snorkelflow.client.fm_suite

Warm Start related functions. Provides functions for estimating costs, viewing available methods, running jobs, and monitoring progress. Allows configuring and kicking off warm start workflows to generate labeling functions using foundation models.

Functions

create_prompt_lf(node, model_name, ...[, ...])	Kick off a prompt inference job on a given node.
get_available_warm_start_methods(node)	Get the available warm start methods for a given node.
get_warm_start_cost(node, configs[, split, ...])	Estimate the additional cost in USD of warm start methods for a given node.
preview_prompt_lf(node, model_name, ...[, ...])	Preview a prompt inference job on a single data split of a given node.
run_lf_inference(node, lf_uid[, ...])	Run the LF on new data.
run_warm_start(node, warm_start_method, ...)	Kick off an FM Warm Start job on a given node.