Merge branch 'main' into ODSC-64668/report_misformat

Author: ahosler

THIRD_PARTY_LICENSES.txt

@@ -24,6 +24,12 @@ autots
 * Source code: https://github.com/winedarksea/AutoTS
 * Project home: https://winedarksea.github.io/AutoTS/build/html/index.html
 
+autogen
+* Copyright (c) 2024 Microsoft Corporation.
+* License: MIT License
+* Source code: https://github.com/microsoft/autogen
+* Project home: microsoft.github.io/autogen/
+
 bokeh
 * Copyright Copyright (c) 2012 - 2021, Anaconda, Inc., and Bokeh Contributors
 * License: BSD 3-Clause "New" or "Revised" License

ads/aqua/common/utils.py

@@ -788,13 +788,14 @@ def get_ocid_substring(ocid: str, key_len: int) -> str:
     return ocid[-key_len:] if ocid and len(ocid) > key_len else ""
 
 
-def upload_folder(os_path: str, local_dir: str, model_name: str) -> str:
+def upload_folder(os_path: str, local_dir: str, model_name: str, exclude_pattern: str = None) -> str:
     """Upload the local folder to the object storage
 
     Args:
         os_path (str): object storage URI with prefix. This is the path to upload
         local_dir (str): Local directory where the object is downloaded
         model_name (str): Name of the huggingface model
+        exclude_pattern (optional, str): The matching pattern of files to be excluded from uploading.
     Retuns:
         str: Object name inside the bucket
     """
@@ -804,6 +805,8 @@ def upload_folder(os_path: str, local_dir: str, model_name: str) -> str:
     auth_state = AuthState()
     object_path = os_details.filepath.rstrip("/") + "/" + model_name + "/"
     command = f"oci os object bulk-upload --src-dir {local_dir} --prefix {object_path} -bn {os_details.bucket} -ns {os_details.namespace} --auth {auth_state.oci_iam_type} --profile {auth_state.oci_key_profile} --no-overwrite"
+    if exclude_pattern:
+        command += f" --exclude {exclude_pattern}"
     try:
         logger.info(f"Running: {command}")
         subprocess.check_call(shlex.split(command))

ads/aqua/constants.py

@@ -35,6 +35,7 @@
 AQUA_MODEL_ARTIFACT_CONFIG_MODEL_NAME = "_name_or_path"
 AQUA_MODEL_ARTIFACT_CONFIG_MODEL_TYPE = "model_type"
 AQUA_MODEL_ARTIFACT_FILE = "model_file"
+HF_METADATA_FOLDER = ".cache/"
 HF_LOGIN_DEFAULT_TIMEOUT = 2
 
 TRAINING_METRICS_FINAL = "training_metrics_final"

ads/aqua/extension/model_handler.py

@@ -129,6 +129,8 @@ def post(self, *args, **kwargs):
             str(input_data.get("download_from_hf", "false")).lower() == "true"
         )
         inference_container_uri = input_data.get("inference_container_uri")
+        allow_patterns = input_data.get("allow_patterns")
+        ignore_patterns = input_data.get("ignore_patterns")
 
         return self.finish(
             AquaModelApp().register(
@@ -141,6 +143,8 @@ def post(self, *args, **kwargs):
                 project_id=project_id,
                 model_file=model_file,
                 inference_container_uri=inference_container_uri,
+                allow_patterns=allow_patterns,
+                ignore_patterns=ignore_patterns,
             )
         )
 

ads/aqua/model/entities.py

@@ -289,6 +289,8 @@ class ImportModelDetails(CLIBuilderMixin):
     project_id: Optional[str] = None
     model_file: Optional[str] = None
     inference_container_uri: Optional[str] = None
+    allow_patterns: Optional[List[str]] = None
+    ignore_patterns: Optional[List[str]] = None
 
     def __post_init__(self):
         self._command = "model register"

ads/aqua/model/model.py

@@ -40,6 +40,7 @@
     AQUA_MODEL_ARTIFACT_CONFIG_MODEL_TYPE,
     AQUA_MODEL_ARTIFACT_FILE,
     AQUA_MODEL_TYPE_CUSTOM,
+    HF_METADATA_FOLDER,
     LICENSE_TXT,
     MODEL_BY_REFERENCE_OSS_PATH_KEY,
     README,
@@ -1274,6 +1275,8 @@ def _download_model_from_hf(
         model_name: str,
         os_path: str,
         local_dir: str = None,
+        allow_patterns: List[str] = None,
+        ignore_patterns: List[str] = None,
     ) -> str:
         """This helper function downloads the model artifact from Hugging Face to a local folder, then uploads
         to object storage location.
@@ -1283,6 +1286,12 @@ def _download_model_from_hf(
         model_name (str): The huggingface model name.
         os_path (str): The OS path where the model files are located.
         local_dir (str): The local temp dir to store the huggingface model.
+        allow_patterns (list): Model files matching at least one pattern are downloaded.
+            Example: ["*.json"] will download all .json files. ["folder/*"] will download all files under `folder`.
+            Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
+        ignore_patterns (list): Model files matching any of the patterns are not downloaded.
+            Example: ["*.json"] will ignore all .json files. ["folder/*"] will ignore all files under `folder`.
+            Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
 
         Returns
         -------
@@ -1293,30 +1302,19 @@ def _download_model_from_hf(
         if not local_dir:
             local_dir = os.path.join(os.path.expanduser("~"), "cached-model")
         local_dir = os.path.join(local_dir, model_name)
-        retry = 10
-        i = 0
-        huggingface_download_err_message = None
-        while i < retry:
-            try:
-                # Download to cache folder. The while loop retries when there is a network failure
-                snapshot_download(repo_id=model_name)
-            except Exception as e:
-                huggingface_download_err_message = str(e)
-                i += 1
-            else:
-                break
-        if i == retry:
-            raise Exception(
-                f"Could not download the model {model_name} from https://huggingface.co with message {huggingface_download_err_message}"
-            )
         os.makedirs(local_dir, exist_ok=True)
-        # Copy the model from the cache to destination
-        snapshot_download(repo_id=model_name, local_dir=local_dir)
-        # Upload to object storage
+        snapshot_download(
+            repo_id=model_name,
+            local_dir=local_dir,
+            allow_patterns=allow_patterns,
+            ignore_patterns=ignore_patterns,
+        )
+        # Upload to object storage and skip .cache/huggingface/ folder
         model_artifact_path = upload_folder(
             os_path=os_path,
             local_dir=local_dir,
             model_name=model_name,
+            exclude_pattern=f"{HF_METADATA_FOLDER}*"
         )
 
         return model_artifact_path
@@ -1335,6 +1333,12 @@ def register(
                 os_path (str): Object storage destination URI to store the downloaded model. Format: oci://bucket-name@namespace/prefix
                 inference_container (str): selects service defaults
                 finetuning_container (str): selects service defaults
+                allow_patterns (list): Model files matching at least one pattern are downloaded.
+                    Example: ["*.json"] will download all .json files. ["folder/*"] will download all files under `folder`.
+                    Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
+                ignore_patterns (list): Model files matching any of the patterns are not downloaded.
+                    Example: ["*.json"] will ignore all .json files. ["folder/*"] will ignore all files under `folder`.
+                    Patterns are Standard Wildcards (globbing patterns) and rules can be found here: https://docs.python.org/3/library/fnmatch.html
 
         Returns:
             AquaModel:
@@ -1381,6 +1385,8 @@ def register(
                 model_name=model_name,
                 os_path=import_model_details.os_path,
                 local_dir=import_model_details.local_dir,
+                allow_patterns=import_model_details.allow_patterns,
+                ignore_patterns=import_model_details.ignore_patterns,
             ).rstrip("/")
         else:
             artifact_path = import_model_details.os_path.rstrip("/")