TensorFlow Runtime Tracer

This project is a web application to monitor and trace TensorFlow scripts in the runtime on the op level.

It starts a web server upon the execution of the script. The web interface keeps track of all the session runs and can trace the execution on demand.

The goal of this tool is to facilitate the process of performance tuning with minimal code changes and insignificant runtime overhead. Both Higher-level (tf.estimator.Estimator) and Low-level (tf.train.MonitoredTrainingSession and co) APIs are supported. It also supports horovod and IBM Distributed Deep Learning (DDL). The tracing session can be saved, reloaded, and distributed effortlessly.

Some screenshots here.

Installation

Use pip to install:

pip install tensorflow-tracer

Quick Start

1. Install tensorflow-tracer and run an example:

   $ pip3 install tensorflow-tracer
   $ git clone https://github.com/xldrx/tensorflow-tracer.git
   $ python3 ./tensorflow-tracer/examples/estimator-example.py 

2. Browse to: http://0.0.0.0:9999

How to Use

1. Add tftracer to your code:

   Estimator API:

   from tftracer import TracingServer
   ...
   
   tracing_server = TracingServer()
   estimator.train(input_fn, hooks=[tracing_server.hook]) 

   Low-Level API:

   from tftracer import TracingServer
   ...
   tracing_server = TracingServer()
   with tf.train.MonitoredTrainingSession(hooks=[tracing_server.hook]):
      ...

   [More examples here]

2. Run your code and browse to:

http://0.0.0.0:9999

How to Trace an Existing Code

If you want to trace an existing script without any modification use tftracer.hook_inject Please note that this is experimental and may cause unexpected errors:

1. Add the following to the beggining of the main script: .. code-block:: python

   import tftracer
   tftracer.hook_inject()
   ...
   

2. Run your code and browse to http://0.0.0.0:9999

Command line

Tracing sessions can be stored either through the web interface or by calling tracing_server.save_session(filename).

To reload a session, run this in the terminal:

tftracer filename

Then browse to:

http://0.0.0.0:9999

API

Full Documentation is here.

Known Bugs/Limitations

• Only Python3 is supported.
• The web interface loads javascript/css libraries remotely (e.g. vue.js, ui-kit, jquery, jquery-ui, Google Roboto, awesome-icons, ... ). Therefore an active internet connection is needed to properly render the interface. The tracing server does not require any remote connection.
• All traces are kept in the memory while tracing server is running.
• Tracing uses tf.train.SessionRunHook and is unable to trace auxiliary runs such as init_op.
• The tracing capability is limited to what tf.RunMetadata offers. For example, CUPTI events are missing when tracing a distributed job.
• HTTPS is not supported.

Frequently Asked Questions

How to trace/visualize just one session run?

Use tftracer.Timeline. for example:

    from tftracer import Timeline
    ...
    with tf.train.MonitoredTrainingSession() as sess:
       with Timeline() as tl:
        sess.run(fetches, **tl.kwargs)
    ...
    tl.visualize(filename)

Comparision to TensorBoard?

The nature of this project is a short-lived light-weight interactive tracing interface to monitor and trace execution on the op-level. In comparison TensorBoard is a full-featured tool to inspect the application on many levels:

• tftracer does not make any assumption about the dataflow DAG. There is no need to add any additional op to the data flow dag (i.e. tf.summary) or having a global step.

• tftracer runs as a thread and lives from the start of the execution and lasts until the end of it. TensorBoard runs as a separate process and can outlive the main script.

Cite this tool

@misc{hashemi-tftracer-2018,
  author = {Sayed Hadi Hashemi},
  title = {TensorFlow Runtime Tracer},
  year = {2018},
  publisher = {GitHub},
  journal = {GitHub repository},
  howpublished = {\url{https://github.com/xldrx/tensorflow-tracer}},
}