modlee.model.callbacks module
- class modlee.model.callbacks.DataMetafeaturesCallback(data_snapshot_size=10000000.0, DataMetafeatures=<class 'modlee.data_metafeatures.DataMetafeatures'>, *args, **kwargs)[source]
Bases:
ModleeCallback
Callback to calculate and log data meta-features.
- class modlee.model.callbacks.LogCodeTextCallback(kwargs_to_cache={}, *args, **kwargs)[source]
Bases:
ModleeCallback
Callback to log the model as code and text.
- class modlee.model.callbacks.LogModalityTaskCallback[source]
Bases:
ModleeCallback
Logs the modality and task
- class modlee.model.callbacks.LogModelCheckpointCallback(monitor='val_loss', filename='model_checkpoint', temp_dir_path='/Users/brad/Github_Modlee/modlee_pypi/src/modlee/tmp/checkpoints', save_top_k=1, mode='min', verbose=True, *args, **kwargs)[source]
Bases:
ModelCheckpoint
Callback to log the best performing model in a training routine based on Loss value
- class modlee.model.callbacks.LogONNXCallback[source]
Bases:
ModleeCallback
Callback for logging the model in its ONNX representations. Deprecated, will be combined with LogCodeTextCallback.
- class modlee.model.callbacks.LogOutputCallback(*args, **kwargs)[source]
Bases:
Callback
Callback to log the output metrics for each batch.
- class modlee.model.callbacks.LogParamsCallback[source]
Bases:
Callback
Callback to log parameters at the start of training.
- class modlee.model.callbacks.LogTransformsCallback[source]
Bases:
ModleeCallback
Logs transforms applied to the dataset, if applied with torchvision.transforms
- class modlee.model.callbacks.ModelMetafeaturesCallback(ModelMetafeatures=<class 'modlee.model_metafeatures.ModelMetafeatures'>)[source]
Bases:
ModleeCallback