[DEVX-219]: Model Training

[sanjaychelliah]

What

• Integrating Model Training in SDK
• Adding support functions for training

Features

Model Training

# Note: CLARIFAI_PAT must be set as env variable.
from clarifai.client.app import App
from clarifai.client.model import Model

"""
Create model with trainable model_type
"""
app = App(user_id="user_id", app_id="app_id")
model = app.create_model(model_id="model_id", model_type_id="visual-classifier")
               (or)
model = Model('url')

"""
List training templates for the model_type
"""
templates = model.list_training_templates()
print(templates)

"""
Get parameters for the model.
"""
params = model.get_params(template='classification_basemodel_v1', yaml_file='model_params.yaml'
)

"""
Update the model params yaml and pass it to model.train()
"""
model_version_id = model.train('model_params.yaml')

"""
Training status and saving logs
"""
status = model.training_status(version_id=model_version_id,training_logs=True)
print(status)

Model Training support functionalities

# Note: CLARIFAI_PAT must be set as env variable.
from clarifai.client.app import App
from clarifai.client.model import Model

"""
show trainable model types
"""
app = App(user_id="user_id", app_id="app_id")
print(app.list_trainable_model_types())

"""
Show individual param information
"""
params = model.get_params(template='classification_basemodel_v1', yaml_file='model_params.yaml')
print(model.get_param_info(param = 'dataset_id))

"""
Update specific params for the model
"""
model.update_params({'template': 'xxx'})
model.train()

Dataset functionalities added

• dataset.create_version()
• dataset.list_versions()
• dataset.delete_version()

TODO

• Pytest tests in the upcoming PR
• More examples in examples repo

[ackizilkale]

looks good to me.

[sainivedh]

Great Work. Added some comments

[sainivedh]

Some comments could be helpful to maintain in future

[sanjaychelliah]

Sure. We can add comments as parsing the API response Multimodeltypes. But every loop detail, not sure, its just indexing. Thats why kept the iter variable names clearly.

[sainivedh]

Suggested change

	def response_to_templates(response: MultiModelTypeResponse, model_type_id: str) -> List[str]:
	def _response_to_templates(response: MultiModelTypeResponse, model_type_id: str) -> List[str]:

Shall we mark all these function as internal use ? wdyt

[sanjaychelliah]

The functions are stored in internal utils and we are following the same structure in Workflows. So I think this is fine.

[sainivedh]

lgtm!