python3.6 Successful Attempt

[kkkmax]

No description provided.

[kkkmax]

thanks@andabi project

• Because I wanted to use Python 3 to reproduce the project, I meet many problems. I wrote about the problems I encountered in the process and how I solved them. I think it can help other friends, that's all.

• Installing CONDA virtual environment
  $ conda create –n voice python=3.6
  $ source activate voice3

• Installing cuda + cudnn
  conda install cudatoolkit=8.0 -c https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/linux-64/

At this point, he will automatically install the cudnn version, and tensorflow 1.4 will work well in the cuda8 + cudnn6 environment.

• Installation of required bags
  tensorflow==1.4
  tensorflow-gpu==1.4
  librosa==0.6.2
  pyyaml
  tensorflow-plot==0.3.0
  tensorpack 0.9.0

I'm just a couple of trains that I've run successfully in the right environment.

• GET RIGHT data_path
  You can choose any of the following to do it
  1）change data_path in default.yaml data_path and change data_load.py
  wav_file.replace("WAV.wav", "PHN").replace("WAV", "PHN")
  2）change data_path in default.yaml data_path and change timit dataset .WAV file -->.wav file
  For example：
  wav_file = '/datasets/timit/TIMIT/TRAIN/DR3/MTJM0/SA1.WAV'
  filename, extension = os.path.splitext(wav_file)
  new_wav_file= filename+extension.lower()
  you can get: '/datasets/timit/TIMIT/TRAIN/DR3/MTJM0/SA1.wav'

• Increase the number of threads or batch_size you can accelerate training
  Of course, it depends on your hardware.

[Lukelluke]

For later visitors guidence: detailed .yaml description in py36:

name: deepvoice
channels:
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/main/
  - https://mirrors.tuna.tsinghua.edu.cn/anaconda/pkgs/free/
  - defaults
dependencies:
  - _libgcc_mutex=0.1=main
  - backports=1.0=pyhd3eb1b0_2
  - backports.weakref=1.0rc1=py36_0
  - blas=1.0=mkl
  - bleach=1.5.0=py36_0
  - ca-certificates=2020.10.14=0
  - certifi=2020.12.5=py36h06a4308_0
  - cudatoolkit=8.0=3
  - cudnn=6.0.21=cuda8.0_0
  - html5lib=0.9999999=py36_0
  - importlib-metadata=2.0.0=py_1
  - intel-openmp=2020.2=254
  - ld_impl_linux-64=2.33.1=h53a641e_7
  - libedit=3.1.20191231=h14c3975_1
  - libffi=3.3=he6710b0_2
  - libgcc=7.2.0=h69d50b8_2
  - libgcc-ng=9.1.0=hdf63c60_0
  - libprotobuf=3.13.0.1=hd408876_0
  - libstdcxx-ng=9.1.0=hdf63c60_0
  - markdown=3.3.3=py36h06a4308_0
  - mkl=2020.2=256
  - mkl-service=2.3.0=py36he8ac12f_0
  - mkl_fft=1.2.0=py36h23d657b_0
  - mkl_random=1.1.1=py36h0573a6f_0
  - ncurses=6.2=he6710b0_1
  - numpy=1.19.2=py36h54aff64_0
  - numpy-base=1.19.2=py36hfa32c7d_0
  - openssl=1.1.1h=h7b6447c_0
  - pip=20.3.1=py36h06a4308_0
  - protobuf=3.13.0.1=py36he6710b0_1
  - python=3.6.12=hcff3b4d_2
  - readline=8.0=h7b6447c_0
  - setuptools=51.0.0=py36h06a4308_2
  - six=1.15.0=py36h06a4308_0
  - sqlite=3.33.0=h62c20be_0
  - tensorflow-gpu=1.3.0=0
  - tensorflow-gpu-base=1.3.0=py36cuda8.0cudnn6.0_1
  - tensorflow-tensorboard=1.5.1=py36hf484d3e_1
  - tk=8.6.10=hbc83047_0
  - werkzeug=1.0.1=py_0
  - wheel=0.36.1=pyhd3eb1b0_0
  - xz=5.2.5=h7b6447c_0
  - zipp=3.4.0=pyhd3eb1b0_0
  - zlib=1.2.11=h7b6447c_3
  - pip:
    - audioread==2.1.9
    - biwrap==0.1.6
    - cffi==1.14.4
    - cycler==0.10.0
    - decorator==4.4.2
    - joblib==0.17.0
    - kiwisolver==1.3.1
    - librosa==0.6.2
    - llvmlite==0.31.0
    - matplotlib==3.3.3
    - msgpack==1.0.0
    - msgpack-numpy==0.4.7.1
    - numba==0.48.0
    - pillow==8.0.1
    - pycparser==2.20
    - pydub==0.24.1
    - pyparsing==2.4.7
    - python-dateutil==2.8.1
    - pyyaml==5.3.1
    - pyzmq==20.0.0
    - resampy==0.2.2
    - scikit-learn==0.23.2
    - scipy==1.5.4
    - soundfile==0.10.3.post1
    - tabulate==0.8.7
    - tensorboard==1.7.0
    - tensorflow-plot==0.3.0
    - tensorpack==0.9.0
    - termcolor==1.1.0
    - threadpoolctl==2.1.0
    - tqdm==4.54.1
prefix: /home/hsj/miniconda3/envs/deepvoice