Fix the yaml parsing issue. (#36424)

* Fix the yaml parsing issue.

* Fix tests.

* Fix pylint

---------

Co-authored-by: Sagar Sumant <[email protected]>

Author: sagarsumant

sdk/ml/azure-ai-ml/azure/ai/ml/_schema/_finetuning/constants.py

@@ -0,0 +1,19 @@
+# ---------------------------------------------------------
+# Copyright (c) Microsoft Corporation. All rights reserved.
+# ---------------------------------------------------------
+
+# pylint: disable=protected-access
+
+
+class SnakeCaseFineTuningTaskTypes:
+    CHAT_COMPLETION = "chat_completion"
+    TEXT_COMPLETION = "text_completion"
+    TEXT_CLASSIFICATION = "text_classification"
+    QUESTION_ANSWERING = "question_answering"
+    TEXT_SUMMARIZATION = "text_summarization"
+    TOKEN_CLASSIFICATION = "token_classification"
+    TEXT_TRANSLATION = "text_translation"
+    IMAGE_CLASSIFICATION = "image_classification"
+    IMAGE_INSTANCE_SEGMENTATION = "image_instance_segmentation"
+    IMAGE_OBJECT_DETECTION = "image_object_detection"
+    VIDEO_MULTI_OBJECT_TRACKING = "video_multi_object_tracking"

sdk/ml/azure-ai-ml/azure/ai/ml/_schema/_finetuning/finetuning_vertical.py

@@ -5,11 +5,11 @@
 # pylint: disable=unused-argument
 
 from azure.ai.ml._schema._finetuning.finetuning_job import FineTuningJobSchema
+from azure.ai.ml._schema._finetuning.constants import SnakeCaseFineTuningTaskTypes
 from azure.ai.ml._schema.core.fields import NestedField, StringTransformedEnum, UnionField
 from azure.ai.ml.constants import JobType
-from azure.ai.ml._utils.utils import camel_to_snake
+from azure.ai.ml._utils.utils import snake_to_camel
 from azure.ai.ml._schema.job.input_output_entry import MLTableInputSchema, DataInputSchema, ModelInputSchema
-from azure.ai.ml._restclient.v2024_01_01_preview.models import FineTuningTaskType
 from azure.ai.ml.constants._job.finetuning import FineTuningConstants
 from azure.ai.ml._utils._experimental import experimental
 
@@ -24,19 +24,19 @@ class FineTuningVerticalSchema(FineTuningJobSchema):
     validation_data = UnionField([NestedField(MLTableInputSchema), NestedField(DataInputSchema)])
     task = StringTransformedEnum(
         allowed_values=[
-            FineTuningTaskType.CHAT_COMPLETION,
-            FineTuningTaskType.TEXT_COMPLETION,
-            FineTuningTaskType.TEXT_CLASSIFICATION,
-            FineTuningTaskType.QUESTION_ANSWERING,
-            FineTuningTaskType.TEXT_SUMMARIZATION,
-            FineTuningTaskType.TOKEN_CLASSIFICATION,
-            FineTuningTaskType.TEXT_TRANSLATION,
-            FineTuningTaskType.IMAGE_CLASSIFICATION,
-            FineTuningTaskType.IMAGE_INSTANCE_SEGMENTATION,
-            FineTuningTaskType.IMAGE_OBJECT_DETECTION,
-            FineTuningTaskType.VIDEO_MULTI_OBJECT_TRACKING,
+            SnakeCaseFineTuningTaskTypes.CHAT_COMPLETION,
+            SnakeCaseFineTuningTaskTypes.TEXT_COMPLETION,
+            SnakeCaseFineTuningTaskTypes.TEXT_CLASSIFICATION,
+            SnakeCaseFineTuningTaskTypes.QUESTION_ANSWERING,
+            SnakeCaseFineTuningTaskTypes.TEXT_SUMMARIZATION,
+            SnakeCaseFineTuningTaskTypes.TOKEN_CLASSIFICATION,
+            SnakeCaseFineTuningTaskTypes.TEXT_TRANSLATION,
+            SnakeCaseFineTuningTaskTypes.IMAGE_CLASSIFICATION,
+            SnakeCaseFineTuningTaskTypes.IMAGE_INSTANCE_SEGMENTATION,
+            SnakeCaseFineTuningTaskTypes.IMAGE_OBJECT_DETECTION,
+            SnakeCaseFineTuningTaskTypes.VIDEO_MULTI_OBJECT_TRACKING,
         ],
-        casing_transform=camel_to_snake,
+        casing_transform=snake_to_camel,
         data_key=FineTuningConstants.TaskType,
         required=True,
     )

sdk/ml/azure-ai-ml/azure/ai/ml/constants/_job/finetuning.py

@@ -13,3 +13,17 @@ class FineTuningConstants:
     TaskType = "task"
     ModelProvider = "model_provider"
     HyperParameters = "hyperparameters"
+
+
+class FineTuningTaskTypes:
+    CHAT_COMPLETION = "ChatCompletion"
+    TEXT_COMPLETION = "TextCompletion"
+    TEXT_CLASSIFICATION = "TextClassification"
+    QUESTION_ANSWERING = "QuestionAnswering"
+    TEXT_SUMMARIZATION = "TextSummarization"
+    TOKEN_CLASSIFICATION = "TokenClassification"
+    TEXT_TRANSLATION = "TextTranslation"
+    IMAGE_CLASSIFICATION = "ImageClassification"
+    IMAGE_INSTANCE_SEGMENTATION = "ImageInstanceSegmentation"
+    IMAGE_OBJECT_DETECTION = "ImageObjectDetection"
+    VIDEO_MULTI_OBJECT_TRACKING = "VideoMultiObjectTracking"

sdk/ml/azure-ai-ml/tests/finetuning_job/conftest.py

@@ -14,7 +14,6 @@
 
 @pytest.fixture
 def text_completion_dataset() -> Tuple[Input, Input]:
-    # Classification Dataset
     training_data = Input(type=AssetTypes.URI_FILE, path=TEXT_COMPLETION_TRAIN_DATASET)
     validation_data = Input(type=AssetTypes.URI_FILE, path=TEXT_COMPLETION_VALIDATION_DATASET)
     return training_data, validation_data

sdk/ml/azure-ai-ml/tests/finetuning_job/unittests/test_azure_openai_finetuning_job_schema.py

@@ -51,7 +51,7 @@ def expected_azure_openai_finetuning_job_full(
     train_dataset, validation_dataset, mlflow_model_gpt4, hyperparameters
 ) -> RestFineTuningJob:
     custom_model_finetuning_job = AzureOpenAIFineTuningJob(
-        task="chat_completion",
+        task="chatCompletion",
         model=mlflow_model_gpt4,
         training_data=train_dataset,
         validation_data=validation_dataset,

sdk/ml/azure-ai-ml/tests/finetuning_job/unittests/test_custom_model_finetuning_job_schema.py

@@ -46,7 +46,7 @@ def expected_custom_model_finetuning_job_full(
     train_dataset, validation_dataset, mlflow_model_llama, hyperparameters
 ) -> RestFineTuningJob:
     custom_model_finetuning_job = CustomModelFineTuningJob(
-        task="text_completion",
+        task="textCompletion",
         model=mlflow_model_llama,
         training_data=train_dataset,
         validation_data=validation_dataset,