How can I use weight of params from a previous iteration of training in Azure custom vision

Question

iteration = trainer.train_project(
    project_id,
    training_type="Advanced", # Possible values include: 'Regular', 'Advanced'
    reserved_budget_in_hours=8, 
    force_train=False, 
    notification_email_address="test@testcompany.com",
    selected_tags=trainable_tags, # list | [str]
    custom_base_model_info=None, # Information of the previously trained iteration
    custom_headers=None, 
    raw=False
)

I want to ask how to use custom_base_model_info from the previous iteration. (in python)

Answer 1

@Tinnapop Try adding the import statement and passing the project_id and iteration_id along with the train_project()

from azure.cognitiveservices.vision.customvision.training.models import CustomBaseModelInfo

iteration = trainer.train_project(
    project_id,
    training_type="Advanced", # Possible values include: 'Regular', 'Advanced'
    reserved_budget_in_hours=8, 
    force_train=False, 
    notification_email_address="test@testcompany.com",
    selected_tags=trainable_tags, # list | [str]
    custom_base_model_info=CustomBaseModelInfo(project_id=<your_project_id, iteration_id=<your_iteration_id>>), # Information of the previously trained iteration
    custom_headers=None, 
    raw=False
)

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.

Reference for CustomBaseModelInfo

If this answers your query, do click Accept Answer and Yes for was this answer helpful. And, if you have any further query do let us know.