2.3 KiB
| title |
|---|
| TensorFlow MNIST |
The tensorflow_mnist.py example demonstrates the integration of ClearML into code that uses TensorFlow and Keras to train a neural network on the Keras built-in MNIST handwritten digits dataset.
The script builds a TensorFlow Keras model, and trains and tests it with the following:
- Loss objective function - tf.keras.metrics.SparseCategoricalCrossentropy
- Accuracy metric - tf.keras.metrics.SparseCategoricalAccuracy
- Model checkpointing - tf.clearml.Checkpoint and tf.train.CheckpointManager
When the script runs, it creates an experiment named Tensorflow v2 mnist with summaries, which is associated with the
examples project.
Scalars
The loss and accuracy metric scalar plots appear in the experiment's page in the ClearML web UI under RESULTS > SCALARS. Resource utilization plots, which are titled :monitor: machine, also appear in the SCALARS tab.
Hyperparameters
ClearML automatically logs TensorFlow Definitions. They appear in CONFIGURATIONS > HYPER PARAMETERS > TF_DEFINE.
Console
All console output appears in RESULTS > CONSOLE.
Artifacts
Models created by the experiment appear in the experiment’s ARTIFACTS tab. ClearML automatically logs and tracks models and any snapshots created using TensorFlow.
Clicking on a model’s name takes you to the model’s page, where you can view the model’s details and access the model.
