(0.35, 1.05),
+    )
+
+    utils.save_fig(fig, "memristive-validation", metric=metric)
+
+
+def nonideality_agnosticism(metric: str = "error", norm_rows=True, include_val_label=False):
+    training_labels = {
+        "nonreg__64__none_none__ideal": "Ideal",
+        "nonreg__64__0.000997_0.00351__IVNL:2.13_0.0953": r"Low $I$-$V$ nonlin. [$\mathrm{SiO}_x$]",
+        "reg__64__0.000997_0.00351__IVNL:2.13_0.0953": r"Low $I$-$V$ nonlin. [$\mathrm{SiO}_x$] (reg.)",
+        "nonreg__64__7.72e-07_2.73e-06__IVNL:2.99_0.369": r"High $I$-$V$ nonlin. [$\mathrm{SiO}_x$]",
+        "reg__64__7.72e-07_2.73e-06__IVNL:2.99_0.369": r"High $I$-$V$ nonlin. [$\mathrm{SiO}_x$] (reg.)",
+        "nonreg__64__7.72e-07_2.73e-06__StuckOff:0.05": r"Stuck at $G_\mathrm{off}$",
+        "nonreg__64__4.36e-05_0.000978__StuckDistr:0.101_1.77e-05": r"Stuck [$\mathrm{Ta/HfO}_2$]",
+        "nonreg__64__7.72e-07_2.73e-06__D2DLN:0.25_0.25": "More uniform D2D var.",
+        "reg__64__7.72e-07_2.73e-06__D2DLN:0.25_0.25": "More uniform D2D var. (reg.)",
+        "nonreg__64__7.72e-07_2.73e-06__D2DLN:0.05_0.5": "Less uniform D2D var.",
+        "reg__64__7.72e-07_2.73e-06__D2DLN:0.05_0.5": "Less uniform D2D var. (reg.)",
+        "nonreg__64__7.72e-07_2.73e-06__IVNL:2.99_0.369+StuckOn:0.05": r"High $I$-$V$ nonlin. [$\mathrm{SiO}_x$] + stuck at $G_\mathrm{on}$",
+        "nonreg__64__7.72e-07_2.73e-06__D2DLN:0.5_0.5": "High D2D var.",
+    }
+    inference_labels = {
+        "none_none__ideal": training_labels["nonreg__64__none_none__ideal"],
+        "0.000997_0.00351__IVNL:2.13_0.0953": training_labels[
+            "nonreg__64__0.000997_0.00351__IVNL:2.13_0.0953"
+        ],
+        "7.72e-07_2.73e-06__IVNL:2.99_0.369": training_labels[
+            "nonreg__64__7.72e-07_2.73e-06__IVNL:2.99_0.369"
+        ],
+        "7.72e-07_2.73e-06__StuckOff:0.05": training_labels[
+            "nonreg__64__7.72e-07_2.73e-06__StuckOff:0.05"
+        ],
+        "4.36e-05_0.000978__StuckDistr:0.101_1.77e-05": training_labels[
+            "nonreg__64__4.36e-05_0.000978__StuckDistr:0.101_1.77e-05"
+        ],
+        "7.72e-07_2.73e-06__D2DLN:0.25_0.25": training_labels[
+            "nonreg__64__7.72e-07_2.73e-06__D2DLN:0.25_0.25"
+        ],
+        "7.72e-07_2.73e-06__D2DLN:0.05_0.5": training_labels[
+            "nonreg__64__7.72e-07_2.73e-06__D2DLN:0.05_0.5"
+        ],
+        "7.72e-07_2.73e-06__IVNL:2.99_0.369+StuckOn:0.05": training_labels[
+            "nonreg__64__7.72e-07_2.73e-06__IVNL:2.99_0.369+StuckOn:0.05"
+        ],
+        "7.72e-07_2.73e-06__D2DLN:0.5_0.5": training_labels[
+            "nonreg__64__7.72e-07_2.73e-06__D2DLN:0.5_0.5"
+        ],
+    }
+    df = pd.DataFrame(
+        columns=[training_labels[key] for key in training_labels],
+        index=[inference_labels[key] for key in inference_labels],
+    )
+    df = df.astype(float)
+    iterators = simulations.nonideality_agnosticism.get_iterators()
+    for iterator in iterators:
+        training_label = training_labels[iterator.training.label()]
+        ys = [
+            iterator.test_metric(metric, inference_idx=idx)
+            for idx in range(len(iterator.inferences))
+        ]
+        for inference, y in zip(iterator.inferences, ys):
+            inference_label = inference_labels[inference.label()]
+            df.at[inference_label, training_label] = np.median(y)
+
+    fig, axes = utils.fig_init(2, 0.5)
+
+    filename = "nonideality-agnosticism"
+    if not norm_rows:
+        filename += "-not-norm"
+    if include_val_label:
+        filename += "-with-val-label"
+
+    utils.add_heatmap(
+        fig, axes, df, x_ticks=df.columns, y_ticks=df.index, metric=metric, norm_rows=norm_rows
+    )
+
+    axes.set_ylabel(utils.axis_label("inference"))
+    axes.set_xlabel(utils.axis_label("training"))
+
+    if include_val_label:
+        axes.text(
+            1.05,
+            0.5,
+            utils.axis_label("error", prepend="median"),
+            horizontalalignment="center",
+            verticalalignment="center",
+            rotation=-90,
+            transform=axes.transAxes,
+            fontsize=utils.Config.AXIS_LABEL_FONT_SIZE,
+        )
+
+    utils.save_fig(fig, filename, metric=metric)
diff --git a/awarememristor/plotting/supporting_figures.py b/awarememristor/plotting/supporting_figures.py
new file mode 100644
index 0000000..d49889d
--- /dev/null
+++ b/awarememristor/plotting/supporting_figures.py
@@ -0,0 +1,236 @@
+import matplotlib.pyplot as plt
+import numpy as np
+
+from awarememristor import simulations
+from awarememristor.plotting import utils
+
+
+def all_iv_curves_full_range():
+    fig, axes = utils.fig_init(2, 0.6, fig_shape=(1, 1))
+
+    data = simulations.data.load_SiO_x_multistate()
+    voltages, currents = simulations.data.all_SiO_x_curves(data, max_voltage=5.0)
+
+    colors = utils.color_list()[:-1]
+    num_colors = len(colors)
+    for idx in range(voltages.shape[0]):
+        voltage_curve = voltages[idx, :]
+        current_curve = currents[idx, :]
+        color_idx = idx % num_colors
+        axes.plot(
+            voltage_curve,
+            current_curve,
+            linewidth=utils.Config.LINEWIDTH,
+            color=colors[color_idx],
+        )
+
+    axes.set_xlim(left=0.0)
+    axes.set_xlabel(utils.axis_label("voltage"))
+    axes.yaxis.get_offset_text().set_fontsize(utils.Config.TICKS_FONT_SIZE)
+
+    axes.set_ylabel(utils.axis_label("current"))
+    axes.set_yscale("log")
+
+    utils.save_fig(fig, "all-SiO_x-IV-curves-full-range", is_supporting=True)
+
+
+def switching():
+    fig, axes = utils.fig_init(2, 0.5, fig_shape=(1, 1))
+
+    data = simulations.data.load_SiO_x_switching()
+    data[:, 0, :] = np.abs(data[:, 0, :])
+
+    colors = [utils.color_dict()[color_name] for color_name in ["blue", "orange"]]
+    labels = ["SET", "RESET"]
+    labels_x = [0.15, 1 - 0.15]
+    for idx, (color, label, label_x) in enumerate(zip(colors, labels, labels_x)):
+        voltage_curve = data[:, 1, idx]
+        current_curve = np.abs(data[:, 0, idx])
+        line = axes.plot(
+            voltage_curve,
+            current_curve,
+            linewidth=utils.Config.LINEWIDTH,
+            color=color,
+        )
+        utils.add_arrow(line[0], 60)
+        utils.add_arrow(line[0], -60)
+        utils.add_text(
+            axes,
+            label,
+            (label_x, 0.88),
+            fontsize=utils.Config.TEXT_LABEL_SIZE,
+            color=color,
+        )
+
+    axes.set_xlabel(utils.axis_label("voltage"))
+
+    axes.set_ylabel(utils.axis_label("current", prepend="absolute"))
+    axes.set_ylim(bottom=5e-8, top=5e-3)
+    axes.set_yscale("log")
+
+    utils.save_fig(fig, "SiO_x-switching", is_supporting=True)
+
+
+def _training_curves_multiple_panels(
+    width_num_cols,
+    height_frac,
+    fig_shape,
+    iterators,
+    metric,
+    figure_name,
+    inference_idxs=None,
+    y_lim=None,
+):
+    fig, axes = utils.fig_init(
+        width_num_cols, height_frac, fig_shape=fig_shape, sharex=True, sharey=True
+    )
+    if inference_idxs is None:
+        inference_idxs = [0 for _ in range(len(iterators))]
+
+    for training_idx, linestyle in enumerate(utils.get_linestyles()):
+        for i in range(len(iterators)):
+            iterators[i].training.repeat_idx = training_idx
+
+        for idx, (iterator, inference_idx) in enumerate(zip(iterators, inference_idxs)):
+            if len(axes.shape) == 1:
+                axis = axes[idx]
+            else:
+                i, j = np.unravel_index(idx, axes.shape)
+                axis = axes[i, j]
+            utils.plot_training_curves(
+                axis,
+                iterator,
+                metric=metric,
+                inference_idx=inference_idx,
+                linestyle=linestyle,
+                is_many=True,
+            )
+            if len(axes.shape) == 1:
+                axis.set_xlabel(utils.axis_label("epoch"))
+                if idx == 0:
+                    axis.set_ylabel(utils.axis_label(metric))
+            else:
+                if i + 1 == axes.shape[0]:
+                    axis.set_xlabel(utils.axis_label("epoch"))
+                if j == 0:
+                    axis.set_ylabel(utils.axis_label(metric))
+            if y_lim is not None:
+                axis.set_ylim(top=y_lim)
+
+    utils.add_legend(
+        fig,
+        labels=["Training", "Validation", "Test (nonideal)"],
+        ncol=3,
+        bbox_to_anchor=(0.5, 1.03),
+    )
+
+    utils.save_fig(fig, f"{figure_name}-training", is_supporting=True, metric=metric)
+
+
+def iv_nonlinearity_training(metric="error"):
+    iterators = simulations.iv_nonlinearity.get_iterators()
+    # Same training, different inference.
+    iterators.insert(3, iterators[0])
+    inference_idxs = [0, 0, 0, 1, 0, 0]
+
+    _training_curves_multiple_panels(
+        2,
+        0.55,
+        (2, 3),
+        iterators,
+        metric,
+        "iv-nonlinearity",
+        inference_idxs=inference_idxs,
+    )
+
+
+def weight_implementation_standard_weights_training(metric="error"):
+    iterators = simulations.weight_implementation.get_nonideal_iterators()[:4]
+
+    _training_curves_multiple_panels(
+        (2, 3),
+        0.77,
+        (2, 2),
+        iterators,
+        metric,
+        "weight-implementation-standard-weights",
+    )
+
+
+def weight_implementation_double_weights_training(metric="error"):
+    iterators = [
+        simulations.weight_implementation.get_ideal_iterator()
+    ] + simulations.weight_implementation.get_nonideal_iterators()[4:]
+    # Same training, different inference.
+    iterators.insert(3, iterators[0])
+    inference_idxs = [0, 0, 0, 1, 0, 0]
+
+    _training_curves_multiple_panels(
+        2,
+        0.55,
+        (2, 3),
+        iterators,
+        metric,
+        "weight-implementation-double-weights",
+        inference_idxs=inference_idxs,
+        y_lim=95,
+    )
+
+
+def memristive_validation_training(metric="error"):
+    iterators = [
+        *simulations.memristive_validation.get_iterators(),
+        simulations.memristive_validation.get_iterators()[1],
+    ]
+    iterators[1].training.is_standard_validation_mode = True
+
+    # Only first five will be plotted.
+    _training_curves_multiple_panels(
+        2,
+        0.32,
+        (1, 3),
+        iterators,
+        metric,
+        "memristive-validation",
+        y_lim=95,
+    )
+
+
+def stuck_off_training(metric="error"):
+    iterators = simulations.stuck_off.get_iterators()
+
+    _training_curves_multiple_panels(
+        (2, 3),
+        0.45,
+        (1, 2),
+        iterators,
+        metric,
+        "stuck-off",
+    )
+
+
+def high_iv_nonlinearity_and_stuck_on_training(metric="error"):
+    iterators = simulations.iv_nonlinearity_and_stuck_on.get_iterators()
+
+    _training_curves_multiple_panels(
+        (2, 3),
+        0.45,
+        (1, 2),
+        iterators,
+        metric,
+        "high-iv-nonlinearity-and-stuck-on",
+    )
+
+
+def stuck_distribution_training(metric="error"):
+    iterators = simulations.stuck_distribution.get_iterators()
+
+    _training_curves_multiple_panels(
+        (2, 3),
+        0.45,
+        (1, 2),
+        iterators,
+        metric,
+        "stuck-distribution",
+        y_lim=95,
+    )
diff --git a/awarememristor/plotting/utils.py b/awarememristor/plotting/utils.py
new file mode 100644
index 0000000..5c17f95
--- /dev/null
+++ b/awarememristor/plotting/utils.py
@@ -0,0 +1,550 @@
+import copy
+import os
+from typing import Union
+
+import matplotlib
+import matplotlib.pyplot as plt
+import matplotlib.transforms as mtransforms
+import numpy as np
+from numpy import ma
+
+
+def _cm_to_in(length: float) -> float:
+    return length / 2.54
+
+
+class Config:
+    AXIS_LABEL_FONT_SIZE: float = 12
+    LEGEND_FONT_SIZE: float = 8
+    TICKS_FONT_SIZE: float = 8
+    SUBPLOT_LABEL_SIZE: float = 12
+    TEXT_LABEL_SIZE: float = 10
+    LINEWIDTH: float = 0.75
+    MARKER_SIZE: float = 0.5
+    BOXPLOT_LINEWIDTH: float = 0.75
+    # Advanced Science
+    ONE_COLUMN_WIDTH: float = _cm_to_in(8.5)
+    TWO_COLUMNS_WIDTH: float = _cm_to_in(17.8)
+    TWO_THIRDS_COLUMN_WIDTH: float = 2 / 3 * TWO_COLUMNS_WIDTH
+    COL_WIDTHS: dict[Union[int, tuple[int, int]], float] = {
+        1: ONE_COLUMN_WIDTH,
+        2: TWO_COLUMNS_WIDTH,
+        (2, 3): TWO_THIRDS_COLUMN_WIDTH,
+    }
+
+
+def color_list() -> list[str]:
+    """Return colors of Okabe-Ito colorblind-friendly palette.
+
+    Returns:
+        HEX color codes.
+    """
+    colors = [
+        "#E69F00",
+        "#56B4E9",
+        "#009E73",
+        "#F0E442",
+        "#0072B2",
+        "#D55E00",
+        "#CC79A7",
+        "#000000",
+    ]
+    return colors
+
+
+def color_dict() -> dict[str, str]:
+    """Return same as `colors_list()` but dict."""
+    color_names = [
+        "orange",
+        "sky-blue",
+        "bluish-green",
+        "yellow",
+        "blue",
+        "vermilion",
+        "reddish-purple",
+        "black",
+    ]
+    colors = dict(zip(color_names, color_list()))
+    return colors
+
+
+def get_linestyles():
+    return [
+        "solid",
+        "dotted",
+        "dashed",
+        "dashdot",
+        (0, (1, 10)),  # "loosely dotted"
+    ]
+
+
+def fig_init(
+    width_num_cols: Union[int, tuple[int, int]],
+    height_frac: float,
+    fig_shape: tuple[int, int] = (1, 1),
+    sharex=False,
+    sharey=False,
+    scaled_position: tuple[float, float] = (-0.06, 1.04),
+    custom_fig: matplotlib.figure = None,
+) -> tuple[matplotlib.figure, matplotlib.axes]:
+    width = Config.COL_WIDTHS[width_num_cols]
+    height = height_frac * width
+    if custom_fig is not None:
+        fig = custom_fig
+        fig.set_size_inches(width, height)
+        axes = fig.axes
+        axes = np.array(axes)
+        fig_shape = axes.shape
+    else:
+        fig, axes = plt.subplots(
+            *fig_shape,
+            sharex=sharex,
+            sharey=sharey,
+            figsize=(width, height),
+        )
+        fig.tight_layout()
+
+    if fig_shape == (1, 1):
+        temp_axes = np.array([axes])
+    else:
+        temp_axes = axes
+
+    for idx, axis in enumerate(temp_axes.flatten()):
+        axis.xaxis.label.set_size(Config.AXIS_LABEL_FONT_SIZE)
+        axis.yaxis.label.set_size(Config.AXIS_LABEL_FONT_SIZE)
+        axis.tick_params(axis="both", which="both", labelsize=Config.TICKS_FONT_SIZE)
+        if fig_shape != (1, 1):
+            add_subfigure_label(axis, idx, scaled_position, Config.SUBPLOT_LABEL_SIZE)
+
+    return fig, axes
+
+
+def add_subfigure_label(
+    axis,
+    letter_idx: int,
+    normalised_position: tuple[float, float],
+    fontsize: float = Config.SUBPLOT_LABEL_SIZE,
+    is_lowercase: bool = True,
+):
+    ascii_idx = 65 + letter_idx
+    if is_lowercase:
+        ascii_idx += 32
+    add_text(axis, chr(ascii_idx), normalised_position, fontsize=fontsize, fontweight="bold")
+
+
+def add_text(
+    axis,
+    text: str,
+    normalised_position: tuple[float, float],
+    fontsize: float = Config.TEXT_LABEL_SIZE,
+    fontweight: str = "normal",
+    color: str = None,
+):
+    axis.text(
+        *normalised_position,
+        text,
+        horizontalalignment="center",
+        verticalalignment="center",
+        transform=axis.transAxes,
+        fontweight=fontweight,
+        fontsize=fontsize,
+        color=color,
+    )
+
+
+def plot_training_curves(
+    axis,
+    iterator,
+    metric="error",
+    inference_idx=0,
+    linestyle="solid",
+    is_many=False,
+):
+    colors = color_dict()
+
+    # Training curve.
+    x_training, y_training = iterator.training_curves(metric)
+    plot_curve(
+        axis,
+        x_training,
+        y_training,
+        colors["orange"],
+        metric=metric,
+        linestyle=linestyle,
+        is_many=is_many,
+    )
+
+    # Validation curve.
+    x_validation, y_validation = iterator.validation_curves(metric)
+    plot_curve(
+        axis,
+        x_validation,
+        y_validation,
+        colors["sky-blue"],
+        metric=metric,
+        linestyle=linestyle,
+        is_many=is_many,
+    )
+
+    # Testing (during training) curve.
+    x_training_testing, y_training_testing = iterator.training_testing_curves(
+        metric, iterator.inferences[inference_idx]
+    )
+    plot_curve(
+        axis,
+        x_training_testing,
+        y_training_testing,
+        colors["reddish-purple"],
+        metric=metric,
+        linestyle=linestyle,
+        is_many=is_many,
+    )
+
+    axis.set_yscale("log")
+    axis.set_xlim([0, len(x_training)])
+
+
+def plot_curve(axis, x, y, color, metric="error", linestyle="solid", is_many=False):
+    if metric in ["accuracy", "error"]:
+        y = 100 * y
+    lw = Config.LINEWIDTH
+    if is_many:
+        lw /= 2
+    if len(y.shape) > 1:
+        alpha = 0.25
+        if is_many:
+            alpha /= 2
+        y_min = np.min(y, axis=1)
+        y_max = np.max(y, axis=1)
+        y_median = np.median(y, axis=1)
+        axis.fill_between(x, y_min, y_max, color=color, alpha=alpha, linewidth=0)
+        axis.plot(x, y_median, color=color, linewidth=lw / 2, linestyle=linestyle)
+    else:
+        if is_many:
+            x = np.concatenate((x[::20], [x[-1]]))
+            y = np.concatenate((y[::20], [y[-1]]))
+            lw /= 2  # Make all curves the same linewidth.
+        axis.plot(x, y, color=color, linewidth=lw, linestyle=linestyle)
+
+
+def _numpify(x):
+    try:  # In case `tf.Tensor`
+        x = x.numpy()
+    except AttributeError:
+        pass
+
+    return x
+
+
+def plot_scatter(axis, x, y, color, alpha=1.0, random_proportion=None):
+    x = _numpify(x)
+    y = _numpify(y)
+    x = x.flatten()
+    y = y.flatten()
+    if random_proportion:
+        np.random.seed(0)
+        num_points = x.size
+        num_reduced_points = int(np.around(random_proportion * num_points))
+        random_idxs = np.random.choice(num_points, num_reduced_points)
+        x = x[random_idxs]
+        y = y[random_idxs]
+    axis.scatter(
+        x,
+        y,
+        color=color,
+        marker="x",
+        s=Config.MARKER_SIZE,
+        linewidth=Config.MARKER_SIZE,
+        alpha=alpha,
+    )
+
+
+def plot_boxplot(
+    axis,
+    y,
+    color,
+    x=None,
+    metric="error",
+    is_x_log=False,
+    linewidth_scaling=1.0,
+    linear_width: float = 0.25,
+):
+    y = y.flatten()
+    if metric in ["accuracy", "error"]:
+        y = 100 * y
+
+    positions = None
+    if x is not None:
+        try:
+            x = x.flatten()
+        except AttributeError:
+            pass
+        positions = [np.mean(x)]
+    if is_x_log and positions is not None:
+        widths = [
+            10 ** (np.log10(positions[0]) + linear_width / 2.0)
+            - 10 ** (np.log10(positions[0]) - linear_width / 2.0)
+        ]
+    else:
+        widths = [linear_width]
+    boxplot = axis.boxplot(
+        y,
+        positions=positions,
+        widths=widths,
+        sym=color,
+    )
+    plt.setp(boxplot["fliers"], marker="x", markersize=4 * Config.MARKER_SIZE, markeredgewidth=0.5)
+    for element in ["boxes", "whiskers", "fliers", "means", "medians", "caps"]:
+        plt.setp(
+            boxplot[element], color=color, linewidth=linewidth_scaling * Config.BOXPLOT_LINEWIDTH
+        )
+
+    if is_x_log:
+        axis.set_xscale("log")
+    else:
+        axis.set_xscale("linear")
+    axis.set_yscale("log")
+
+    return boxplot
+
+
+def add_boxplot_legend(axis, boxplots, labels, linewdith=1.0, loc="upper right"):
+    leg = axis.legend(
+        [boxplot["boxes"][0] for boxplot in boxplots],
+        labels,
+        fontsize=Config.LEGEND_FONT_SIZE,
+        frameon=False,
+        loc=loc,
+    )
+    for line in leg.get_lines():
+        line.set_linewidth(linewdith)
+
+
+def add_legend(
+    fig,
+    labels=None,
+    ncol=1,
+    loc="center",
+    bbox_to_anchor=(0.5, 1.0),
+    linewidth=1.0,
+    frameon=False,
+    handles=None,
+):
+    if handles is None:
+        leg = fig.legend(labels, ncol=ncol, loc=loc, bbox_to_anchor=bbox_to_anchor, frameon=frameon)
+    else:
+        leg = fig.legend(
+            handles=handles, ncol=ncol, loc=loc, bbox_to_anchor=bbox_to_anchor, frameon=frameon
+        )
+    for line in leg.get_lines():
+        line.set_linewidth(linewidth)
+
+
+def save_fig(fig, name: str, is_supporting: bool = False, metric: str = "error"):
+    dir_name = "plots"
+    os.makedirs(dir_name, exist_ok=True)
+    if metric != "error":
+        name += f"-{metric}"
+    if is_supporting:
+        name = f"supporting-information--{name}"
+    path = os.path.join(dir_name, f"{name}.pdf")
+    fig.savefig(path, bbox_inches="tight", transparent=True)
+
+
+def axis_label(var_name: str, prepend: str = None, unit_prefix: str = "") -> str:
+    if var_name == "accuracy":
+        label = "accuracy (%)"
+    elif var_name == "error":
+        label = "error (%)"
+    elif var_name == "loss":
+        label = "loss"
+    elif var_name == "epoch":
+        label = "epoch"
+    elif var_name == "inference":
+        label = "inference"
+    elif var_name == "training":
+        label = "training"
+    elif var_name == "power-consumption":
+        label = f"power consumption ({unit_prefix}W)"
+    elif var_name == "d2d-uniformity":
+        label = "uniformity of D2D variability"
+    elif var_name == "checkpoint":
+        label = "checkpoint"
+    elif var_name == "conductance":
+        label = f"conductance ({unit_prefix}S)"
+    elif var_name == "voltage":
+        label = f"voltage ({unit_prefix}V)"
+    elif var_name == "current":
+        label = f"current ({unit_prefix}A)"
+    elif var_name == "nonlinearity-parameter":
+        label = "nonlinearity parameter"
+    elif var_name == "pulse-number":
+        label = "pulse number"
+    elif var_name == "g-plus":
+        label = rf"$G_{{+}}$ ({unit_prefix}S)"
+    elif var_name == "g-minus":
+        label = rf"$G_{{-}}$ ({unit_prefix}S)"
+    else:
+        raise ValueError(f'Unrecognised variable name "{var_name}".')
+
+    if prepend is not None:
+        label = f"{prepend} {label}"
+
+    first_letter = label[0].upper()
+    if len(label) > 1:
+        label = first_letter + label[1:]
+    else:
+        label = first_letter
+
+    return label
+
+
+def _get_luminance(r, g, b):
+    """Adapted from <https://stackoverflow.com/a/596243/17322548>."""
+    return 0.299 * r + 0.587 * g + 0.114 * b
+
+
+def _annotate_heatmap(
+    im: matplotlib.image,
+    data: np.array = None,
+    valfmt: Union[str, matplotlib.ticker.StrMethodFormatter] = "{x:.2f}",
+    textcolors: tuple[str, str] = ("black", "white"),
+    threshold: float = 0.5,
+    norm_rows: bool = False,
+    **textkw,
+):
+    """Annotate a heatmap. Adapted from
+    <https://matplotlib.org/stable/gallery/images_contours_and_fields/image_annotated_heatmap.html>.
+
+    Args:
+        im: The image to be labelled.
+        data: Data used to annotate. If `None`, the image's data is used.
+        valfmt: The format of the annotations inside the heatmap. This should
+            either use the string format method, e.g. `{x:.2f}`, or be a
+            `matplotlib.ticker.Formatter`.
+        textcolors: A pair of colors. The first is used for values below a
+            threshold, the second for those above.
+        threshold: Value in data units according to which the colors from
+            textcolors are applied. If `None` (the default) uses the middle of
+            the colormap as separation.
+        **kwargs: All other arguments are forwarded to each call to `text` used to create the text labels.
+    """
+    if not isinstance(data, (list, np.ndarray)):
+        data = im.get_array()
+
+    # Set default alignment to center, but allow it to be overwritten by textkw.
+    kw = dict(horizontalalignment="center", verticalalignment="center")
+    kw.update(textkw)
+
+    # Get the formatter in case a string is supplied
+    if isinstance(valfmt, str):
+        valfmt = matplotlib.ticker.StrMethodFormatter(valfmt)
+
+    # Loop over the data and create a `Text` for each "pixel".
+    # Change the text's color depending on the data.
+    texts = []
+    for i in range(data.shape[0]):
+        if norm_rows:
+            colors = im.cmap(matplotlib.colors.LogNorm()(data[i, :]))
+            luminance = _get_luminance(colors[:, 0], colors[:, 1], colors[:, 2])
+        else:
+            colors = im.cmap(matplotlib.colors.LogNorm()(data))
+            luminance = _get_luminance(colors[:, :, 0], colors[:, :, 1], colors[:, :, 2])
+        for j in range(data.shape[1]):
+            if norm_rows:
+                cell_luminance = luminance[j]
+            else:
+                cell_luminance = luminance[i, j]
+            kw.update(color=textcolors[int(cell_luminance > threshold)])
+            text = im.axes.text(j, i, valfmt(data[i, j], None), **kw)
+            texts.append(text)
+
+    return texts
+
+
+def add_heatmap(fig, axis, data, x_ticks=None, y_ticks=None, metric="error", norm_rows=False):
+    if metric in ["accuracy", "error"]:
+        data = 100 * data
+
+    data = data.to_numpy()
+    if norm_rows:
+        num_rows = data.shape[0]
+        row_indices = np.arange(num_rows)
+        for i in range(num_rows):
+            row_data = ma.array(copy.deepcopy(data))
+            row_data[row_indices != i, :] = ma.masked
+            image = axis.imshow(row_data, norm=matplotlib.colors.LogNorm(), cmap="cividis")
+    else:
+        image = axis.imshow(data, norm=matplotlib.colors.LogNorm(), cmap="cividis")
+
+    if x_ticks is not None:
+        axis.set_xticks(np.arange(len(x_ticks)))
+        axis.set_xticklabels(x_ticks)
+        axis.tick_params(top=True, bottom=False, labeltop=True, labelbottom=False)
+        plt.setp(axis.get_xticklabels(), rotation=-45, ha="right", rotation_mode="anchor")
+        axis.xaxis.set_label_position("top")
+    if y_ticks is not None:
+        axis.set_yticks(np.arange(len(y_ticks)))
+        axis.set_yticklabels(y_ticks)
+
+    if not norm_rows:
+        cbar = fig.colorbar(image, ax=axis)
+        cbar.ax.set_ylabel(
+            axis_label(metric, prepend="median"),
+            rotation=-90,
+            fontsize=Config.AXIS_LABEL_FONT_SIZE,
+            va="bottom",
+        )
+        cbar.ax.tick_params(axis="both", which="both", labelsize=Config.TICKS_FONT_SIZE)
+
+    _annotate_heatmap(
+        image,
+        data=data,
+        valfmt="{x:.1f}",
+        textcolors=("white", "black"),
+        size=Config.TICKS_FONT_SIZE,
+        norm_rows=norm_rows,
+    )
+
+
+def add_histogram(axis, values: np.ndarray, color: str, bins: int = 100, alpha: float = 1.0):
+    try:  # In case `tf.Tensor`
+        values = values.numpy()
+    except AttributeError:
+        pass
+    values = values.flatten()
+    axis.hist(values, bins=bins, color=color, alpha=alpha)
+
+
+def add_arrow(
+    line,
+    start_idx: int,
+    direction: str = "right",
+    size: float = 15,
+    color: str = None,
+    linewidth: float = None,
+):
+    """Adds arrow to a curve.
+
+    Adapted from <https://stackoverflow.com/a/34018322/17322548>.
+    """
+    if color is None:
+        color = line.get_color()
+    if linewidth is None:
+        linewidth = line.get_linewidth()
+
+    xdata = line.get_xdata()
+    ydata = line.get_ydata()
+
+    if direction == "right":
+        end_idx = start_idx + 1
+    else:
+        end_idx = start_idx - 1
+
+    line.axes.annotate(
+        "",
+        xytext=(xdata[start_idx], ydata[start_idx]),
+        xy=(xdata[end_idx], ydata[end_idx]),
+        arrowprops=dict(linewidth=linewidth, arrowstyle="->", color=color),
+        size=size,
+    )
diff --git a/awarememristor/simulations/__init__.py b/awarememristor/simulations/__init__.py
new file mode 100644
index 0000000..8031db0
--- /dev/null
+++ b/awarememristor/simulations/__init__.py
@@ -0,0 +1,7 @@
+from awarememristor.simulations import (data, ideal, iv_nonlinearity,
+                                        iv_nonlinearity_and_stuck_on,
+                                        iv_nonlinearity_cnn,
+                                        memristive_validation,
+                                        nonideality_agnosticism,
+                                        stuck_distribution, stuck_off,
+                                        weight_implementation)
diff --git a/awarememristor/simulations/data.py b/awarememristor/simulations/data.py
new file mode 100644
index 0000000..4ab8007
--- /dev/null
+++ b/awarememristor/simulations/data.py
@@ -0,0 +1,182 @@
+import os
+from typing import Optional
+
+import h5py
+import numpy as np
+import openpyxl
+import pandas as pd
+import requests
+from scipy.io import loadmat
+
+
+def load_SiO_x_multistate() -> np.ndarray:
+    """Load SiO_x data from multiple conductance states.
+
+    Returns:
+        Array of shape `(2, num_states, num_points)`. The first dimension
+            combines current and voltage values.
+    """
+    path = os.path.join(_create_and_get_data_dir(), "SiO_x-multistate-data.mat")
+    _validate_data_path(path, url="https://zenodo.org/record/5762184/files/excelDataCombined.mat")
+    data = loadmat(path)["data"]
+    data = np.flip(data, axis=2)
+    data = np.transpose(data, (1, 2, 0))
+    data = data[:2, :, :]
+    return data
+
+
+def _workbook_sheet_to_ndarray(workbook, sheet_name: str):
+    sheet = workbook[sheet_name]
+    return pd.DataFrame(sheet.values).to_numpy()[1:, :]
+
+
+def load_SiO_x_switching() -> np.ndarray:
+    """Load SiO_x switching data.
+
+    Returns:
+        Array of shape `(num_points, 2, 2)`. The second dimension combines
+            current and voltage values, while the third combined SET and RESET
+            modes.
+    """
+    path = os.path.join(_create_and_get_data_dir(), "SiO_x-switching-data.xlsx")
+    _validate_data_path(
+        path,
+        url="https://zenodo.org/record/5762184/files/Ordinary%20I-V%20data%20%28full%20cycle%29.xlsx",
+    )
+    worksheet = openpyxl.load_workbook(filename=path)
+    set_data = _workbook_sheet_to_ndarray(worksheet, "SET")[:, :2]
+    reset_data = _workbook_sheet_to_ndarray(worksheet, "RESET")[:, :2]
+    data = np.stack([set_data, reset_data], axis=-1)
+    return data
+
+
+def all_SiO_x_curves(data, max_voltage=0.5, voltage_step=0.005):
+    num_points = int(max_voltage / voltage_step) + 1
+
+    data = data[:, :, :num_points]
+    voltages = data[1, :, :]
+    currents = data[0, :, :]
+
+    return voltages, currents
+
+
+def low_high_n_SiO_x_curves(data):
+    # Arbitrary, but 11 results in a similar G_on/G_off ratio.
+    NUM_LOW_N_CURVES = 11
+
+    voltages, currents = all_SiO_x_curves(data)
+
+    num_points = voltages.shape[1]
+    half_voltage_idx = int(num_points / 2)
+    resistances = voltages[:, half_voltage_idx] / currents[:, half_voltage_idx]
+    indices = np.argsort(resistances)
+    resistances = resistances[indices]
+    voltages = voltages[indices, :]
+    currents = currents[indices, :]
+
+    low_n_ratio = resistances[NUM_LOW_N_CURVES - 1] / resistances[0]
+
+    high_n_R_off = resistances[-1]
+    idx = len(indices) - 2
+    while True:
+        # Stop whenever we exceed G_on/G_off ratio of low-nonlinearity region.
+        if high_n_R_off / resistances[idx] > low_n_ratio:
+            break
+        idx -= 1
+
+    low_n_voltages = voltages[:NUM_LOW_N_CURVES, :]
+    low_n_currents = currents[:NUM_LOW_N_CURVES, :]
+    high_n_voltages = voltages[idx:, :]
+    high_n_currents = currents[idx:, :]
+
+    return (low_n_voltages, low_n_currents), (high_n_voltages, high_n_currents)
+
+
+def nonlinearity_parameter(current_curve):
+    num_points = len(current_curve)
+    half_voltage_idx = int(num_points / 2)
+    return current_curve[-1] / current_curve[half_voltage_idx]
+
+
+def G_at_half_voltage(voltage_curve, current_curve):
+    num_points = len(current_curve)
+    half_voltage_idx = int(num_points / 2)
+    return current_curve[half_voltage_idx] / voltage_curve[half_voltage_idx]
+
+
+def low_high_n_SiO_x_vals(data, is_high_nonlinearity):
+    curves = low_high_n_SiO_x_curves(data)
+    if is_high_nonlinearity:
+        idx = 1
+    else:
+        idx = 0
+    voltage_curves, current_curves = curves[idx]
+
+    n = [nonlinearity_parameter(curve) for curve in current_curves]
+    n_avg, n_std = np.mean(n), np.std(n, ddof=1)
+
+    G_on = G_at_half_voltage(voltage_curves[0, :], current_curves[0, :])
+    G_off = G_at_half_voltage(voltage_curves[-1, :], current_curves[-1, :])
+    return G_off, G_on, n_avg, n_std
+
+
+def load_Ta_HfO2():
+    """Load Ta/HfO2 data.
+
+    Returns:
+        Array of shape `(num_cycles, num_pulses, num_bit_lines, num_word_lines)`.
+            The first half of `num_pulses` denotes potentiation, while the second
+            half denotes depression.
+    """
+    path = os.path.join(_create_and_get_data_dir(), "Ta_HfO2-data.mat")
+    _validate_data_path(path)
+    f = h5py.File(path, "r")
+    data = f.get("G_reads")
+    data = np.array(data)
+    return data
+
+
+def extract_G_off_and_G_on(data: np.ndarray) -> tuple[float, float]:
+    shape = data.shape
+    data = np.reshape(data, (shape[0] * shape[1], shape[2] * shape[3]))
+    G_offs = np.min(data, axis=0)
+    G_off = np.median(G_offs)
+    G_ons = np.max(data, axis=0)
+    G_on = np.median(G_ons)
+
+    return G_off, G_on
+
+
+def extract_stuck(data: np.ndarray, G_off: float, G_on: float) -> tuple[list[float], float]:
+    median_range = G_on - G_off
+    shape = data.shape
+    data = np.reshape(data, (shape[0] * shape[1], shape[2] * shape[3]))
+    mins = np.min(data, axis=0)
+    maxs = np.max(data, axis=0)
+    ranges = maxs - mins
+    means = np.mean(data, axis=0)
+    stuck_values = means[np.where(ranges < stuck_device_threshold(median_range))]
+    probability_stuck = stuck_values.shape[0] / means.shape[0]
+    return stuck_values.tolist(), probability_stuck
+
+
+def stuck_device_threshold(median_range):
+    return median_range / 2
+
+
+def _validate_data_path(path: str, url: Optional[str] = None) -> None:
+    if os.path.isfile(path):
+        return
+
+    if url is None:
+        raise ValueError(f'Data file "{path}" does not exist and the URL has not been provided.')
+
+    with open(path, "wb") as file:
+        response = requests.get(url)
+        file.write(response.content)
+
+
+def _create_and_get_data_dir() -> str:
+    dir_name = ".data"
+    os.makedirs(dir_name, exist_ok=True)
+    return dir_name
diff --git a/awarememristor/simulations/devices.py b/awarememristor/simulations/devices.py
new file mode 100644
index 0000000..a87661b
--- /dev/null
+++ b/awarememristor/simulations/devices.py
@@ -0,0 +1,102 @@
+from typing import Any
+
+from awarememristor.crossbar.nonidealities import (D2DLognormal,
+                                                   IVNonlinearity, Nonideality,
+                                                   StuckAtGOff, StuckAtGOn,
+                                                   StuckDistribution)
+from awarememristor.simulations import data
+
+
+def ideal() -> dict[str, Any]:
+    return {"G_off": None, "G_on": None, "nonidealities": []}
+
+
+def SiO_x_V_ref() -> dict[str, float]:
+    exp_data = data.load_SiO_x_multistate()
+    (voltages, _), (_, _) = data.low_high_n_SiO_x_curves(exp_data)
+    V_ref = voltages[0][-1] / 2
+
+    return {"V_ref": float(V_ref)}
+
+
+def _SiO_x_G(is_high_nonlinearity: bool) -> dict[str, float]:
+    exp_data = data.load_SiO_x_multistate()
+    G_off, G_on, _, _ = data.low_high_n_SiO_x_vals(exp_data, is_high_nonlinearity)
+    return {
+        "G_off": float(G_off),
+        "G_on": float(G_on),
+    }
+
+
+def _SiO_x_nonidealities(is_high_nonlinearity: bool) -> dict[str, list[Nonideality]]:
+    exp_data = data.load_SiO_x_multistate()
+    _, _, n_avg, n_std = data.low_high_n_SiO_x_vals(exp_data, is_high_nonlinearity)
+    V_ref = SiO_x_V_ref()["V_ref"]
+    return {
+        "nonidealities": [IVNonlinearity(V_ref, float(n_avg), float(n_std))],
+    }
+
+
+def SiO_x(is_high_nonlinearity: bool) -> dict[str, Any]:
+    return {
+        **_SiO_x_G(is_high_nonlinearity),
+        **_SiO_x_nonidealities(is_high_nonlinearity),
+    }
+
+
+def stuck_off() -> dict[str, Any]:
+    G = _SiO_x_G(True)
+    return {
+        **G,
+        "nonidealities": [
+            StuckAtGOff(G["G_off"], 0.05),
+        ],
+    }
+
+
+def SiO_x_high_nonlinearity_and_stuck_on() -> dict[str, Any]:
+    is_high_nonlinearity = True
+    G = _SiO_x_G(is_high_nonlinearity)
+    nonidealities = _SiO_x_nonidealities(is_high_nonlinearity)["nonidealities"] + [
+        StuckAtGOn(G["G_on"], 0.05)
+    ]
+    return {
+        **G,
+        "nonidealities": nonidealities,
+    }
+
+
+def more_uniform_d2d() -> dict[str, Any]:
+    G = _SiO_x_G(True)
+    return {
+        **G,
+        "nonidealities": [D2DLognormal(G["G_off"], G["G_on"], 0.25, 0.25)],
+    }
+
+
+def less_uniform_d2d() -> dict[str, Any]:
+    G = _SiO_x_G(True)
+    return {
+        **G,
+        "nonidealities": [D2DLognormal(G["G_off"], G["G_on"], 0.05, 0.5)],
+    }
+
+
+def high_magnitude_more_uniform_d2d() -> dict[str, Any]:
+    G = _SiO_x_G(True)
+    return {
+        **G,
+        "nonidealities": [D2DLognormal(G["G_off"], G["G_on"], 0.5, 0.5)],
+    }
+
+
+def Ta_HfO2() -> dict[str, Any]:
+    exp_data = data.load_Ta_HfO2()
+    G_off, G_on = data.extract_G_off_and_G_on(exp_data)
+    G_off, G_on = float(G_off), float(G_on)
+    vals, p = data.extract_stuck(exp_data, G_off, G_on)
+    return {
+        "G_off": G_off,
+        "G_on": G_on,
+        "nonidealities": [StuckDistribution(vals, p)],
+    }
diff --git a/awarememristor/simulations/ideal.py b/awarememristor/simulations/ideal.py
new file mode 100644
index 0000000..9db8fe7
--- /dev/null
+++ b/awarememristor/simulations/ideal.py
@@ -0,0 +1,49 @@
+from awarememristor.simulations import (iv_nonlinearity,
+                                        iv_nonlinearity_and_stuck_on,
+                                        iv_nonlinearity_cnn,
+                                        memristive_validation,
+                                        stuck_distribution, stuck_off,
+                                        weight_implementation)
+from awarememristor.training.iterator import Iterator
+
+
+def get_mnist_iterator():
+    iterators = [
+        iv_nonlinearity.get_ideal_iterator(),
+        stuck_off.get_ideal_iterator(),
+        iv_nonlinearity_and_stuck_on.get_ideal_iterator(),
+        stuck_distribution.get_ideal_iterator(),
+        weight_implementation.get_ideal_iterator(),
+        memristive_validation.get_ideal_iterator(),
+    ]
+
+    return Iterator(
+        "mnist",
+        iterators[0].training,
+        [inference for iterator in iterators for inference in iterator.inferences],
+    )
+
+
+def get_cifar10_iterator():
+    iterators = [
+        iv_nonlinearity_cnn.get_ideal_iterator(),
+    ]
+
+    return Iterator(
+        "cifar10",
+        iterators[0].training,
+        [inference for iterator in iterators for inference in iterator.inferences],
+    )
+
+
+def get_iterators():
+    return [
+        get_mnist_iterator(),
+        get_cifar10_iterator(),
+    ]
+
+
+def main():
+    for iterator in get_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/simulations/iv_nonlinearity.py b/awarememristor/simulations/iv_nonlinearity.py
new file mode 100644
index 0000000..0d1f2ad
--- /dev/null
+++ b/awarememristor/simulations/iv_nonlinearity.py
@@ -0,0 +1,39 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(training_setup, inference_setups, is_regularized):
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(
+        **utils.get_training_params(), is_regularized=is_regularized, **training_setup
+    )
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    return custom_iterator(devices.ideal(), [devices.SiO_x(False), devices.SiO_x(True)], False)
+
+
+def get_nonideal_iterators():
+    return [
+        custom_iterator(devices.SiO_x(False), [devices.SiO_x(False)], False),
+        custom_iterator(devices.SiO_x(False), [devices.SiO_x(False)], True),
+        custom_iterator(devices.SiO_x(True), [devices.SiO_x(True)], False),
+        custom_iterator(devices.SiO_x(True), [devices.SiO_x(True)], True),
+    ]
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/simulations/iv_nonlinearity_and_stuck_on.py b/awarememristor/simulations/iv_nonlinearity_and_stuck_on.py
new file mode 100644
index 0000000..f79d838
--- /dev/null
+++ b/awarememristor/simulations/iv_nonlinearity_and_stuck_on.py
@@ -0,0 +1,39 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(training_setup, inference_setups):
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(**utils.get_training_params(), is_regularized=False, **training_setup)
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    return custom_iterator(devices.ideal(), [devices.SiO_x_high_nonlinearity_and_stuck_on()])
+
+
+def get_nonideal_iterators():
+    iterators = [
+        custom_iterator(
+            devices.SiO_x_high_nonlinearity_and_stuck_on(),
+            [devices.SiO_x_high_nonlinearity_and_stuck_on()],
+        ),
+    ]
+
+    return iterators
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/simulations/iv_nonlinearity_cnn.py b/awarememristor/simulations/iv_nonlinearity_cnn.py
new file mode 100644
index 0000000..af49995
--- /dev/null
+++ b/awarememristor/simulations/iv_nonlinearity_cnn.py
@@ -0,0 +1,34 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "cifar10"
+
+
+def custom_iterator(training_setup, inference_setups):
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(**utils.get_training_params(), is_regularized=False, **training_setup)
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    return custom_iterator(devices.ideal(), [devices.SiO_x(True)])
+
+
+def get_nonideal_iterators():
+    return [
+        custom_iterator(devices.SiO_x(True), [devices.SiO_x(True)]),
+    ]
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/simulations/memristive_validation.py b/awarememristor/simulations/memristive_validation.py
new file mode 100644
index 0000000..4133c0c
--- /dev/null
+++ b/awarememristor/simulations/memristive_validation.py
@@ -0,0 +1,55 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(
+    training_setup, inference_setups, use_combined_validation, num_training_repeats: int = None
+):
+    training_params = utils.get_training_params()
+    if num_training_repeats is not None:
+        training_params["num_repeats"] = num_training_repeats
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(
+        **training_params,
+        is_regularized=False,
+        use_combined_validation=use_combined_validation,
+        **training_setup
+    )
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    return custom_iterator(devices.ideal(), [devices.high_magnitude_more_uniform_d2d()], False)
+
+
+def get_nonideal_iterators():
+    return [
+        custom_iterator(
+            devices.high_magnitude_more_uniform_d2d(),
+            [devices.high_magnitude_more_uniform_d2d()],
+            True,
+            # Validation is utilized during training, so to evaluate the
+            # effectiveness of different methods, we need to increase the
+            # sample size of trained networks.
+            num_training_repeats=100,
+        ),
+    ]
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.training.is_standard_validation_mode = False
+        iterator.infer()
+        iterator.training.is_standard_validation_mode = True
+        iterator.infer()
diff --git a/awarememristor/simulations/nonideality_agnosticism.py b/awarememristor/simulations/nonideality_agnosticism.py
new file mode 100644
index 0000000..24dfebd
--- /dev/null
+++ b/awarememristor/simulations/nonideality_agnosticism.py
@@ -0,0 +1,59 @@
+from awarememristor.simulations import (devices, ideal, iv_nonlinearity,
+                                        iv_nonlinearity_and_stuck_on,
+                                        memristive_validation,
+                                        stuck_distribution, stuck_off, utils,
+                                        weight_implementation)
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(training_setup, inference_setups, is_regularized):
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(
+        **utils.get_training_params(), is_regularized=is_regularized, **training_setup
+    )
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_iterators():
+    inference_setups = [
+        devices.ideal(),
+        devices.SiO_x(False),
+        devices.SiO_x(True),
+        devices.stuck_off(),
+        devices.SiO_x_high_nonlinearity_and_stuck_on(),
+        devices.more_uniform_d2d(),
+        devices.less_uniform_d2d(),
+        devices.high_magnitude_more_uniform_d2d(),
+        devices.Ta_HfO2(),
+    ]
+
+    iterators = [
+        ideal.get_mnist_iterator(),
+        *iv_nonlinearity.get_nonideal_iterators(),
+        *iv_nonlinearity_and_stuck_on.get_nonideal_iterators(),
+        *stuck_off.get_nonideal_iterators(),
+        *weight_implementation.get_nonideal_iterators()[-4:],
+        *stuck_distribution.get_nonideal_iterators(),
+        memristive_validation.get_nonideal_iterators()[0],
+    ]
+    inferences = [
+        Inference(**utils.get_inference_params(), **inference_setup)
+        for inference_setup in inference_setups
+    ]
+
+    for idx, iterator in enumerate(iterators):
+        # Use the same number of repeats for all training setups.
+        iterator.training.num_repeats = utils.get_training_params()["num_repeats"]
+        for inference in inferences:
+            if inference not in iterator.inferences:
+                iterators[idx].inferences.append(inference)
+
+    return iterators
+
+
+def main():
+    for iterator in get_iterators():
+        iterator.infer()
diff --git a/awarememristor/simulations/stuck_distribution.py b/awarememristor/simulations/stuck_distribution.py
new file mode 100644
index 0000000..0873165
--- /dev/null
+++ b/awarememristor/simulations/stuck_distribution.py
@@ -0,0 +1,34 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(training_setup, inference_setups):
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(**utils.get_training_params(), is_regularized=False, **training_setup)
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    return custom_iterator(devices.ideal(), [devices.Ta_HfO2()])
+
+
+def get_nonideal_iterators():
+    return [
+        custom_iterator(devices.Ta_HfO2(), [devices.Ta_HfO2()]),
+    ]
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/simulations/stuck_off.py b/awarememristor/simulations/stuck_off.py
new file mode 100644
index 0000000..8968b6f
--- /dev/null
+++ b/awarememristor/simulations/stuck_off.py
@@ -0,0 +1,34 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(training_setup, inference_setups):
+    inferences = [Inference(**utils.get_inference_params(), **setup) for setup in inference_setups]
+    training = Training(**utils.get_training_params(), is_regularized=False, **training_setup)
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    return custom_iterator(devices.ideal(), [devices.stuck_off()])
+
+
+def get_nonideal_iterators():
+    return [
+        custom_iterator(devices.stuck_off(), [devices.stuck_off()]),
+    ]
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/simulations/utils.py b/awarememristor/simulations/utils.py
new file mode 100644
index 0000000..a13e3ca
--- /dev/null
+++ b/awarememristor/simulations/utils.py
@@ -0,0 +1,30 @@
+def get_training_params():
+    return {
+        "num_repeats": 5,
+        "num_epochs": 1000,
+        "batch_size": 64,
+    }
+
+
+def get_inference_params():
+    return {
+        "num_repeats": 25,
+    }
+
+
+def get_energy_efficiency(
+    avg_power: float,
+    num_neurons_lst: list[int] = [784, 25, 10],
+    read_time: float = 50e-9,
+):
+    num_synapses = get_num_synapses(num_neurons_lst)
+    energy_efficiency = (2 * num_synapses) / (read_time * avg_power)
+    return energy_efficiency
+
+
+def get_num_synapses(num_neurons_lst: list[int]):
+    num_synapses = 0
+    for idx, num_neurons in enumerate(num_neurons_lst[:-1]):
+        num_synapses += (num_neurons + 1) * num_neurons_lst[idx + 1]
+
+    return num_synapses
diff --git a/awarememristor/simulations/weight_implementation.py b/awarememristor/simulations/weight_implementation.py
new file mode 100644
index 0000000..df00fca
--- /dev/null
+++ b/awarememristor/simulations/weight_implementation.py
@@ -0,0 +1,82 @@
+from awarememristor.simulations import devices, utils
+from awarememristor.training.iterator import Inference, Iterator, Training
+
+DATASET = "mnist"
+
+
+def custom_iterator(
+    training_setup,
+    inference_setups,
+    is_regularized=False,
+    force_standard_w=False,
+    mapping_rule="default",
+):
+    inferences = [
+        Inference(mapping_rule=mapping_rule, **utils.get_inference_params(), **setup)
+        for setup in inference_setups
+    ]
+    training = Training(
+        **utils.get_training_params(),
+        is_regularized=is_regularized,
+        force_standard_w=force_standard_w,
+        mapping_rule=mapping_rule,
+        **training_setup
+    )
+
+    return Iterator(DATASET, training, inferences)
+
+
+def get_ideal_iterator():
+    iterator = custom_iterator(
+        devices.ideal(), [devices.more_uniform_d2d(), devices.less_uniform_d2d()], False
+    )
+
+    return iterator
+
+
+def get_nonideal_iterators():
+    return [
+        custom_iterator(
+            devices.more_uniform_d2d(),
+            [devices.more_uniform_d2d()],
+            force_standard_w=True,
+            mapping_rule="avg",
+        ),
+        custom_iterator(
+            devices.more_uniform_d2d(),
+            [devices.more_uniform_d2d()],
+            force_standard_w=True,
+        ),
+        custom_iterator(
+            devices.less_uniform_d2d(),
+            [devices.less_uniform_d2d()],
+            force_standard_w=True,
+            mapping_rule="avg",
+        ),
+        custom_iterator(
+            devices.less_uniform_d2d(),
+            [devices.less_uniform_d2d()],
+            force_standard_w=True,
+        ),
+        custom_iterator(devices.more_uniform_d2d(), [devices.more_uniform_d2d()]),
+        custom_iterator(
+            devices.more_uniform_d2d(), [devices.more_uniform_d2d()], is_regularized=True
+        ),
+        custom_iterator(devices.less_uniform_d2d(), [devices.less_uniform_d2d()]),
+        custom_iterator(
+            devices.less_uniform_d2d(), [devices.less_uniform_d2d()], is_regularized=True
+        ),
+    ]
+
+
+def get_iterators():
+    return [
+        get_ideal_iterator(),
+        *get_nonideal_iterators(),
+    ]
+
+
+def main():
+    for iterator in get_nonideal_iterators():
+        iterator.train(use_test_callback=True)
+        iterator.infer()
diff --git a/awarememristor/training/__init__.py b/awarememristor/training/__init__.py
new file mode 100644
index 0000000..c426286
--- /dev/null
+++ b/awarememristor/training/__init__.py
@@ -0,0 +1 @@
+from awarememristor.training import iterator
diff --git a/awarememristor/training/architecture.py b/awarememristor/training/architecture.py
new file mode 100644
index 0000000..1c69491
--- /dev/null
+++ b/awarememristor/training/architecture.py
@@ -0,0 +1,229 @@
+import numpy as np
+import tensorflow as tf
+from tensorflow.keras import constraints, layers, models
+
+from awarememristor import crossbar
+from awarememristor.training import utils
+
+
+def get_model(iterator, custom_weights=None, custom_weights_path=None):
+    num_hidden_neurons = 25
+    if iterator.dataset == "mnist":
+        model = models.Sequential()
+        model.add(layers.Flatten(input_shape=(28, 28)))
+        model.add(MemristorDense(784, num_hidden_neurons, iterator))
+        model.add(layers.Activation("sigmoid"))
+        model.add(MemristorDense(num_hidden_neurons, 10, iterator))
+        model.add(layers.Activation("softmax"))
+    elif iterator.dataset == "cifar10":
+        model = models.Sequential()
+
+        # Convolutional layers
+        model.add(layers.Conv2D(32, (3, 3), activation="relu", input_shape=(32, 32, 3)))
+        model.add(layers.MaxPooling2D((2, 2)))
+        model.add(layers.Conv2D(64, (3, 3), activation="relu"))
+        model.add(layers.MaxPooling2D((2, 2)))
+        # Ensure inputs bounded between 0 and 1 for first memristive layer.
+        model.add(layers.Conv2D(64, (3, 3), activation=tf.keras.layers.ReLU(max_value=1.0)))
+
+        # Fully connected layers
+        model.add(layers.Flatten())
+        model.add(MemristorDense(1024, num_hidden_neurons, iterator))
+        model.add(layers.Activation("sigmoid"))
+        model.add(MemristorDense(num_hidden_neurons, 10, iterator))
+        model.add(layers.Activation("softmax"))
+    else:
+        raise ValueError(f"Dataset {iterator.dataset} is not recognised!")
+
+    if custom_weights is not None:
+        model.set_weights(custom_weights)
+    elif custom_weights_path is not None:
+        model.load_weights(custom_weights_path)
+    elif not iterator.is_training:
+        model.load_weights(iterator.weights_path())
+
+    model.compile(
+        optimizer=tf.keras.optimizers.SGD(),
+        loss=tf.keras.losses.SparseCategoricalCrossentropy(),
+        metrics=["accuracy"],
+    )
+    return model
+
+
+class MemristorDense(layers.Layer):
+    def __init__(self, n_in, n_out, iterator, **kwargs):
+        self.n_in = n_in
+        self.n_out = n_out
+        self.iterator = iterator
+        super(MemristorDense, self).__init__(**kwargs)
+
+    # Adding this function removes an issue with custom layer checkpoint
+    def get_config(self):
+        config = super().get_config().copy()
+        config.update(
+            {
+                "n_in": self.n_in,
+                "n_out": self.n_out,
+            }
+        )
+        return config
+
+    # Create trainable weights and biases
+    def build(self, input_shape):
+        stdv = 1 / np.sqrt(self.n_in)
+
+        kwargs = {}
+        if self.iterator.training.is_regularized:
+            reg_gamma = 1e-4
+            kwargs["regularizer"] = tf.keras.regularizers.l1(reg_gamma)
+
+        if self.iterator.training.uses_double_weights():
+            self.w_pos = self.add_weight(
+                shape=(self.n_in, self.n_out),
+                initializer=tf.keras.initializers.RandomNormal(mean=0.5, stddev=stdv),
+                name="weights_pos",
+                trainable=True,
+                constraint=constraints.NonNeg(),
+                **kwargs,
+            )
+
+            self.w_neg = self.add_weight(
+                shape=(self.n_in, self.n_out),
+                initializer=tf.keras.initializers.RandomNormal(mean=0.5, stddev=stdv),
+                name="weights_neg",
+                trainable=True,
+                constraint=constraints.NonNeg(),
+                **kwargs,
+            )
+
+            self.b_pos = self.add_weight(
+                shape=(self.n_out,),
+                initializer=tf.keras.initializers.Constant(value=0.5),
+                name="biases_pos",
+                trainable=True,
+                constraint=constraints.NonNeg(),
+                **kwargs,
+            )
+
+            self.b_neg = self.add_weight(
+                shape=(self.n_out,),
+                initializer=tf.keras.initializers.Constant(value=0.5),
+                name="biases_neg",
+                trainable=True,
+                constraint=constraints.NonNeg(),
+                **kwargs,
+            )
+        else:
+            self.w = self.add_weight(
+                shape=(self.n_in, self.n_out),
+                initializer=tf.keras.initializers.RandomNormal(mean=0.0, stddev=stdv),
+                name="weights",
+                trainable=True,
+                **kwargs,
+            )
+
+            self.b = self.add_weight(
+                shape=(self.n_out,),
+                initializer=tf.keras.initializers.Constant(value=0.0),
+                name="biases",
+                trainable=True,
+                **kwargs,
+            )
+
+    def combined_weights(self):
+        if self.iterator.training.uses_double_weights():
+            b_pos = tf.expand_dims(self.b_pos, axis=0)
+            b_neg = tf.expand_dims(self.b_neg, axis=0)
+            combined_weights_pos = tf.concat([self.w_pos, b_pos], 0)
+            combined_weights_neg = tf.concat([self.w_neg, b_neg], 0)
+
+            # Interleave positive and negative weights
+            combined_weights = tf.reshape(
+                tf.concat(
+                    [
+                        combined_weights_pos[..., tf.newaxis],
+                        combined_weights_neg[..., tf.newaxis],
+                    ],
+                    axis=-1,
+                ),
+                [tf.shape(combined_weights_pos)[0], -1],
+            )
+        else:
+            bias = tf.expand_dims(self.b, axis=0)
+            combined_weights = tf.concat([self.w, bias], 0)
+
+        return combined_weights
+
+    def call(self, x, mask=None):
+        if (
+            not self.iterator.training.uses_double_weights()
+            and not self.iterator.current_stage().is_nonideal()
+        ):
+            return tf.tensordot(x, self.w, axes=1) + self.b
+
+        ones = tf.ones([tf.shape(x)[0], 1])
+        inputs = tf.concat([x, ones], 1)
+
+        self.out = self.memristive_outputs(inputs, self.combined_weights())
+
+        return self.out
+
+    def memristive_outputs(self, x, weights):
+        current_stage = self.iterator.current_stage()
+
+        if current_stage.is_nonideal():
+            G_off = current_stage.G_off
+            G_on = current_stage.G_on
+            k_V = current_stage.k_V()
+        else:
+            # Handle case when training is aware, but inference assumes no
+            # nonidealities. This will not affect accuracy, but will affect
+            # power consumption.
+            G_off = self.iterator.training.G_off
+            G_on = self.iterator.training.G_on
+            k_V = self.iterator.training.k_V()
+
+        # Mapping inputs onto voltages.
+        V = crossbar.map.x_to_V(x, k_V)
+
+        # Mapping weights onto conductances.
+        if self.iterator.training.uses_double_weights():
+            G, max_weight = crossbar.map.double_w_to_G(weights, G_off, G_on)
+        else:
+            G, max_weight = crossbar.map.w_to_G(weights, G_off, G_on, current_stage.mapping_rule)
+
+        # Linearity-preserving nonidealities
+        for nonideality in current_stage.nonidealities:
+            if isinstance(nonideality, crossbar.nonidealities.LinearityPreserving):
+                G = nonideality.disturb_G(G)
+
+        # Linearity-nonpreserving nonidealities
+        I = None
+        I_ind = None
+        for nonideality in current_stage.nonidealities:
+            if isinstance(nonideality, crossbar.nonidealities.LinearityNonpreserving):
+                I, I_ind = nonideality.compute_I(V, G)
+
+        # Ideal case for computing output currents.
+        if I is None or I_ind is None:
+            if self.iterator.is_training:
+                I = crossbar.ideal.compute_I(V, G)
+            else:
+                I, I_ind = crossbar.ideal.compute_I_all(V, G)
+
+        if self.iterator.compute_power:
+            power_path = self.iterator.power_path()
+            P_avg = utils.compute_avg_crossbar_power(V, I_ind)
+            with open(power_path, mode="a", encoding="utf-8"):
+                tf.print(P_avg, output_stream=f"file://{power_path}")
+
+        # Converting to outputs.
+        y_disturbed = crossbar.map.I_to_y(I, k_V, max_weight, G_on, G_off)
+
+        return y_disturbed
+
+    def get_output_shape_for(self, input_shape):
+        return (input_shape[0], self.n_out)
+
+    def compute_output_shape(self, input_shape):
+        return (input_shape[0], self.n_out)
diff --git a/awarememristor/training/callbacks.py b/awarememristor/training/callbacks.py
new file mode 100644
index 0000000..fe7ff15
--- /dev/null
+++ b/awarememristor/training/callbacks.py
@@ -0,0 +1,310 @@
+import copy
+import os
+import time
+from abc import ABC, abstractmethod
+from typing import Any
+
+import numpy as np
+import tensorflow as tf
+
+from awarememristor.training import architecture
+
+
+class Callback(ABC):
+    """Abstract class that requires implementation of callback name."""
+
+    @staticmethod
+    @abstractmethod
+    def name() -> str:
+        """Returns name of the callback."""
+
+
+class Checkpoint:
+    """Used only to mark certain callbacks as checkpoint callbacks."""
+
+
+class MemristiveCallback(tf.keras.callbacks.Callback):
+    """Computes a metric multiple times in order to take the stochastic nature
+    of memristive devices into account."""
+
+    def __init__(self, iterator, history=None) -> None:
+        self.iterator = copy.copy(iterator)
+        self.validation_freq = 20
+        if iterator.training.memristive_validation_freq is not None:
+            self.validation_freq = iterator.training.memristive_validation_freq
+        self.testing_freq = 20
+        self.num_repeats = 20
+        self.history = history
+
+    def should_skip_epoch(self, epoch, is_validation=False) -> bool:
+        freq = self.testing_freq
+        if is_validation:
+            freq = self.validation_freq
+        # Will evaluate on first epoch and then every `freq` epochs.
+        if epoch != 0 and (epoch + 1) % freq != 0:
+            return True
+        return False
+
+    def evaluate(self, model, data, num_repeats: int = None):
+        if num_repeats is None:
+            num_repeats = self.num_repeats
+
+        accuracy = []
+        loss = []
+
+        start_time = time.time()
+        for _ in range(num_repeats):
+            single_loss, single_accuracy = model.evaluate(data, verbose=0)
+            loss.append(single_loss)
+            accuracy.append(single_accuracy)
+        num_total_batches = data.cardinality().numpy() * num_repeats
+
+        end_time = time.time()
+        duration = int(end_time - start_time)
+        if num_repeats == 1:
+            loss = loss[0]
+            accuracy = accuracy[0]
+
+        return loss, accuracy, duration, num_total_batches
+
+    def evaluation_results_str(
+        self,
+        num_total_batches: int,
+        duration: int,
+        loss: float,
+        accuracy: float,
+        prepend_metrics: str = None,
+        label: str = None,
+    ) -> str:
+        str_ = f"{num_total_batches}/{num_total_batches}"
+        str_ += f" - {duration}s - "
+        if prepend_metrics:
+            str_ += f"{prepend_metrics}_"
+        str_ += f"loss: {loss:.4f} - "
+        if prepend_metrics:
+            str_ += f"{prepend_metrics}_"
+        str_ += f"accuracy: {accuracy:.4f}"
+        if label is not None:
+            str_ += f" [{label}]"
+
+        return str_
+
+    def saving_weights_str(
+        self, accuracy: float, previous_best_accuracy: float, prepend: str = None
+    ) -> str:
+        str_ = ""
+        if prepend is not None:
+            str_ += f"{prepend}_"
+        str_ += f"accuracy ({accuracy:.4f}) improved over previous best result ({previous_best_accuracy:.4f}). Saving weights..."
+        return str_
+
+    def info(self) -> dict[str, Any]:
+        return {
+            "history": self.history,
+        }
+
+
+class TestCallback(MemristiveCallback, Callback):
+    """Compute test accuracy for all inference setups during training."""
+
+    def __init__(self, iterator) -> None:
+        MemristiveCallback.__init__(self, iterator)
+        self.iterator.is_training = False
+        self.history = [
+            {
+                "label": inference.label(),
+                "epoch_no": [],
+                "loss": [],
+                "accuracy": [],
+            }
+            for inference in self.iterator.inferences
+        ]
+
+    def on_epoch_end(self, epoch, logs=None):
+        if self.should_skip_epoch(epoch, is_validation=False):
+            return
+
+        model_weights = self.model.get_weights()
+
+        for inference_idx, inference in enumerate(self.iterator.inferences):
+            self.iterator.inference_idx = inference_idx
+            data = self.iterator.data("testing")
+            callback_model = architecture.get_model(self.iterator, custom_weights=model_weights)
+            loss, accuracy, duration, num_total_batches = self.evaluate(callback_model, data)
+            self.history[inference_idx]["loss"].append(loss)
+            self.history[inference_idx]["accuracy"].append(accuracy)
+            self.history[inference_idx]["epoch_no"].append(epoch + 1)
+            results_str = self.evaluation_results_str(
+                num_total_batches,
+                duration,
+                np.median(loss),
+                np.median(accuracy),
+                prepend_metrics="median_test",
+                label=inference.nonideality_label(),
+            )
+            print(results_str)
+
+        self.iterator.inference_idx = 0
+
+    @staticmethod
+    def name():
+        return "memristive_test"
+
+
+class MemristiveCheckpoint(MemristiveCallback, Callback, Checkpoint):
+    """Evaluate accuracy on validation set multiple times to provide a more reliable measure of
+    learning progress.
+    """
+
+    def __init__(self, iterator) -> None:
+        MemristiveCallback.__init__(self, iterator)
+        self.iterator.is_training = True
+        self.best_median_val_accuracy = 0.0
+        self.history = {"epoch_no": [], "loss": [], "accuracy": []}
+
+    def on_epoch_end(self, epoch, logs=None):
+        if self.should_skip_epoch(epoch, is_validation=True):
+            return
+
+        data = self.iterator.data("validation")
+        loss, accuracy, duration, num_total_batches = self.evaluate(self.model, data)
+
+        self.history["loss"].append(loss)
+        self.history["accuracy"].append(accuracy)
+        self.history["epoch_no"].append(epoch + 1)
+
+        median_val_loss = np.median(loss)
+        median_val_accuracy = np.median(accuracy)
+        print(
+            self.evaluation_results_str(
+                num_total_batches,
+                duration,
+                median_val_loss,
+                median_val_accuracy,
+                prepend_metrics="median_val",
+            )
+        )
+
+        if median_val_accuracy > self.best_median_val_accuracy:
+            print(
+                self.saving_weights_str(
+                    median_val_accuracy, self.best_median_val_accuracy, prepend="median_val"
+                )
+            )
+            self.best_median_val_accuracy = median_val_accuracy
+            self.model.save_weights(self.iterator.weights_path())
+
+    @staticmethod
+    def name():
+        return "memristive_checkpoint"
+
+
+class StandardCheckpoint(tf.keras.callbacks.ModelCheckpoint, Callback, Checkpoint):
+    """Same as `tf.keras.callbacks.ModelCheckpoint`, but with a `name()`."""
+
+    def __init__(self, iterator) -> None:
+        tf.keras.callbacks.ModelCheckpoint.__init__(
+            self,
+            iterator.weights_path(),
+            monitor="val_accuracy",
+            save_best_only=True,
+        )
+
+    @staticmethod
+    def name():
+        return "standard_checkpoint"
+
+
+class CombinedCheckpoint(MemristiveCallback, Callback, Checkpoint):
+    """Used to test the effectiveness of memristive validation.
+
+    Two validation techniques (standard and memristive) are applied at the same
+    time during training.
+    """
+
+    def __init__(self, iterator) -> None:
+        MemristiveCallback.__init__(self, iterator)
+        self.iterator.is_training = True
+        self.best_median_val_accuracy = 0.0
+        self.best_standard_val_accuracy = 0.0
+        self.history = {
+            "epoch_no": [],
+            "loss": [],
+            "accuracy": [],
+            "standard_epoch_no": [],
+            "standard_loss": [],
+            "standard_accuracy": [],
+        }
+
+    def on_epoch_end(self, epoch, logs=None):
+        data = self.iterator.data("validation")
+
+        if self.should_skip_epoch(epoch, is_validation=True):
+            single_loss, single_accuracy, duration, num_total_batches = self.evaluate(
+                self.model,
+                data,
+                num_repeats=1,
+            )
+            self.history["standard_loss"].append(single_loss)
+            self.history["standard_accuracy"].append(single_accuracy)
+            self.history["standard_epoch_no"].append(epoch + 1)
+            print(
+                self.evaluation_results_str(
+                    num_total_batches, duration, single_loss, single_accuracy, prepend_metrics="val"
+                )
+            )
+        else:
+            loss, accuracy, duration, num_total_batches = self.evaluate(self.model, data)
+
+            self.history["loss"].append(loss)
+            self.history["accuracy"].append(accuracy)
+            self.history["epoch_no"].append(epoch + 1)
+
+            median_val_loss = np.median(loss)
+            median_val_accuracy = np.median(accuracy)
+            print(
+                self.evaluation_results_str(
+                    num_total_batches,
+                    duration,
+                    median_val_loss,
+                    median_val_accuracy,
+                    prepend_metrics="median_val",
+                )
+            )
+
+            single_loss = loss[0]
+            single_accuracy = accuracy[0]
+            self.history["standard_loss"].append(single_loss)
+            self.history["standard_accuracy"].append(single_accuracy)
+            self.history["standard_epoch_no"].append(epoch + 1)
+            print(
+                self.evaluation_results_str(
+                    0, 0, single_loss, single_accuracy, prepend_metrics="val"
+                )
+            )
+
+            if median_val_accuracy > self.best_median_val_accuracy:
+                self.iterator.training.is_standard_validation_mode = True
+                os.makedirs(self.iterator.weights_dir(), exist_ok=True)
+                print(
+                    self.saving_weights_str(
+                        median_val_accuracy, self.best_median_val_accuracy, prepend="median_val"
+                    )
+                )
+                self.best_median_val_accuracy = median_val_accuracy
+                self.model.save_weights(self.iterator.weights_path())
+
+        if single_accuracy > self.best_standard_val_accuracy:
+            self.iterator.training.is_standard_validation_mode = False
+            os.makedirs(self.iterator.weights_dir(), exist_ok=True)
+            print(
+                self.saving_weights_str(
+                    single_accuracy, self.best_standard_val_accuracy, prepend="val"
+                )
+            )
+            self.best_standard_val_accuracy = single_accuracy
+            self.model.save_weights(self.iterator.weights_path())
+
+    @staticmethod
+    def name():
+        return "combined_checkpoint"
diff --git a/awarememristor/training/iterator.py b/awarememristor/training/iterator.py
new file mode 100644
index 0000000..e2073c5
--- /dev/null
+++ b/awarememristor/training/iterator.py
@@ -0,0 +1,498 @@
+import os
+import pickle
+import warnings
+from typing import Any
+
+import numpy as np
+import tensorflow as tf
+import tensorflow_datasets as tfds
+
+from awarememristor.crossbar.nonidealities import (LinearityNonpreserving,
+                                                   LinearityPreserving,
+                                                   Nonideality)
+from awarememristor.simulations import devices
+from awarememristor.training import callbacks, network, utils
+
+warnings.simplefilter("default")
+
+
+class Iterable:
+    """Used to keep track of training or inference repeats."""
+
+    def __init__(self, num_repeats: int) -> None:
+        self.repeat_idx = 0
+        self.num_repeats = num_repeats
+
+    def __eq__(self, other):
+        return self.repeat_idx == other.repeat_idx and self.num_repeats == other.num_repeats
+
+
+class Stage(Iterable):
+    """Used for training and inference."""
+
+    def __init__(
+        self,
+        G_off: float = None,
+        G_on: float = None,
+        nonidealities: list[Nonideality] = None,
+        mapping_rule: str = "default",
+        num_repeats: int = 0,
+    ) -> None:
+        self.G_off = G_off
+        self.G_on = G_on
+        if nonidealities is None:
+            nonidealities = []
+        self.nonidealities = nonidealities
+        self.mapping_rule = mapping_rule
+        self.validate_nonidealities()
+        Iterable.__init__(self, num_repeats)
+
+    def __eq__(self, other):
+        return (
+            self.G_off == other.G_off
+            and self.G_on == other.G_on
+            and self.nonidealities == other.nonidealities
+            and self.mapping_rule == other.mapping_rule
+            and Iterable.__eq__(self, other)
+        )
+
+    def conductance_label(self) -> str:
+        if self.G_off is None and self.G_on is None:
+            return "none_none"
+
+        return f"{self.G_off:.3g}_{self.G_on:.3g}"
+
+    def nonideality_label(self) -> str:
+        if len(self.nonidealities) == 0:
+            return "ideal"
+
+        l = "+".join(nonideality.label() for nonideality in self.nonidealities)
+        if self.mapping_rule != "default":
+            l += f"__{self.mapping_rule}"
+        return l
+
+    def label(self) -> str:
+        return f"{self.conductance_label()}__{self.nonideality_label()}"
+
+    def is_nonideal(self) -> bool:
+        return len(self.nonidealities) > 0
+
+    def k_V(self) -> float:
+        for nonideality in self.nonidealities:
+            if isinstance(nonideality, LinearityNonpreserving):
+                return nonideality.k_V()
+
+        # Except for power consumption, `k_V` makes no difference for
+        # linearity-preserving nonidealities, thus using the same value as for
+        # SiO_x devices.
+        return 2 * devices.SiO_x_V_ref()["V_ref"]
+
+    def validate_nonidealities(self) -> None:
+        """Multiple linearity-preserving nonidealities or multiple
+        linearity-nonpreserving nonidealities are not currently supported; this
+        function makes sure that this condition is satisfied."""
+        num_linearity_preserving = 0
+        num_linearity_nonpreserving = 0
+        for nonideality in self.nonidealities:
+            if isinstance(nonideality, LinearityPreserving):
+                num_linearity_preserving += 1
+            elif isinstance(nonideality, LinearityNonpreserving):
+                num_linearity_nonpreserving += 1
+
+        for num, nonideality_type in zip(
+            [num_linearity_preserving, num_linearity_nonpreserving],
+            ["linearity-preserving", "linearity-nonpreserving"],
+        ):
+            if num > 1:
+                raise ValueError(
+                    f"Current implementation does not support more than one {nonideality_type} nonideality."
+                )
+
+
+class Training(Stage, Iterable):
+    """Training configuration."""
+
+    def __init__(
+        self,
+        batch_size: int = 1,
+        validation_split: float = 0.2,
+        num_epochs: int = 1,
+        is_regularized: bool = False,
+        num_repeats: int = 0,
+        G_off: float = None,
+        G_on: float = None,
+        nonidealities: list[Nonideality] = None,
+        use_combined_validation: bool = False,
+        memristive_validation_freq: int = None,
+        mapping_rule: str = "default",
+        force_standard_w: bool = False,
+    ) -> None:
+        self.batch_size = batch_size
+        self.num_epochs = num_epochs
+        self.is_regularized = is_regularized
+        self.validation_split = validation_split
+        self.use_combined_validation = use_combined_validation
+        self.is_standard_validation_mode = False
+        self.memristive_validation_freq = memristive_validation_freq
+        self.force_standard_w = force_standard_w
+        Stage.__init__(
+            self,
+            G_off=G_off,
+            G_on=G_on,
+            nonidealities=nonidealities,
+            mapping_rule=mapping_rule,
+            num_repeats=num_repeats,
+        )
+
+    def regularized_label(self) -> str:
+        if self.is_regularized:
+            return "reg"
+        else:
+            return "nonreg"
+
+    def label(self) -> str:
+        l = f"{self.regularized_label()}__{self.batch_size}__{Stage.label(self)}"
+        if self.memristive_validation_freq is not None:
+            l += f"__val_freq_{self.memristive_validation_freq}"
+        if self.force_standard_w:
+            l += "__standard_w"
+        return l
+
+    def network_label(self) -> str:
+        return f"network-{self.repeat_idx}"
+
+    def uses_double_weights(self) -> bool:
+        return self.is_nonideal() and not self.force_standard_w
+
+
+class Inference(Stage):
+    """Inference configuration."""
+
+    def __init__(
+        self,
+        num_repeats: int = 0,
+        G_off: float = None,
+        G_on: float = None,
+        nonidealities: list[Nonideality] = None,
+        mapping_rule: str = "default",
+    ) -> None:
+        self.num_repeats = num_repeats
+        Stage.__init__(
+            self,
+            G_off=G_off,
+            G_on=G_on,
+            nonidealities=nonidealities,
+            mapping_rule=mapping_rule,
+            num_repeats=num_repeats,
+        )
+
+    def repeat_label(self) -> str:
+        return f"repeat-{self.repeat_idx}"
+
+
+class Iterator:
+    """A helper class used in simulations of memristive neural networks.
+
+    It is used to combine different training and inference setups.
+    """
+
+    def __init__(self, dataset: str, training: Training, inferences: list[Inference]) -> None:
+        self.dataset = dataset
+        self.training = training
+        self.inferences = inferences
+        self.compute_power = False
+        self.is_training = False
+        self.inference_idx = 0
+        self.test_batch_size = 100  # divisor of the size of the test set
+        self.__training_data = None
+        self.__validation_data = None
+        self.__testing_data = None
+        self.__train_split_boundary = int(100 * (1 - self.training.validation_split))
+
+    def data(self, subset: str) -> tf.data.Dataset:
+        if subset == "training":
+            if self.__training_data is not None:
+                return self.__training_data
+            split = f"train[:{self.__train_split_boundary}%]"
+        elif subset == "validation":
+            if self.__validation_data is not None:
+                return self.__validation_data
+            split = f"train[{self.__train_split_boundary}%:]"
+        elif subset == "testing":
+            if self.__testing_data is not None:
+                return self.__testing_data
+            split = "test"
+        else:
+            raise ValueError(f'Subset "{subset}" is not recognised!')
+
+        ds = tfds.load(
+            self.dataset,
+            split=split,
+            as_supervised=True,
+            shuffle_files=True,
+        )
+        size = ds.cardinality().numpy()
+
+        ds = ds.map(utils.normalize_img, num_parallel_calls=tf.data.AUTOTUNE)
+        if subset == "testing":
+            ds = ds.batch(self.test_batch_size)
+            ds = ds.cache()
+        else:
+            ds = ds.cache()
+            ds = ds.shuffle(size)
+            ds = ds.batch(self.training.batch_size)
+            if self.dataset == "cifar10" and subset == "training":
+                data_augmentation = tf.keras.Sequential(
+                    [
+                        tf.keras.layers.RandomTranslation(0.1, 0.1),
+                        tf.keras.layers.RandomFlip("horizontal"),
+                    ]
+                )
+                ds = ds.map(
+                    lambda x, y: (data_augmentation(x, training=True), y),
+                    num_parallel_calls=tf.data.AUTOTUNE,
+                )
+        ds = ds.prefetch(tf.data.AUTOTUNE)
+
+        if subset == "training":
+            self.__training_data = ds
+        elif subset == "validation":
+            self.__validation_data = ds
+        elif subset == "testing":
+            self.__testing_data = ds
+
+        print(f'Loaded dataset "{self.dataset}" ({subset}): {size} examples.')
+
+        return ds
+
+    def training_dir(self) -> str:
+        return os.path.join(os.getcwd(), "models", self.dataset, self.training.label())
+
+    def network_dir(self) -> str:
+        return os.path.join(self.training_dir(), self.training.network_label())
+
+    def weights_dir(self) -> str:
+        if self.training.use_combined_validation:
+            if self.training.is_standard_validation_mode:
+                return os.path.join(self.network_dir(), "standard-validation")
+            return os.path.join(self.network_dir(), "memristive-validation")
+        return self.network_dir()
+
+    def weights_path(self) -> str:
+        filename = "model.h5"
+        return os.path.join(self.weights_dir(), filename)
+
+    def info_path(self) -> str:
+        return os.path.join(self.network_dir(), "info.pkl")
+
+    def inference_nonideality_dir(self) -> str:
+        return os.path.join(self.weights_dir(), self.inferences[self.inference_idx].label())
+
+    def inference_repeat_dir(self) -> str:
+        return os.path.join(
+            self.inference_nonideality_dir(),
+            self.inferences[self.inference_idx].repeat_label(),
+        )
+
+    def power_path(self) -> str:
+        return os.path.join(self.inference_repeat_dir(), "power.csv")
+
+    def loss_path(self) -> str:
+        return os.path.join(self.inference_repeat_dir(), "loss.csv")
+
+    def accuracy_path(self) -> str:
+        return os.path.join(self.inference_repeat_dir(), "accuracy.csv")
+
+    def info(self) -> dict[str, Any]:
+        with open(self.info_path(), "rb") as pickle_file:
+            return pickle.load(pickle_file)
+
+    def current_stage(self) -> Stage:
+        if self.is_training:
+            return self.training
+        return self.inferences[self.inference_idx]
+
+    def training_curves(self, metric: str) -> tuple[np.ndarray, np.ndarray]:
+        if metric == "error":
+            y = self.info()["history"]["accuracy"]
+        else:
+            y = self.info()["history"][metric]
+
+        num_epochs = len(y)
+        x = np.arange(1, num_epochs + 1)
+
+        y = np.array(y)
+        if metric == "error":
+            y = 1 - y
+
+        return x, y
+
+    def _checkpoint_from_info(self) -> str:
+        if not self.training.is_nonideal():
+            return callbacks.StandardCheckpoint.name()
+        if self.training.use_combined_validation:
+            return callbacks.CombinedCheckpoint.name()
+        return callbacks.MemristiveCheckpoint.name()
+
+    def validation_curves(self, metric: str) -> tuple[np.ndarray, np.ndarray]:
+        checkpoint_name = self._checkpoint_from_info()
+        info = self.info()
+        if checkpoint_name == callbacks.StandardCheckpoint.name():
+            if metric == "error":
+                y = info["history"]["val_accuracy"]
+            else:
+                y = info["history"]["val_" + metric]
+            num_epochs = len(y)
+            x = np.arange(1, num_epochs + 1)
+        elif checkpoint_name == callbacks.MemristiveCheckpoint.name():
+            history = info["callback_infos"][callbacks.MemristiveCheckpoint.name()]["history"]
+            x = history["epoch_no"]
+            x = np.array(x)
+            if metric == "error":
+                y = history["accuracy"]
+            else:
+                y = history[metric]
+        elif checkpoint_name == callbacks.CombinedCheckpoint.name():
+            prepend = ""
+            if self.training.is_standard_validation_mode:
+                prepend = "standard_"
+            history = info["callback_infos"][callbacks.CombinedCheckpoint.name()]["history"]
+            x = history[f"{prepend}epoch_no"]
+            x = np.array(x)
+            if metric == "error":
+                y = history[f"{prepend}accuracy"]
+            else:
+                y = history[f"{prepend}{metric}"]
+
+        y = np.array(y)
+        if metric == "error":
+            y = 1 - y
+
+        return x, y
+
+    def training_testing_curves(self, metric: str, inference: Inference):
+        """Data from test callbacks during training."""
+
+        history = self.info()["callback_infos"]["memristive_test"]["history"][
+            self._memristive_test_callback_idx(inference)
+        ]
+
+        if metric == "error":
+            y = history["accuracy"]
+        else:
+            y = history[metric]
+
+        x = np.array(history["epoch_no"])
+
+        y = np.array(y)
+        if metric == "error":
+            y = 1 - y
+
+        return x, y
+
+    def _memristive_test_callback_idx(self, inference: Inference) -> int:
+        """Number of inferences might not equal the number of memristive test callbacks."""
+        label = inference.label()
+        for idx, history in enumerate(self.info()["callback_infos"]["memristive_test"]["history"]):
+            try:
+                if history["label"] == label:
+                    return idx
+            except KeyError:
+                break
+
+        raise ValueError("Index not found.")
+
+    def _test_metric_existing(self, inference_idx: int, metric: str = "accuracy") -> np.ndarray:
+        """Return test metric for which we already have data."""
+        self.inference_idx = inference_idx
+        inference = self.inferences[self.inference_idx]
+        y = np.zeros((self.training.num_repeats, inference.num_repeats))
+
+        initial_training_repeat_idx = self.training.repeat_idx
+        self.training.repeat_idx = 0
+        for i in range(self.training.num_repeats):
+            inference.repeat_idx = 0
+            for j in range(inference.num_repeats):
+                if metric == "accuracy":
+                    filename = self.accuracy_path()
+                elif metric == "loss":
+                    filename = self.loss_path()
+                elif metric == "avg_power":
+                    filename = self.power_path()
+                val = np.genfromtxt(filename)
+                if metric == "avg_power":
+                    val = np.mean(val)
+                    # Two synaptic layers.
+                    val = 2 * val
+
+                y[i, j] = val
+
+                inference.repeat_idx += 1
+
+            self.training.repeat_idx += 1
+
+        self.training.repeat_idx = initial_training_repeat_idx
+        self.inference_idx = 0
+        return y
+
+    def test_metric(self, metric: str, inference_idx: int = 0) -> np.ndarray:
+        if metric in "error":
+            values = self._test_metric_existing(inference_idx, metric="accuracy")
+        else:
+            values = self._test_metric_existing(inference_idx, metric=metric)
+
+        if metric == "error":
+            values = 1 - values
+
+        return values
+
+    def train(self, use_test_callback: bool = False) -> None:
+        self.is_training = True
+
+        for _ in range(self.training.num_repeats):
+            if os.path.isdir(self.network_dir()):
+                warnings.warn(
+                    f'Training directory "{self.network_dir()}" already exists. Skipping...'
+                )
+                self.training.repeat_idx += 1
+                continue
+            # New callbacks in each iteration because iterator changes.
+            train_callbacks: list[callbacks.Callback] = []
+
+            if use_test_callback:
+                train_callbacks.append(callbacks.TestCallback(self))
+
+            if self.training.use_combined_validation:
+                train_callbacks.append(callbacks.CombinedCheckpoint(self))
+            elif not self.training.is_nonideal():
+                train_callbacks.append(callbacks.StandardCheckpoint(self))
+            else:
+                train_callbacks.append(callbacks.MemristiveCheckpoint(self))
+
+            network.train(self, train_callbacks)
+            self.training.repeat_idx += 1
+
+        self.training.repeat_idx = 0
+
+    def infer(self) -> None:
+        self.is_training = False
+        self.compute_power = True
+        for idx in range(len(self.inferences)):
+            self.inference_idx = idx
+            if os.path.isdir(self.inference_nonideality_dir()):
+                warnings.warn(
+                    f'Inference directory "{self.inference_nonideality_dir()}" already exists. Skipping...'
+                )
+                continue
+            inference = self.inferences[self.inference_idx]
+            for _ in range(self.training.num_repeats):
+                for _ in range(inference.num_repeats):
+                    network.infer(self)
+                    inference.repeat_idx += 1
+
+                inference.repeat_idx = 0
+                self.training.repeat_idx += 1
+
+            self.training.repeat_idx = 0
+        self.inference_idx = 0
+        self.compute_power = False
diff --git a/awarememristor/training/network.py b/awarememristor/training/network.py
new file mode 100644
index 0000000..d8243df
--- /dev/null
+++ b/awarememristor/training/network.py
@@ -0,0 +1,63 @@
+import os
+import pickle
+
+import tensorflow as tf
+
+import awarememristor.training.callbacks as callbacks_
+from awarememristor.training.architecture import get_model
+
+
+def train(iterator, callbacks: list[callbacks_.Callback]) -> None:
+    """Trains using iterator settings and saves information once done."""
+    os.makedirs(iterator.weights_dir(), exist_ok=True)
+
+    validation_data = None
+    num_checkpoint_callbacks = 0
+    for callback in callbacks:
+        if isinstance(callback, callbacks_.Checkpoint):
+            num_checkpoint_callbacks += 1
+            if isinstance(callback, callbacks_.StandardCheckpoint):
+                validation_data = iterator.data("validation")
+
+    if num_checkpoint_callbacks != 1:
+        raise ValueError("One checkpoint callback must be supplied during training!")
+
+    model = get_model(iterator)
+
+    history = model.fit(
+        iterator.data("training"),
+        validation_data=validation_data,
+        verbose=2,
+        epochs=iterator.training.num_epochs,
+        callbacks=callbacks,
+    )
+
+    info = {
+        "history": history.history,
+        "validation_split": iterator.training.validation_split,
+        "batch_size": iterator.training.batch_size,
+        "callback_infos": {},
+    }
+    for callback in callbacks:
+        if isinstance(callback, callbacks_.MemristiveCallback):
+            if callback.name() in info["callback_infos"]:
+                raise KeyError(f'Callback "{callback.name()}" already exists!')
+            info["callback_infos"][callback.name()] = callback.info()
+
+    with open(iterator.info_path(), "wb") as handle:
+        pickle.dump(info, handle)
+
+
+def infer(iterator) -> None:
+    """Performs inference using iterator settings and saves metrics to separate files."""
+    os.makedirs(iterator.inference_repeat_dir(), exist_ok=True)
+
+    model = get_model(iterator)
+
+    score = model.evaluate(iterator.data("testing"), verbose=0)
+
+    print(f"Test loss: {score[0]:.4f}\nTest accuracy: {score[1]:.4f}")
+
+    for var, path in zip(score, [iterator.loss_path(), iterator.accuracy_path()]):
+        with open(path, mode="a", encoding="utf-8"):
+            tf.print(var, output_stream=f"file://{path}")
diff --git a/awarememristor/training/utils.py b/awarememristor/training/utils.py
new file mode 100644
index 0000000..f91c429
--- /dev/null
+++ b/awarememristor/training/utils.py
@@ -0,0 +1,43 @@
+import tensorflow as tf
+
+
+def _compute_device_power(V: tf.Tensor, I_ind: tf.Tensor) -> tf.Tensor:
+    """Compute power dissipated by individual devices in a crossbar.
+
+    Args:
+        V: Voltages in shape `p x m` with `p` examples applied across `m` word lines.
+        I_ind: Currents in shape `p x m x n` generated by the individual devices in crossbar with
+            `m` word lines and `n` bit lines.
+
+    Returns:
+        Power in shape `p x m x n`.
+    """
+    # $P = |V| |I|$ for individual devices. All devices in the same word line
+    # of the crossbar (row of G) are applied with the same voltage.
+    P_ind = tf.einsum("ij,ijk->ijk", tf.math.abs(V), tf.math.abs(I_ind))
+
+    return P_ind
+
+
+def compute_avg_crossbar_power(V: tf.constant, I_ind: tf.constant) -> float:
+    """Compute average power dissipated by a crossbar.
+
+    Args:
+        V: Voltages in shape `p x m` with `p` examples applied across `m` word lines.
+        I_ind: Currents in shape `p x m x n` generated by the individual devices in crossbar with
+            `m` word lines and `n` bit lines.
+
+    Returns:
+        Average power dissipated by a crossbar.
+    """
+    P = _compute_device_power(V, I_ind)
+    P_sum = tf.math.reduce_sum(P)
+    # To get average power consumption **per crossbar** we divide by number of examples.
+    P_avg = P_sum / tf.cast(tf.shape(V)[0], tf.float32)
+
+    return P_avg
+
+
+def normalize_img(image: tf.constant, label: str):
+    """Normalize images: `uint8` -> `float32`."""
+    return tf.cast(image, tf.float32) / 255.0, label
diff --git a/crossbar/__init__.py b/crossbar/__init__.py
deleted file mode 100644
index 2004578..0000000
--- a/crossbar/__init__.py
+++ /dev/null
@@ -1,2 +0,0 @@
-import crossbar.map
-import crossbar.nonlinear_IV
diff --git a/crossbar/map.py b/crossbar/map.py
deleted file mode 100644
index 51ba6ff..0000000
--- a/crossbar/map.py
+++ /dev/null
@@ -1,165 +0,0 @@
-import tensorflow as tf
-
-
-def I_to_y(I, k_V, max_weight, G_max, G_min):
-    """Converts output currents of a dot-product engine onto synaptic layer inputs.
-
-    Parameters
-    ----------
-    I : ndarray
-        Output currents of shape `p x 2n`
-    k_V : float
-        Voltage scaling factor.
-    max_weight : float
-        Assumed maximum weight.
-    G_max : float
-        Maximum conductance of electroformed memristors.
-    G_min : float
-        Minimum conductance of electroformed memristors.
-
-    Returns
-    ----------
-    y : ndarray
-        Outputs of shape `p x n` of a synaptic layer implemented using
-        memristive crossbars.
-    """
-    I_total = I[:, 0::2] - I[:, 1::2]
-    y = I_total_to_y(I_total, k_V, max_weight, G_max, G_min)
-    return y
-
-
-def I_total_to_y(I_total, k_V, max_weight, G_max, G_min):
-    """Converts total output currents of a dot-product engine onto synaptic layer
-    inputs.
-
-    Parameters
-    ----------
-    I_total : ndarray
-        Total output currents of shape `p x n`
-    k_V : float
-        Voltage scaling factor.
-    max_weight : float
-        Assumed maximum weight.
-    G_max : float
-        Maximum conductance of electroformed memristors.
-    G_min : float, optional
-        Minimum conductance of electroformed memristors.
-
-    Returns
-    ----------
-    y : ndarray
-        Outputs of shape `p x n` of a synaptic layer implemented using
-        memristive crossbars.
-    """
-    k_G = compute_k_G(max_weight, G_max, G_min)
-    k_I = compute_k_I(k_V, k_G)
-    y = I_total/k_I
-    return y
-
-
-def clip_weights(weights, max_weight):
-    """Clips weights below 0 and above max_weight.
-
-    Parameters
-    ----------
-    weights : ndarray
-        Synaptic weights.
-    max_weight : float
-        Assumed maximum weight.
-
-    Returns
-    ----------
-    new_weights : ndarray
-        Clipped weights.
-    """
-    weights = tf.clip_by_value(weights, 0.0, max_weight)
-
-    return weights
-
-
-def compute_k_G(max_weight, G_max, G_min):
-    """Computes conductance scaling factor.
-
-    Parameters
-    ----------
-    max_weight : float
-        Assumed maximum weight.
-    G_max : float
-        Maximum conductance of electroformed memristors.
-    G_min : float, optional
-        Minimum conductance of electroformed memristors.
-
-    Returns
-    ----------
-    float
-        Conductance scaling factor.
-    """
-    k_G = (G_max-G_min)/max_weight
-
-    return k_G
-
-
-def compute_k_I(k_V, k_G):
-    """Computes current scaling factor.
-
-    Parameters
-    ----------
-    k_V : float
-        Voltage scaling factor.
-    k_G : float
-        Conductance scaling factor.
-
-    Returns
-    ----------
-    float
-        Current scaling factor.
-    """
-    return k_V*k_G
-
-
-def x_to_V(x, k_V):
-    """Maps inputs (to a synaptic layer) onto voltages.
-
-    Parameters
-    ----------
-    x : ndarray
-        Synaptic inputs.
-    k_V : float
-        Voltage scaling factor.
-
-    Returns
-    ----------
-    ndarray
-        Voltages.
-    """
-    return k_V*x
-
-
-def w_params_to_G(weight_params, max_weight, G_min, G_max):
-    """Maps weight parameters onto conductances.
-
-    Parameters
-    ----------
-    weights_params : ndarray
-        Weight parameters of shape `m x 2n`. These are used to
-        train each conductance (instead of pair of conductances)
-        directly.
-    max_weight : float
-        Assumed maximum weight.
-    G_min : float
-        Minimum conductance of electroformed memristors.
-    G_max : float
-        Maximum conductance of electroformed memristors.
-
-    Returns
-    ----------
-    ndarray
-        Conductances of shape `m x 2n`.
-    """
-    weights_params = clip_weights(weight_params, max_weight)
-
-    k_G = compute_k_G(max_weight, G_max, G_min)
-    G = k_G*weight_params + G_min
-
-    return G
-
diff --git a/crossbar/nonlinear_IV.py b/crossbar/nonlinear_IV.py
deleted file mode 100644
index 4cbbb2f..0000000
--- a/crossbar/nonlinear_IV.py
+++ /dev/null
@@ -1,97 +0,0 @@
-import tensorflow as tf
-
-
-def compute_I(V, G, V_ref, G_min, G_max, n_avg, n_std=tf.constant(0.0)):
-    """Computes output currents of a crossbar consisting of devices suffering
-    from I/V non-linearities.
-
-    Parameters
-    ----------
-    V : ndarray
-        Voltages of shape `p x m`.
-    G : ndarray
-        Conductances of shape `m x n`.
-    V_ref :
-        Reference voltage values of length r (in increasing order) or voltage
-        at which the devices behave Ohmically.
-    G_min : float
-        Minimum conductance of electroformed memristors.
-    G_max : float
-        Maximum conductance of electroformed memristors.
-    n_avg : tf.constant
-        Average value of non-linearity parameter.
-    n_std: tf.constant, optional
-        Standard deviation of non-linearity parameter.
-
-    Returns
-    ----------
-    I : ndarray
-        Output currents of shape `p x n`.
-    I_ind : ndarray
-        Currents of shape `p x m x n` produced by each of the conductances in
-        the crossbar array.
-    """
-    I_ind = compute_currents(G_min, G_max, n_avg, V_ref, G, V, n_std=n_std)
-    I = add_I_BL(I_ind)
-
-    return I, I_ind
-
-
-def compute_currents(G_min, G_max, n_avg, V_ref, G, V, n_std=tf.constant(0.0)):
-    """Compute current values by modelling I-V behaviour using nonlinearity
-    parameter.
-
-    Parameters
-    ----------
-    G_min : float
-        Minimum conductance of electroformed memristors.
-    G_max : float
-        Maximum conductance of electroformed memristors.
-    n_avg : tf.constant
-        Average value of non-linearity parameter.
-    V_ref : float
-        Voltage at which the devices behave Ohmically.
-    G : ndarray
-        Conductances of shape `m x n`.
-    V : ndarray
-        Voltages of shape `p x m`.
-    n_std: tf.constant, optional
-        Standard deviation of non-linearity parameter.
-
-    Returns
-    ----------
-    I : ndarray
-        Currents of shape `p x m x n` produced by each of the conductances in
-        the crossbar array.
-    """
-    epsilon = 1e-4
-
-    exponent = tf.math.log((tf.math.abs(V)+epsilon)/V_ref)/tf.math.log(2.0)
-
-    if n_std == tf.constant(0.0):
-        n = n_avg
-        I = tf.sign(tf.expand_dims(V, axis=-1)) * V_ref * tf.expand_dims(G, axis=0) * n ** (tf.expand_dims(exponent, axis=-1))
-    else:
-        n = tf.random.normal(G.get_shape().as_list(), mean=n_avg, stddev=n_std, dtype=tf.float32)
-        I = tf.sign(tf.expand_dims(V, axis=-1)) * V_ref * tf.expand_dims(G, axis=0) * tf.expand_dims(n, axis=0) ** (tf.expand_dims(exponent, axis=-1))
-
-    return I
-
-
-def add_I_BL(I_ind):
-    """Adds currents along the bit lines.
-
-    Parameters
-    ----------
-    I_ind : ndarray
-        Currents of shape `p x m x n` produced by each of the conductances in
-        the crossbar array.
-
-    Returns
-    ----------
-    I : ndarray
-        Output currents of shape `p x n`.
-    """
-    I = tf.math.reduce_sum(I_ind, axis=1)
-    return I
-
diff --git a/memristor_utils.py b/memristor_utils.py
deleted file mode 100755
index 86f2327..0000000
--- a/memristor_utils.py
+++ /dev/null
@@ -1,199 +0,0 @@
-import numpy as np
-
-import tensorflow as tf
-from tensorflow.keras import backend as K
-from tensorflow.keras.layers import Layer
-
-# Import memristor non-idealities
-import crossbar
-
-
-def disturbed_outputs_i_v_non_linear(x, weights, group_idx=None, log_dir_full_path=None):
-    if group_idx is None:
-        group_idx = 0
-
-    max_weight = tf.math.reduce_max(tf.math.abs(weights))
-    V_ref = tf.constant(0.25)
-
-    G_min_lst = tf.constant([1/983.3, 1/10170, 1/1401000])
-    G_max_lst = tf.constant([1/281.3, 1/2826, 1/385700])
-    n_avg_lst = tf.constant([2.132, 2.596, 2.986])
-    n_std_lst = tf.constant([0.095, 0.088, 0.378])
-
-    G_min = G_min_lst[group_idx]
-    G_max = G_max_lst[group_idx]
-    n_avg= n_avg_lst[group_idx]
-    n_std= n_std_lst[group_idx]
-
-    # Mapping weights onto conductances.
-    G = crossbar.map.w_params_to_G(weights, max_weight, G_min, G_max)
-
-    k_V = 2*V_ref
-
-    # Mapping inputs onto voltages.
-    V = crossbar.map.x_to_V(x, k_V)
-
-    # Computing currents
-    I, I_ind = crossbar.nonlinear_IV.compute_I(
-            V, G, V_ref, G_min, G_max, n_avg, n_std=n_std)
-    if log_dir_full_path is not None:
-        log_file_full_path = "{}/power.csv".format(log_dir_full_path)
-        open(log_file_full_path, "a").close()
-        P_avg = compute_avg_crossbar_power(V, I_ind)
-        tf.print(P_avg, output_stream="file://{}".format(log_file_full_path))
-
-    # Converting to outputs.
-    y_disturbed = crossbar.map.I_to_y(I, k_V, max_weight, G_max, G_min)
-
-    tf.debugging.assert_all_finite(
-        y_disturbed, "nan in outputs", name=None
-    )
-
-    return y_disturbed
-
-
-def compute_device_power(V, I_ind):
-    """Computes power dissipated by individual devices in a crossbar.
-
-    Parameters
-    ----------
-    V : tf.Tensor
-        Voltages in shape (p x m) with p examples applied across m
-        word lines.
-    I_ind : tf.Tensor
-        Currents in shape (p x m x n) generated by the individual
-        devices in crossbar with m word lines and n bit lines.
-
-    Returns
-    ----------
-    P_ind : tf.Tensor
-        Power in shape (p x m x n).
-    """
-    # $P = VI$ for individual devices. All devices in the same word
-    # line of the crossbar (row of G) are applied with the same voltage.
-    P_ind = tf.einsum('ij,ijk->ijk', V, I_ind)
-
-    return P_ind
-
-
-def compute_avg_crossbar_power(V, I_ind):
-    """Computes average power dissipated by a crossbar.
-
-    Parameters
-    ----------
-    V : tf.Tensor
-        Voltages in shape (p x m) with p examples applied across m
-        word lines.
-    I_ind : tf.Tensor
-        Currents in shape (p x m x n) generated by the individual
-        devices in crossbar with m word lines and n bit lines.
-
-    Returns
-    ----------
-    P_avg : tf.Tensor
-        Average power dissipated by a crossbar.
-    """
-    P = compute_device_power(V, I_ind)
-    P_sum = tf.math.reduce_sum(P)
-    # To get average power consumption **per crossbar** we divide by
-    # number of examples.
-    P_avg = P_sum/tf.cast(tf.shape(V)[0], tf.float32)
-
-    return P_avg
-
-
-class memristor_dense(Layer):
-    def __init__(self, n_in, n_out, group_idx=None, is_regularized=True, log_dir_full_path=None, **kwargs):
-        self.n_in=n_in
-        self.n_out=n_out
-        self.group_idx = group_idx
-        self.is_regularized = is_regularized
-        self.log_dir_full_path = log_dir_full_path
-        super(memristor_dense, self).__init__(**kwargs)
-
-    # Adding this funcion removes an issue with custom layer checkpoint
-    def get_config(self):
-
-        config = super().get_config().copy()
-        config.update({
-            'n_in': self.n_in,
-            'n_out': self.n_out,
-        })
-        return config
-
-    # Create trainable weights and biases
-    def build(self, input_shape):
-        stdv=1/np.sqrt(self.n_in)
-        kwargs = {}
-        if self.is_regularized:
-            reg_gamma = 1e-4
-            kwargs["regularizer"] = tf.keras.regularizers.l1(reg_gamma)
-
-        self.w_pos = self.add_weight(
-            shape=(self.n_in,self.n_out),
-            initializer=tf.keras.initializers.RandomNormal(mean=0.5, stddev=stdv),
-            name="weights_pos",
-            trainable=True,
-            **kwargs
-        )
-
-        self.w_neg = self.add_weight(
-            shape=(self.n_in,self.n_out),
-            initializer=tf.keras.initializers.RandomNormal(mean=0.5, stddev=stdv),
-            name="weights_neg",
-            trainable=True,
-            **kwargs
-        )
-
-        self.b_pos = self.add_weight(
-            shape=(self.n_out,),
-            initializer=tf.keras.initializers.Constant(value=0.5),
-            name="biasess_pos",
-            trainable=True,
-            **kwargs
-        )
-
-        self.b_neg = self.add_weight(
-            shape=(self.n_out,),
-            initializer=tf.keras.initializers.Constant(value=0.5),
-            name="biasess_neg",
-            trainable=True,
-            **kwargs
-        )
-
-
-    def call(self, x,mask=None):
-
-        # Clip inputs within 0 and 1
-        #x = tf.clip_by_value(x, 0.0, 1.0)
-
-        # Non-ideality-aware training
-        bias_pos = tf.expand_dims(self.b_pos, axis=0)
-        bias_neg = tf.expand_dims(self.b_neg, axis=0)
-        combined_weights_pos = tf.concat([self.w_pos, bias_pos], 0)
-        combined_weights_neg = tf.concat([self.w_neg, bias_neg], 0)
-        ones = tf.ones([tf.shape(x)[0], 1])
-        inputs = tf.concat([x, ones], 1)
-
-        is_aware = True
-        if is_aware:
-            # Interleave positive and negative weights
-            combined_weights = tf.reshape(
-            tf.concat([combined_weights_pos[...,tf.newaxis], combined_weights_neg[...,tf.newaxis]], axis=-1),
-            [tf.shape(combined_weights_pos)[0],-1])
-
-            self.out = self.apply_output_disturbance(inputs, combined_weights)
-        else:
-            self.out = K.dot(x, self.w) + self.b
-
-        return self.out
-
-    def apply_output_disturbance(self, inputs, weights):
-        disturbed_outputs = disturbed_outputs_i_v_non_linear(inputs, weights, group_idx=self.group_idx, log_dir_full_path=self.log_dir_full_path)
-        return disturbed_outputs
-
-    def get_output_shape_for(self,input_shape):
-        return (input_shape[0], self.n_out)
-    def compute_output_shape(self,input_shape):
-        return (input_shape[0], self.n_out)
-
diff --git a/model_architectures.py b/model_architectures.py
deleted file mode 100644
index 1b1bfe0..0000000
--- a/model_architectures.py
+++ /dev/null
@@ -1,19 +0,0 @@
-from tensorflow.keras.models import Sequential, Model
-from tensorflow.keras.layers import Activation
-from memristor_utils import *
-
-
-batch_norm_eps=1e-4
-batch_norm_momentum=0.9
-
-def get_model(dataset, batch_size, group_idx=None, is_regularized=True, log_dir_full_path=None):
-	if dataset=='MNIST':
-		model=Sequential()
-		model.add(memristor_dense(n_in=784, n_out=25, group_idx=group_idx, is_regularized=is_regularized, log_dir_full_path=log_dir_full_path, input_shape=[784]))
-        # We will try to introduce non-linearities using dense layers.
-		model.add(Activation('sigmoid'))
-		model.add(memristor_dense(n_in=int(model.output.get_shape()[1]),n_out=10, group_idx=group_idx, log_dir_full_path=log_dir_full_path, is_regularized=is_regularized))
-		model.add(Activation('softmax'))
-	else:
-		raise("Dataset {} is not recognised!".format(dataset))
-	return model
diff --git a/pyproject.toml b/pyproject.toml
new file mode 100644
index 0000000..b3f3cec
--- /dev/null
+++ b/pyproject.toml
@@ -0,0 +1,5 @@
+[tool.black]
+line-length = 100
+
+[tool.isort]
+known_first_party = "awarememristor"
diff --git a/reproduce_paper.py b/reproduce_paper.py
new file mode 100644
index 0000000..29699a8
--- /dev/null
+++ b/reproduce_paper.py
@@ -0,0 +1,42 @@
+"""Reproducing simulations and plots in the paper.
+
+Seeds for the simulations were not provided, thus reproduced data might differ
+*slightly*. The qualitative results should remain the same because multiple
+training and inference iterations were used.
+
+SiO_x data will be downloaded automatically. To get access to Ta/HfO2 data,
+please email [Dovydas Joksas](mailto:dovydas.joksas.15@ucl.ac.uk).
+"""
+from awarememristor import simulations
+from awarememristor.plotting import figures, supporting_figures
+
+# Simulations
+simulations.ideal.main()
+simulations.iv_nonlinearity.main()
+simulations.iv_nonlinearity_and_stuck_on.main()
+simulations.iv_nonlinearity_cnn.main()
+simulations.memristive_validation.main()
+simulations.stuck_distribution.main()
+simulations.stuck_off.main()
+simulations.weight_implementation.main()
+simulations.nonideality_agnosticism.main()
+
+# Figures in the main text
+figures.experimental_data()
+figures.iv_nonlinearity_training()
+figures.iv_nonlinearity_inference()
+figures.iv_nonlinearity_cnn()
+figures.weight_implementation()
+figures.memristive_validation()
+figures.nonideality_agnosticism()
+
+# Figures in the Supporting Information
+supporting_figures.all_iv_curves_full_range()
+supporting_figures.switching()
+supporting_figures.iv_nonlinearity_training()
+supporting_figures.weight_implementation_standard_weights_training()
+supporting_figures.weight_implementation_double_weights_training()
+supporting_figures.memristive_validation_training()
+supporting_figures.stuck_off_training()
+supporting_figures.high_iv_nonlinearity_and_stuck_on_training()
+supporting_figures.stuck_distribution_training()
diff --git a/requirements.txt b/requirements.txt
new file mode 100644
index 0000000..ef30583
--- /dev/null
+++ b/requirements.txt
@@ -0,0 +1,12 @@
+h5py==3.1.0
+KDEpy==1.1.0
+matplotlib==3.4.3
+numpy==1.21.3
+pandas==1.3.4
+pytest==6.2.5
+requests==2.26.0
+scipy==1.7.1
+tensorflow==2.7.0
+tensorflow_datasets==4.4.0
+tensorflow_gpu==2.6.0
+tensorflow_probability==0.14.1
diff --git a/tests/__init__.py b/tests/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/crossbar/__init__.py b/tests/crossbar/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/crossbar/test_ideal.py b/tests/crossbar/test_ideal.py
new file mode 100644
index 0000000..ee61d6c
--- /dev/null
+++ b/tests/crossbar/test_ideal.py
@@ -0,0 +1,176 @@
+"""
+Tests of functions of crossbar.nonlinear_IV
+"""
+import pytest
+import tensorflow as tf
+
+from awarememristor.crossbar import ideal
+from tests import utils
+
+# In the ideal case, the bit-line outputs should represent the vector-matrix
+# product of voltages and conductances.
+compute_I_testdata = [
+    (
+        {
+            "G": tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                ]
+            ),
+            "V": tf.constant(
+                [
+                    [1.0, 0.0, -0.5],
+                ]
+            ),
+        },
+        tf.constant(
+            [
+                [
+                    1.0 * 1.0 + 0.0 * 5.0 + (-0.5) * 9.0,
+                    1.0 * 2.0 + 0.0 * 6.0 + (-0.5) * 10.0,
+                    1.0 * 3.0 + 0.0 * 7.0 + (-0.5) * 11.0,
+                    1.0 * 4.0 + 0.0 * 8.0 + (-0.5) * 12.0,
+                ],
+            ]
+        ),
+    ),
+    (
+        {
+            "G": tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                ]
+            ),
+            "V": tf.constant(
+                [
+                    [1.0, 0.0, -0.4],
+                    [0.0, 2.0, 0.0],
+                ]
+            ),
+        },
+        tf.constant(
+            [
+                [
+                    1.0 * 1.0 + 0.0 * 5.0 + (-0.4) * 9.0,
+                    1.0 * 2.0 + 0.0 * 6.0 + (-0.4) * 10.0,
+                    1.0 * 3.0 + 0.0 * 7.0 + (-0.4) * 11.0,
+                    1.0 * 4.0 + 0.0 * 8.0 + (-0.4) * 12.0,
+                ],
+                [
+                    0.0 * 1.0 + 2.0 * 5.0 + 0.0 * 9.0,
+                    0.0 * 2.0 + 2.0 * 6.0 + 0.0 * 10.0,
+                    0.0 * 3.0 + 2.0 * 7.0 + 0.0 * 11.0,
+                    0.0 * 4.0 + 2.0 * 8.0 + 0.0 * 12.0,
+                ],
+            ]
+        ),
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", compute_I_testdata)
+def test_compute_I(args, expected):
+    I = ideal.compute_I(**args)
+    utils.assert_tf_approx(I, expected)
+
+
+compute_I_all_testdata = [
+    (
+        {
+            "G": tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                ]
+            ),
+            "V": tf.constant(
+                [
+                    [0.2, -0.1, 0.0],
+                ]
+            ),
+        },
+        [
+            tf.constant(
+                [
+                    [
+                        0.2 * 1.0 + (-0.1) * 5.0 + 0.0 * 9.0,
+                        0.2 * 2.0 + (-0.1) * 6.0 + 0.0 * 10.0,
+                        0.2 * 3.0 + (-0.1) * 7.0 + 0.0 * 11.0,
+                        0.2 * 4.0 + (-0.1) * 8.0 + 0.0 * 12.0,
+                    ],
+                ]
+            ),
+            tf.constant(
+                [
+                    [
+                        [0.2 * 1.0, 0.2 * 2.0, 0.2 * 3.0, 0.2 * 4.0],
+                        [(-0.1) * 5.0, (-0.1) * 6.0, (-0.1) * 7.0, (-0.1) * 8.0],
+                        [0.0 * 9.0, 0.0 * 10.0, 0.0 * 11.0, 0.0 * 12.0],
+                    ],
+                ]
+            ),
+        ],
+    ),
+    (
+        {
+            "G": tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                ]
+            ),
+            "V": tf.constant(
+                [
+                    [0.2, -0.1, 0.0],
+                    [0.2, -0.1, 0.0],
+                ]
+            ),
+        },
+        [
+            tf.constant(
+                [
+                    [
+                        0.2 * 1.0 + (-0.1) * 5.0 + 0.0 * 9.0,
+                        0.2 * 2.0 + (-0.1) * 6.0 + 0.0 * 10.0,
+                        0.2 * 3.0 + (-0.1) * 7.0 + 0.0 * 11.0,
+                        0.2 * 4.0 + (-0.1) * 8.0 + 0.0 * 12.0,
+                    ],
+                    [
+                        0.2 * 1.0 + (-0.1) * 5.0 + 0.0 * 9.0,
+                        0.2 * 2.0 + (-0.1) * 6.0 + 0.0 * 10.0,
+                        0.2 * 3.0 + (-0.1) * 7.0 + 0.0 * 11.0,
+                        0.2 * 4.0 + (-0.1) * 8.0 + 0.0 * 12.0,
+                    ],
+                ]
+            ),
+            tf.constant(
+                [
+                    [
+                        [0.2 * 1.0, 0.2 * 2.0, 0.2 * 3.0, 0.2 * 4.0],
+                        [(-0.1) * 5.0, (-0.1) * 6.0, (-0.1) * 7.0, (-0.1) * 8.0],
+                        [0.0 * 9.0, 0.0 * 10.0, 0.0 * 11.0, 0.0 * 12.0],
+                    ],
+                    [
+                        [0.2 * 1.0, 0.2 * 2.0, 0.2 * 3.0, 0.2 * 4.0],
+                        [(-0.1) * 5.0, (-0.1) * 6.0, (-0.1) * 7.0, (-0.1) * 8.0],
+                        [0.0 * 9.0, 0.0 * 10.0, 0.0 * 11.0, 0.0 * 12.0],
+                    ],
+                ]
+            ),
+        ],
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", compute_I_all_testdata)
+def test_compute_I_all(args, expected):
+    I_exp, I_ind_exp = expected
+    I, I_ind = ideal.compute_I_all(**args)
+    utils.assert_tf_approx(I, I_exp)
+    utils.assert_tf_approx(I_ind, I_ind_exp)
diff --git a/tests/crossbar/test_map.py b/tests/crossbar/test_map.py
new file mode 100644
index 0000000..14633a7
--- /dev/null
+++ b/tests/crossbar/test_map.py
@@ -0,0 +1,201 @@
+"""
+Tests of functions of crossbar.nonlinear_IV
+"""
+import pytest
+import tensorflow as tf
+
+from awarememristor import crossbar
+from tests import utils
+
+double_w_to_G_testdata = [
+    (
+        {
+            "double_w": tf.constant(
+                [
+                    [3.75, 2.5, 5.0, 2.5],
+                    [2.5, 0.0, 0.0, 1.25],
+                ]
+            ),
+            "G_off": tf.constant(2.0),
+            "G_on": tf.constant(10.0),
+        },
+        [
+            tf.constant(
+                [
+                    [8.0, 6.0, 10.0, 6.0],
+                    [6.0, 2.0, 2.0, 4.0],
+                ]
+            ),
+            tf.constant(5.0),
+        ],
+    ),
+    (
+        {
+            "double_w": tf.constant(
+                [
+                    [8.0, 0.0],
+                    [2.0, 0.0],
+                ]
+            ),
+            "G_off": tf.constant(1.0),
+            "G_on": tf.constant(3.0),
+        },
+        [
+            tf.constant(
+                [
+                    [3.0, 1.0],
+                    [1.5, 1.0],
+                ]
+            ),
+            tf.constant(8.0),
+        ],
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", double_w_to_G_testdata)
+def test_double_w_to_G(args, expected):
+    G_exp, max_weight_exp = expected
+    G, max_weight = crossbar.map.double_w_to_G(**args)
+    utils.assert_tf_approx(G, G_exp)
+    utils.assert_tf_approx(max_weight, max_weight_exp)
+
+
+w_to_G_testdata = [
+    (
+        {
+            "weights": tf.constant(
+                [
+                    [3.75, 2.5, -5.0],
+                    [-2.5, 0.0, 1.25],
+                ]
+            ),
+            "G_off": tf.constant(2.0),
+            "G_on": tf.constant(10.0),
+        },
+        [
+            tf.constant(
+                [
+                    [8.0, 2.0, 6.0, 2.0, 2.0, 10.0],
+                    [2.0, 6.0, 2.0, 2.0, 4.0, 2.0],
+                ]
+            ),
+            tf.constant(5.0),
+        ],
+    ),
+    (
+        {
+            "weights": tf.constant(
+                [
+                    [4.0],
+                    [-2.0],
+                ]
+            ),
+            "G_off": tf.constant(3.0),
+            "G_on": tf.constant(5.0),
+        },
+        [
+            tf.constant(
+                [
+                    [5.0, 3.0],
+                    [3.0, 4.0],
+                ]
+            ),
+            tf.constant(4.0),
+        ],
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", w_to_G_testdata)
+def test_w_to_G(args, expected):
+    G_exp, max_weight_exp = expected
+    G, max_weight = crossbar.map.w_to_G(**args)
+    utils.assert_tf_approx(G, G_exp)
+    utils.assert_tf_approx(max_weight, max_weight_exp)
+
+
+# Test whether ideal crossbars compute vector-matrix products correctly.
+ideal_dpe_testdata = [
+    (
+        {
+            "x": tf.constant(
+                [
+                    [1.0, -1.0],
+                    [2.0, 0.0],
+                    [0.0, -0.5],
+                ]
+            ),
+            "w": tf.constant(
+                [
+                    [1.0, -2.0, 3.0, -4.0],
+                    [5.0, 6.0, -7.0, 8.0],
+                ]
+            ),
+        },
+        tf.constant(
+            [
+                [
+                    1.0 * 1.0 + (-1.0) * 5.0,
+                    1.0 * (-2.0) + (-1.0) * 6.0,
+                    1.0 * 3.0 + (-1.0) * (-7.0),
+                    1.0 * (-4.0) + (-1.0) * 8.0,
+                ],
+                [
+                    2.0 * 1.0 + 0.0 * 5.0,
+                    2.0 * (-2.0) + 0.0 * 6.0,
+                    2.0 * 3.0 + 0.0 * (-7.0),
+                    2.0 * (-4.0) + 0.0 * 8.0,
+                ],
+                [
+                    0.0 * 1.0 + (-0.5) * 5.0,
+                    0.0 * (-2.0) + (-0.5) * 6.0,
+                    0.0 * 3.0 + (-0.5) * (-7.0),
+                    0.0 * (-4.0) + (-0.5) * 8.0,
+                ],
+            ]
+        ),
+    )
+]
+
+
+@pytest.mark.parametrize(
+    "G_off,G_on",
+    [
+        (tf.constant(0.1), tf.constant(0.2)),
+        (tf.constant(0.2), tf.constant(5.0)),
+    ],
+)
+@pytest.mark.parametrize(
+    "V_ref",
+    [
+        tf.constant(0.2),
+        tf.constant(1.0),
+    ],
+)
+@pytest.mark.parametrize(
+    "is_ideal",
+    [
+        True,
+        False,
+    ],
+)
+@pytest.mark.parametrize("args,expected", ideal_dpe_testdata)
+def test_ideal_dpe(args, expected, G_off, G_on, V_ref, is_ideal):
+    x = args["x"]
+    w = args["w"]
+
+    k_V = 2 * V_ref
+    V = crossbar.map.x_to_V(x, k_V)
+
+    G, max_weight = crossbar.map.w_to_G(w, G_off, G_on)
+
+    if is_ideal:
+        I = crossbar.ideal.compute_I(V, G)
+    else:
+        nonideality = crossbar.nonidealities.IVNonlinearity(V_ref, 2.0, 1e-10)
+        I, _ = nonideality.compute_I(V, G)
+
+    y = crossbar.map.I_to_y(I, k_V, max_weight, G_on, G_off)
+
+    utils.assert_tf_approx(y, expected)
diff --git a/tests/crossbar/test_nonidealities.py b/tests/crossbar/test_nonidealities.py
new file mode 100644
index 0000000..f46569f
--- /dev/null
+++ b/tests/crossbar/test_nonidealities.py
@@ -0,0 +1,349 @@
+import pytest
+import tensorflow as tf
+
+from awarememristor.crossbar import nonidealities
+from tests import utils
+
+# Only special case, i.e. when std = 0.0 for all entries.
+d2d_lognormal_testdata = [
+    (
+        (
+            0.5,
+            0.6,
+            0.0,
+            0.0,
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0],
+                    [4.0, 5.0, 6.0],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [1.0, 2.0, 3.0],
+                [4.0, 5.0, 6.0],
+            ]
+        ),
+    ),
+    (
+        (
+            1.0,
+            6.0,
+            0.0,
+            0.0,
+            tf.constant(
+                [
+                    [1.0, 1.0, 1.0],
+                    [1.0, 1.0, 1.0],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [1.0, 1.0, 1.0],
+                [1.0, 1.0, 1.0],
+            ]
+        ),
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", d2d_lognormal_testdata)
+def test_d2d_lognormal(args, expected):
+    G_off, G_on, R_on_log_std, R_off_log_std, G = args
+    nonideality = nonidealities.D2DLognormal(G_off, G_on, R_on_log_std, R_off_log_std)
+    result = nonideality.disturb_G(G)
+    utils.assert_tf_approx(result, expected)
+
+
+# I feel it is appropriate to use multiplication for expected tensors because
+# it is not the underlying operation that we are testing. Writing it out
+# reveals the logic behind the calculations that *should* take place - Dovydas
+iv_nonlinearity_I_ind_testdata = [
+    (
+        (
+            nonidealities.IVNonlinearity(1.0, 2.0, 1e-10),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                ]
+            ),
+            tf.constant(
+                [
+                    [1.0, 0.0],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [
+                    [1.0 * 1.0, 1.0 * 2.0, 1.0 * 3.0, 1.0 * 4.0],
+                    [0.0 * 5.0, 0.0 * 6.0, 0.0 * 7.0, 0.0 * 8.0],
+                ],
+            ]
+        ),
+    ),
+    (
+        (
+            nonidealities.IVNonlinearity(2.0, 2.0, 1e-10),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                ]
+            ),
+            tf.constant(
+                [
+                    [1.0, 0.5],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [
+                    [1.0 * 1.0, 1.0 * 2.0, 1.0 * 3.0, 1.0 * 4.0],
+                    [0.5 * 5.0, 0.5 * 6.0, 0.5 * 7.0, 0.5 * 8.0],
+                ],
+            ]
+        ),
+    ),
+    (
+        (
+            nonidealities.IVNonlinearity(0.5, 4.0, 1e-10),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                ]
+            ),
+            tf.constant(
+                [
+                    [0.0, 0.5, 1.0],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [
+                    [0.0 * 1.0, 0.0 * 2.0, 0.0 * 3.0, 0.0 * 4.0],
+                    # Baseline because V_ref = 0.5
+                    [0.5 * 5.0, 0.5 * 6.0, 0.5 * 7.0, 0.5 * 8.0],
+                    # Multiplying by additional factor of 4 because 1/0.5 = 2
+                    # and n_avg = 4
+                    [0.5 * 9.0 * 4, 0.5 * 10.0 * 4, 0.5 * 11.0 * 4, 0.5 * 12.0 * 4],
+                ],
+            ]
+        ),
+    ),
+    (
+        (
+            nonidealities.IVNonlinearity(0.2, 3.0, 1e-10),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                ]
+            ),
+            tf.constant(
+                [
+                    [0.0, 0.2],
+                    [0.1, 0.4],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [
+                    [0.0 * 1.0, 0.0 * 2.0, 0.0 * 3.0, 0.0 * 4.0],
+                    # Baseline because V_ref = 0.2
+                    [0.2 * 5.0, 0.2 * 6.0, 0.2 * 7.0, 0.2 * 8.0],
+                ],
+                [
+                    # Dividing by additional factor of 3 because 0.1/0.2 =
+                    # 1/2 and n_avg = 3
+                    [
+                        0.2 * 1.0 / 3.0,
+                        0.2 * 2.0 / 3.0,
+                        0.2 * 3.0 / 3.0,
+                        0.2 * 4.0 / 3.0,
+                    ],
+                    # Multiplying by additional factor of 3 because 0.4/0.2
+                    # = 2 and n_avg = 3
+                    [
+                        0.2 * 5.0 * 3.0,
+                        0.2 * 6.0 * 3.0,
+                        0.2 * 7.0 * 3.0,
+                        0.2 * 8.0 * 3.0,
+                    ],
+                ],
+            ]
+        ),
+    ),
+    (
+        (
+            nonidealities.IVNonlinearity(0.5, 5.0, 1e-10),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                    [13.0, 14.0, 15.0, 16.0],
+                ]
+            ),
+            tf.constant(
+                [
+                    [-0.5, -0.25, -1.0, 0.5],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [
+                    # Baseline because V_ref = 0.5
+                    [-0.5 * 1.0, -0.5 * 2.0, -0.5 * 3.0, -0.5 * 4.0],
+                    # Dividing by additional factor of 5 because -0.25/-0.5
+                    # = 1/2 and n_avg = 5
+                    [
+                        -0.5 * 5.0 / 5.0,
+                        -0.5 * 6.0 / 5.0,
+                        -0.5 * 7.0 / 5.0,
+                        -0.5 * 8.0 / 5.0,
+                    ],
+                    # Multiplying by additional factor of 5 because
+                    # -1.0/-0.5 = 1/2 and n_avg = 5
+                    [
+                        -0.5 * 9.0 * 5.0,
+                        -0.5 * 10.0 * 5.0,
+                        -0.5 * 11.0 * 5.0,
+                        -0.5 * 12.0 * 5.0,
+                    ],
+                    # Baseline because V_ref = 0.5
+                    [0.5 * 13.0, 0.5 * 14.0, 0.5 * 15.0, 0.5 * 16.0],
+                ],
+            ]
+        ),
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", iv_nonlinearity_I_ind_testdata)
+def test_iv_nonlinearity_I_ind(args, expected):
+    nonideality, G, V = args
+    _, result = nonideality.compute_I(V, G)
+    utils.assert_tf_approx(result, expected)
+
+
+iv_nonlinearity_I_testdata = [
+    (
+        (
+            nonidealities.IVNonlinearity(5.0, 2.0, 1e-10),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0, 4.0],
+                    [5.0, 6.0, 7.0, 8.0],
+                    [9.0, 10.0, 11.0, 12.0],
+                ]
+            ),
+            tf.constant(
+                [
+                    [1.0, 0.0, -0.5],
+                    [0.0, 0.25, 0.0],
+                ]
+            ),
+        ),
+        [
+            # With {n_avg = 2, n_std = 0} the bit-line outputs should
+            # represent the vector-matrix product of voltages and
+            # conductances.
+            tf.constant(
+                [
+                    [
+                        1.0 * 1.0 + 0.0 * 5.0 + (-0.5) * 9.0,
+                        1.0 * 2.0 + 0.0 * 6.0 + (-0.5) * 10.0,
+                        1.0 * 3.0 + 0.0 * 7.0 + (-0.5) * 11.0,
+                        1.0 * 4.0 + 0.0 * 8.0 + (-0.5) * 12.0,
+                    ],
+                    [
+                        0.0 * 1.0 + 0.25 * 5.0 + 0.0 * 9.0,
+                        0.0 * 2.0 + 0.25 * 6.0 + 0.0 * 10.0,
+                        0.0 * 3.0 + 0.25 * 7.0 + 0.0 * 11.0,
+                        0.0 * 4.0 + 0.25 * 8.0 + 0.0 * 12.0,
+                    ],
+                ]
+            ),
+            # With {n_avg = 2, n_std = 0} currents should be produced
+            # according to Ohm's law.
+            tf.constant(
+                [
+                    [
+                        [1.0 * 1.0, 1.0 * 2.0, 1.0 * 3.0, 1.0 * 4.0],
+                        [0.0 * 5.0, 0.0 * 6.0, 0.0 * 7.0, 0.0 * 8.0],
+                        [-0.5 * 9.0, -0.5 * 10.0, -0.5 * 11.0, -0.5 * 12.0],
+                    ],
+                    [
+                        [0.0 * 1.0, 0.0 * 2.0, 0.0 * 3.0, 0.0 * 4.0],
+                        [0.25 * 5.0, 0.25 * 6.0, 0.25 * 7.0, 0.25 * 8.0],
+                        [0.0 * 9.0, 0.0 * 10.0, 0.0 * 11.0, 0.0 * 12.0],
+                    ],
+                ]
+            ),
+        ],
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", iv_nonlinearity_I_testdata)
+def test_iv_nonlinearity_I(args, expected):
+    I_exp, I_ind_exp = expected
+    nonideality, G, V = args
+    I, I_ind = nonideality.compute_I(V, G)
+    utils.assert_tf_approx(I, I_exp)
+    utils.assert_tf_approx(I_ind, I_ind_exp)
+
+
+stuck_at_testdata = [
+    (
+        (
+            nonidealities.StuckAt(2.0, 1.0),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0],
+                    [4.0, 5.0, 6.0],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [2.0, 2.0, 2.0],
+                [2.0, 2.0, 2.0],
+            ]
+        ),
+    ),
+    (
+        (
+            nonidealities.StuckAt(5.0, 0.0),
+            tf.constant(
+                [
+                    [1.0, 2.0, 3.0],
+                    [4.0, 5.0, 6.0],
+                ]
+            ),
+        ),
+        tf.constant(
+            [
+                [1.0, 2.0, 3.0],
+                [4.0, 5.0, 6.0],
+            ]
+        ),
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", stuck_at_testdata)
+def test_stuck_at(args, expected):
+    nonideality, G = args
+    result = nonideality.disturb_G(G)
+    utils.assert_tf_approx(result, expected)
diff --git a/tests/crossbar/test_utils.py b/tests/crossbar/test_utils.py
new file mode 100644
index 0000000..e3d35ed
--- /dev/null
+++ b/tests/crossbar/test_utils.py
@@ -0,0 +1,36 @@
+"""
+Tests of functions of crossbar.faulty_devices
+"""
+import pytest
+import tensorflow as tf
+
+from awarememristor.crossbar import utils as crossbar_utils
+from tests import utils
+
+random_bool_tensor_testdata = [
+    (
+        {
+            "shape": [2, 3],
+            "prob_true": 0.0,
+        },
+        tf.constant(
+            [
+                [False, False, False],
+                [False, False, False],
+            ]
+        ),
+    ),
+    (
+        {
+            "shape": [4],
+            "prob_true": 1.0,
+        },
+        tf.constant([True, True, True, True]),
+    ),
+]
+
+
+@pytest.mark.parametrize("args,expected", random_bool_tensor_testdata)
+def test_random_bool_tensor(args, expected):
+    result = crossbar_utils.random_bool_tensor(**args)
+    utils.assert_tf_bool_equal(result, expected)
diff --git a/tests/training/__init__.py b/tests/training/__init__.py
new file mode 100644
index 0000000..e69de29
diff --git a/tests/training/test_iterator.py b/tests/training/test_iterator.py
new file mode 100644
index 0000000..be5da60
--- /dev/null
+++ b/tests/training/test_iterator.py
@@ -0,0 +1,69 @@
+"""
+Tests of functions of training.iterator
+"""
+# pylint: disable=missing-function-docstring
+import pytest
+
+from awarememristor.crossbar import nonidealities
+from awarememristor.training import iterator
+
+nonideality_label_testdata = [
+    (
+        iterator.Stage(),
+        "ideal",
+    ),
+    (
+        iterator.Stage(
+            nonidealities=[nonidealities.IVNonlinearity(0.25, 1.53, 0.625)],
+        ),
+        "IVNL:1.53_0.625",
+    ),
+    (
+        iterator.Stage(
+            nonidealities=[nonidealities.StuckAt(1.20, 0.6341)],
+        ),
+        "Stuck:1.2_0.634",
+    ),
+    (
+        iterator.Stage(
+            nonidealities=[
+                nonidealities.IVNonlinearity(0.25, 1.530, 0.123),
+                nonidealities.StuckAt(1.2344, 0.06341),
+            ]
+        ),
+        "IVNL:1.53_0.123+Stuck:1.23_0.0634",
+    ),
+]
+
+
+@pytest.mark.parametrize("nonideal_instance,expected", nonideality_label_testdata)
+def test_nonideality_label(nonideal_instance, expected):
+    result = nonideal_instance.nonideality_label()
+    assert result == expected
+
+
+nonidealities_exception_testdata = [
+    (
+        [
+            nonidealities.IVNonlinearity(0.25, 3.1, 0.1203),
+            nonidealities.StuckAt(1.23, 0.0009),
+            nonidealities.StuckAt(4.5, 0.1),
+        ],
+        "Current implementation does not support more than one linearity-preserving nonideality.",
+    ),
+    (
+        [
+            nonidealities.IVNonlinearity(0.25, 3.1, 0.1203),
+            nonidealities.IVNonlinearity(0.25, 2.1, 0.1),
+        ],
+        "Current implementation does not support more than one linearity-nonpreserving nonideality.",
+    ),
+]
+
+
+@pytest.mark.parametrize("nonidealities_input,error_msg", nonidealities_exception_testdata)
+def test_nonidealities_exception(nonidealities_input, error_msg):
+    with pytest.raises(Exception) as exc:
+        _ = iterator.Stage(nonidealities=nonidealities_input)
+    assert error_msg in str(exc.value)
+    assert exc.type == ValueError
diff --git a/tests/training/test_utils.py b/tests/training/test_utils.py
new file mode 100644
index 0000000..506ccb0
--- /dev/null
+++ b/tests/training/test_utils.py
@@ -0,0 +1,110 @@
+import pytest
+import tensorflow as tf
+
+from awarememristor.training import utils
+from tests import utils as test_utils
+
+compute_avg_crossbar_power_testdata = [
+    (
+        tf.constant(
+            [
+                [
+                    1.0,
+                    0.0,
+                    2.0,
+                ],
+            ]
+        ),
+        tf.constant(
+            [
+                [
+                    [0.0, 1.0],
+                    [2.0, 3.0],
+                    [4.0, 5.0],
+                ],
+            ]
+        ),
+        19.0,
+    ),
+    (
+        tf.constant(
+            [
+                [
+                    4.0,
+                    1.0,
+                ],
+                [
+                    0.0,
+                    1.0,
+                ],
+            ]
+        ),
+        tf.constant(
+            [
+                [
+                    [0.0, 1.0],
+                    [2.0, 3.0],
+                ],
+                [
+                    [2.0, 4.0],
+                    [0.0, 2.0],
+                ],
+            ],
+        ),
+        5.5,
+    ),
+    (
+        tf.constant(
+            [
+                [
+                    -1.0,
+                    0.0,
+                    2.0,
+                ],
+            ]
+        ),
+        tf.constant(
+            [
+                [
+                    [0.0, 1.0],
+                    [2.0, 3.0],
+                    [-4.0, 5.0],
+                ],
+            ]
+        ),
+        19.0,
+    ),
+    (
+        tf.constant(
+            [
+                [
+                    -4.0,
+                    1.0,
+                ],
+                [
+                    0.0,
+                    -1.0,
+                ],
+            ]
+        ),
+        tf.constant(
+            [
+                [
+                    [0.0, -1.0],
+                    [2.0, -3.0],
+                ],
+                [
+                    [2.0, 4.0],
+                    [0.0, -2.0],
+                ],
+            ],
+        ),
+        5.5,
+    ),
+]
+
+
+@pytest.mark.parametrize("V,I_ind,expected", compute_avg_crossbar_power_testdata)
+def test_compute_avg_crossbar_power(V, I_ind, expected):
+    result = utils.compute_avg_crossbar_power(V, I_ind)
+    assert pytest.approx(result) == expected
diff --git a/tests/utils.py b/tests/utils.py
new file mode 100644
index 0000000..fa462b8
--- /dev/null
+++ b/tests/utils.py
@@ -0,0 +1,12 @@
+import numpy as np
+import tensorflow as tf
+
+
+def assert_tf_approx(a, b):
+    tf.debugging.assert_near(a, b, rtol=1.0e-6, atol=1.0e-6)
+    assert a.shape == b.shape
+
+
+def assert_tf_bool_equal(a, b):
+    # Don't know how to compare boolean arrays using TF, so using numpy.
+    np.testing.assert_array_equal(a, b)