TensorFlow Model Analysis Integration

If you're already using TensorFlow and TensorFlow Model Analysis (TFMA), then Comet works out of the box. Just add these lines of code to your training script:

```python import os os.environ["COMET_AUTO_LOG_TFMA"] = "1"

from comet_ml import Experiment import tensorflow_model_analysis as tfma

experiment = Experiment()

Your TFMA code:

results = tfma.load_eval_result("/path/to/results")

tfma.view.render_slicing_metrics(results) tfma.view.render_slicing_metrics(results, 'a_column') tfma.view.render_plot(results) tfma.view.render_time_series(results) ```

That's it for logging the results!

To see them in the Comet.ml UI, you can use the custom TFMA Viewer Panel at either the Project View, or the Experiment View.

TensorFlow Model Analysis

Summary

To use TFMA with Comet.ml:

  1. Set the Comet.ml Configuration Variable "COMET_AUTO_LOG_TFMA" to "1"
  2. Create a Comet.ml Experiment()
  3. Render your TFMA results as normal
  4. See results in your experiment asset, and the custom "TFMA Viewer" Panel

References

  1. TensorFlow
  2. TensorFlow Model Analysis
  3. TFMA Asset Panel
  4. Setting Comet.ml Configuration Variables