feat: LLM - Added tuning support for chat-bison models

PiperOrigin-RevId: 555782339

Author: Ark-kun

tests/unit/aiplatform/test_language_models.py

@@ -57,6 +57,7 @@
     language_models as preview_language_models,
 )
 from vertexai import language_models
+from vertexai.language_models import _language_models
 from google.cloud.aiplatform_v1 import Execution as GapicExecution
 from google.cloud.aiplatform.compat.types import (
     encryption_spec as gca_encryption_spec,
@@ -471,7 +472,7 @@ def get_endpoint_mock():
 @pytest.fixture
 def mock_get_tuned_model(get_endpoint_mock):
     with mock.patch.object(
-        preview_language_models.TextGenerationModel, "get_tuned_model"
+        _language_models._TunableModelMixin, "get_tuned_model"
     ) as mock_text_generation_model:
         mock_text_generation_model._model_id = (
             test_constants.ModelConstants._TEST_MODEL_RESOURCE_NAME
@@ -634,7 +635,7 @@ def test_text_generation_ga(self):
         ["https://us-central1-kfp.pkg.dev/proj/repo/pack/latest"],
         indirect=True,
     )
-    def test_tune_model(
+    def test_tune_text_generation_model(
         self,
         mock_pipeline_service_create,
         mock_pipeline_job_get,
@@ -680,6 +681,49 @@ def test_tune_model(
                 == _TEST_ENCRYPTION_KEY_NAME
             )
 
+    @pytest.mark.parametrize(
+        "job_spec",
+        [_TEST_PIPELINE_SPEC_JSON],
+    )
+    @pytest.mark.parametrize(
+        "mock_request_urlopen",
+        ["https://us-central1-kfp.pkg.dev/proj/repo/pack/latest"],
+        indirect=True,
+    )
+    def test_tune_chat_model(
+        self,
+        mock_pipeline_service_create,
+        mock_pipeline_job_get,
+        mock_pipeline_bucket_exists,
+        job_spec,
+        mock_load_yaml_and_json,
+        mock_gcs_from_string,
+        mock_gcs_upload,
+        mock_request_urlopen,
+        mock_get_tuned_model,
+    ):
+        """Tests tuning a chat model."""
+        aiplatform.init(project=_TEST_PROJECT, location=_TEST_LOCATION)
+        with mock.patch.object(
+            target=model_garden_service_client.ModelGardenServiceClient,
+            attribute="get_publisher_model",
+            return_value=gca_publisher_model.PublisherModel(
+                _CHAT_BISON_PUBLISHER_MODEL_DICT
+            ),
+        ):
+            model = preview_language_models.ChatModel.from_pretrained("chat-bison@001")
+
+            model.tune_model(
+                training_data=_TEST_TEXT_BISON_TRAINING_DF,
+                tuning_job_location="europe-west4",
+                tuned_model_location="us-central1",
+            )
+            call_kwargs = mock_pipeline_service_create.call_args[1]
+            pipeline_arguments = call_kwargs[
+                "pipeline_job"
+            ].runtime_config.parameter_values
+            assert pipeline_arguments["large_model_reference"] == "chat-bison@001"
+
     @pytest.mark.usefixtures(
         "get_model_with_tuned_version_label_mock",
         "get_endpoint_with_models_mock",

vertexai/_model_garden/_model_garden_models.py

@@ -33,6 +33,7 @@
 _SHORT_MODEL_ID_TO_TUNING_PIPELINE_MAP = {
     "text-bison": "https://us-kfp.pkg.dev/ml-pipeline/large-language-model-pipelines/tune-large-model/v2.0.0",
     "code-bison": "https://us-kfp.pkg.dev/ml-pipeline/large-language-model-pipelines/tune-large-model/v3.0.0",
+    "chat-bison": "https://us-kfp.pkg.dev/ml-pipeline/large-language-model-pipelines/tune-large-chat-model/v3.0.0",
 }
 
 _SDK_PRIVATE_PREVIEW_LAUNCH_STAGE = frozenset(

vertexai/language_models/_language_models.py

@@ -692,7 +692,7 @@ class ChatModel(_ChatModelBase):
     _INSTANCE_SCHEMA_URI = "gs://google-cloud-aiplatform/schema/predict/instance/chat_generation_1.0.0.yaml"
 
 
-class _PreviewChatModel(ChatModel):
+class _PreviewChatModel(ChatModel, _TunableModelMixin):
     _LAUNCH_STAGE = _model_garden_models._SDK_PUBLIC_PREVIEW_LAUNCH_STAGE