Implementation:Googleapis Python genai Tunings Tune

Knowledge Sources	googleapis/python-genai Google Gen AI Python SDK
Domains	Fine_Tuning, Model_Training
Last Updated	2026-02-15 00:00 GMT

Overview

Concrete tool for launching model fine-tuning jobs provided by the google-genai tunings module.

Description

Tunings.tune submits a fine-tuning job to the Google AI service. It accepts a base model ID, a TuningDataset, and optional CreateTuningJobConfig. The method validates inputs, sends the request via the appropriate backend (Gemini Developer API or Vertex AI), and returns a TuningJob object with the job's resource name, state, and metadata. The tuning job runs asynchronously server-side.

Usage

Call client.tunings.tune with the base model, training dataset, and optional config. Store the returned TuningJob's name for monitoring. The job runs in the background; use tunings.get to check progress.

Code Reference

Source Location

Repository: googleapis/python-genai
File: google/genai/tunings.py
Lines: L1725

Signature

class Tunings:
    def tune(
        self,
        *,
        base_model: str,
        training_dataset: types.TuningDatasetOrDict,
        config: Optional[types.CreateTuningJobConfigOrDict] = None,
    ) -> types.TuningJob:
        """Launches a fine-tuning job.

        Args:
            base_model: Base model ID to fine-tune (e.g., 'gemini-1.5-flash-002').
            training_dataset: Training data (TuningDataset or dict).
            config: Optional tuning job configuration.
        """

Import

from google import genai

I/O Contract

Inputs

Name	Type	Required	Description
base_model	str	Yes	Base model to fine-tune (e.g., 'gemini-1.5-flash-002')
training_dataset	TuningDatasetOrDict	Yes	Training data from TuningDataset
config	Optional[CreateTuningJobConfigOrDict]	No	Tuning job configuration

Outputs

Name	Type	Description
TuningJob	types.TuningJob	Job object with .name, .state, .tuned_model, .create_time

Usage Examples

Launch a Tuning Job

from google import genai
from google.genai import types

client = genai.Client(
    vertexai=True,
    project="my-project",
    location="us-central1"
)

tuning_job = client.tunings.tune(
    base_model="gemini-1.5-flash-002",
    training_dataset=types.TuningDataset(
        gcs_uri="gs://my-bucket/training_data.jsonl"
    ),
    config=types.CreateTuningJobConfig(
        epoch_count=5,
        tuned_model_display_name="my-classifier",
    ),
)

print(f"Job name: {tuning_job.name}")
print(f"State: {tuning_job.state}")

Launch with Inline Examples

tuning_job = client.tunings.tune(
    base_model="gemini-1.5-flash-002",
    training_dataset=types.TuningDataset(
        examples=[
            types.TuningExample(text_input="Hello", output="Hi there!"),
            types.TuningExample(text_input="Bye", output="Goodbye!"),
        ]
    ),
)

Related Pages

Page Connections

Double-click a node to navigate. Hold to expand connections.

Principle

Implementation

Heuristic

Environment