Trainer
Used to perform training and inference. Every time you execute fit, predict, or fit_predict using a Trainer object, a job is created and queued for execution. Only one job can be executed at a time. You can use JobManager.show_jobs() to check the status of all jobs.
Parameters
Section titled “Parameters”| Name | Type | Description | Optional |
|---|---|---|---|
connector | SnowflakeConnector | The connector object used for sending requests to the GNN engine. | No |
config | TrainerConfig | Configuration object containing all training hyperparameters. | No |
Returns
Section titled “Returns”An instance of the Trainer class.
Example
Section titled “Example”from relationalai_gnns import Trainer
model_config = TrainerConfig(connector=connector, device="cuda", n_epochs=5, max_iters=2000, text_embedder="model2vec-potion-base-4M")
trainer = Trainer(connector=connector, config=config)