huggingface-hub 0.18.0__py3-none-any.whl → 0.19.0__py3-none-any.whl

huggingface_hub/__init__.py

@@ -46,7 +46,7 @@ import sys
 from typing import TYPE_CHECKING
 
 
-__version__ = "0.18.0"
+__version__ = "0.19.0"
 
 # Alphabetical order of definitions is ensured in tests
 # WARNING: any comment added in this dictionary definition will be lost when
@@ -55,6 +55,13 @@ _SUBMOD_ATTRS = {
     "_commit_scheduler": [
         "CommitScheduler",
     ],
+    "_inference_endpoints": [
+        "InferenceEndpoint",
+        "InferenceEndpointError",
+        "InferenceEndpointStatus",
+        "InferenceEndpointTimeoutError",
+        "InferenceEndpointType",
+    ],
     "_login": [
         "interpreter_login",
         "login",
@@ -136,12 +143,10 @@ _SUBMOD_ATTRS = {
         "CommitOperationAdd",
         "CommitOperationCopy",
         "CommitOperationDelete",
-        "DatasetSearchArguments",
         "GitCommitInfo",
         "GitRefInfo",
         "GitRefs",
         "HfApi",
-        "ModelSearchArguments",
         "RepoUrl",
         "User",
         "UserLikes",
@@ -155,6 +160,7 @@ _SUBMOD_ATTRS = {
         "create_commit",
         "create_commits_on_pr",
         "create_discussion",
+        "create_inference_endpoint",
         "create_pull_request",
         "create_repo",
         "create_tag",
@@ -164,6 +170,7 @@ _SUBMOD_ATTRS = {
         "delete_collection_item",
         "delete_file",
         "delete_folder",
+        "delete_inference_endpoint",
         "delete_repo",
         "delete_space_secret",
         "delete_space_storage",
@@ -176,6 +183,7 @@ _SUBMOD_ATTRS = {
         "get_dataset_tags",
         "get_discussion_details",
         "get_full_repo_name",
+        "get_inference_endpoint",
         "get_model_tags",
         "get_repo_discussions",
         "get_space_runtime",
@@ -184,6 +192,7 @@ _SUBMOD_ATTRS = {
         "like",
         "list_datasets",
         "list_files_info",
+        "list_inference_endpoints",
         "list_liked_repos",
         "list_metrics",
         "list_models",
@@ -195,6 +204,7 @@ _SUBMOD_ATTRS = {
         "merge_pull_request",
         "model_info",
         "move_repo",
+        "pause_inference_endpoint",
         "pause_space",
         "preupload_lfs_files",
         "rename_discussion",
@@ -204,13 +214,16 @@ _SUBMOD_ATTRS = {
         "request_space_hardware",
         "request_space_storage",
         "restart_space",
+        "resume_inference_endpoint",
         "run_as_future",
+        "scale_to_zero_inference_endpoint",
         "set_space_sleep_time",
         "space_info",
         "super_squash_history",
         "unlike",
         "update_collection_item",
         "update_collection_metadata",
+        "update_inference_endpoint",
         "update_repo_visibility",
         "upload_file",
         "upload_folder",
@@ -380,6 +393,13 @@ __getattr__, __dir__, __all__ = _attach(__name__, submodules=[], submod_attrs=_S
 # ```
 if TYPE_CHECKING:  # pragma: no cover
     from ._commit_scheduler import CommitScheduler  # noqa: F401
+    from ._inference_endpoints import (
+        InferenceEndpoint,  # noqa: F401
+        InferenceEndpointError,  # noqa: F401
+        InferenceEndpointStatus,  # noqa: F401
+        InferenceEndpointTimeoutError,  # noqa: F401
+        InferenceEndpointType,  # noqa: F401
+    )
     from ._login import (
         interpreter_login,  # noqa: F401
         login,  # noqa: F401
@@ -457,12 +477,10 @@ if TYPE_CHECKING:  # pragma: no cover
         CommitOperationAdd,  # noqa: F401
         CommitOperationCopy,  # noqa: F401
         CommitOperationDelete,  # noqa: F401
-        DatasetSearchArguments,  # noqa: F401
         GitCommitInfo,  # noqa: F401
         GitRefInfo,  # noqa: F401
         GitRefs,  # noqa: F401
         HfApi,  # noqa: F401
-        ModelSearchArguments,  # noqa: F401
         RepoUrl,  # noqa: F401
         User,  # noqa: F401
         UserLikes,  # noqa: F401
@@ -476,6 +494,7 @@ if TYPE_CHECKING:  # pragma: no cover
         create_commit,  # noqa: F401
         create_commits_on_pr,  # noqa: F401
         create_discussion,  # noqa: F401
+        create_inference_endpoint,  # noqa: F401
         create_pull_request,  # noqa: F401
         create_repo,  # noqa: F401
         create_tag,  # noqa: F401
@@ -485,6 +504,7 @@ if TYPE_CHECKING:  # pragma: no cover
         delete_collection_item,  # noqa: F401
         delete_file,  # noqa: F401
         delete_folder,  # noqa: F401
+        delete_inference_endpoint,  # noqa: F401
         delete_repo,  # noqa: F401
         delete_space_secret,  # noqa: F401
         delete_space_storage,  # noqa: F401
@@ -497,6 +517,7 @@ if TYPE_CHECKING:  # pragma: no cover
         get_dataset_tags,  # noqa: F401
         get_discussion_details,  # noqa: F401
         get_full_repo_name,  # noqa: F401
+        get_inference_endpoint,  # noqa: F401
         get_model_tags,  # noqa: F401
         get_repo_discussions,  # noqa: F401
         get_space_runtime,  # noqa: F401
@@ -505,6 +526,7 @@ if TYPE_CHECKING:  # pragma: no cover
         like,  # noqa: F401
         list_datasets,  # noqa: F401
         list_files_info,  # noqa: F401
+        list_inference_endpoints,  # noqa: F401
         list_liked_repos,  # noqa: F401
         list_metrics,  # noqa: F401
         list_models,  # noqa: F401
@@ -516,6 +538,7 @@ if TYPE_CHECKING:  # pragma: no cover
         merge_pull_request,  # noqa: F401
         model_info,  # noqa: F401
         move_repo,  # noqa: F401
+        pause_inference_endpoint,  # noqa: F401
         pause_space,  # noqa: F401
         preupload_lfs_files,  # noqa: F401
         rename_discussion,  # noqa: F401
@@ -525,13 +548,16 @@ if TYPE_CHECKING:  # pragma: no cover
         request_space_hardware,  # noqa: F401
         request_space_storage,  # noqa: F401
         restart_space,  # noqa: F401
+        resume_inference_endpoint,  # noqa: F401
         run_as_future,  # noqa: F401
+        scale_to_zero_inference_endpoint,  # noqa: F401
         set_space_sleep_time,  # noqa: F401
         space_info,  # noqa: F401
         super_squash_history,  # noqa: F401
         unlike,  # noqa: F401
         update_collection_item,  # noqa: F401
         update_collection_metadata,  # noqa: F401
+        update_inference_endpoint,  # noqa: F401
         update_repo_visibility,  # noqa: F401
         upload_file,  # noqa: F401
         upload_folder,  # noqa: F401

huggingface_hub/_inference_endpoints.py

@@ -0,0 +1,348 @@
+import time
+from dataclasses import dataclass, field
+from datetime import datetime
+from enum import Enum
+from typing import TYPE_CHECKING, Dict, Optional
+
+from .inference._client import InferenceClient
+from .inference._generated._async_client import AsyncInferenceClient
+from .utils import logging, parse_datetime
+
+
+if TYPE_CHECKING:
+    from .hf_api import HfApi
+
+
+logger = logging.get_logger(__name__)
+
+
+class InferenceEndpointError(Exception):
+    """Generic exception when dealing with Inference Endpoints."""
+
+
+class InferenceEndpointTimeoutError(InferenceEndpointError, TimeoutError):
+    """Exception for timeouts while waiting for Inference Endpoint."""
+
+
+class InferenceEndpointStatus(str, Enum):
+    PENDING = "pending"
+    INITIALIZING = "initializing"
+    UPDATING = "updating"
+    UPDATE_FAILED = "updateFailed"
+    RUNNING = "running"
+    PAUSED = "paused"
+    FAILED = "failed"
+    SCALED_TO_ZERO = "scaledToZero"
+
+
+class InferenceEndpointType(str, Enum):
+    PUBlIC = "public"
+    PROTECTED = "protected"
+    PRIVATE = "private"
+
+
+@dataclass
+class InferenceEndpoint:
+    """
+    Contains information about a deployed Inference Endpoint.
+
+    Args:
+        name (`str`):
+            The unique name of the Inference Endpoint.
+        namespace (`str`):
+            The namespace where the Inference Endpoint is located.
+        repository (`str`):
+            The name of the model repository deployed on this Inference Endpoint.
+        status ([`InferenceEndpointStatus`]):
+            The current status of the Inference Endpoint.
+        url (`str`, *optional*):
+            The URL of the Inference Endpoint, if available. Only a deployed Inference Endpoint will have a URL.
+        framework (`str`):
+            The machine learning framework used for the model.
+        revision (`str`):
+            The specific model revision deployed on the Inference Endpoint.
+        task (`str`):
+            The task associated with the deployed model.
+        created_at (`datetime.datetime`):
+            The timestamp when the Inference Endpoint was created.
+        updated_at (`datetime.datetime`):
+            The timestamp of the last update of the Inference Endpoint.
+        type ([`InferenceEndpointType`]):
+            The type of the Inference Endpoint (public, protected, private).
+        raw (`Dict`):
+            The raw dictionary data returned from the API.
+        token (`str`, *optional*):
+            Authentication token for the Inference Endpoint, if set when requesting the API.
+
+    Example:
+    ```python
+    >>> from huggingface_hub import get_inference_endpoint
+    >>> endpoint = get_inference_endpoint("my-text-to-image")
+    >>> endpoint
+    InferenceEndpoint(name='my-text-to-image', ...)
+
+    # Get status
+    >>> endpoint.status
+    'running'
+    >>> endpoint.url
+    'https://my-text-to-image.region.vendor.endpoints.huggingface.cloud'
+
+    # Run inference
+    >>> endpoint.client.text_to_image(...)
+
+    # Pause endpoint to save $$$
+    >>> endpoint.pause()
+
+    # ...
+    # Resume and wait for deployment
+    >>> endpoint.resume()
+    >>> endpoint.wait()
+    >>> endpoint.client.text_to_image(...)
+    ```
+    """
+
+    # Field in __repr__
+    name: str = field(init=False)
+    namespace: str
+    repository: str = field(init=False)
+    status: InferenceEndpointStatus = field(init=False)
+    url: Optional[str] = field(init=False)
+
+    # Other fields
+    framework: str = field(repr=False, init=False)
+    revision: str = field(repr=False, init=False)
+    task: str = field(repr=False, init=False)
+    created_at: datetime = field(repr=False, init=False)
+    updated_at: datetime = field(repr=False, init=False)
+    type: InferenceEndpointType = field(repr=False, init=False)
+
+    # Raw dict from the API
+    raw: Dict = field(repr=False)
+
+    # Internal fields
+    _token: Optional[str] = field(repr=False, compare=False)
+    _api: "HfApi" = field(repr=False, compare=False)
+
+    @classmethod
+    def from_raw(
+        cls, raw: Dict, namespace: str, token: Optional[str] = None, api: Optional["HfApi"] = None
+    ) -> "InferenceEndpoint":
+        """Initialize object from raw dictionary."""
+        if api is None:
+            from .hf_api import HfApi
+
+            api = HfApi()
+        if token is None:
+            token = api.token
+
+        # All other fields are populated in __post_init__
+        return cls(raw=raw, namespace=namespace, _token=token, _api=api)
+
+    def __post_init__(self) -> None:
+        """Populate fields from raw dictionary."""
+        self._populate_from_raw()
+
+    @property
+    def client(self) -> InferenceClient:
+        """Returns a client to make predictions on this Inference Endpoint.
+
+        Raises:
+            [`InferenceEndpointError`]: If the Inference Endpoint is not yet deployed.
+        """
+        if self.url is None:
+            raise InferenceEndpointError(
+                "Cannot create a client for this Inference Endpoint as it is not yet deployed. "
+                "Please wait for the Inference Endpoint to be deployed using `endpoint.wait()` and try again."
+            )
+        return InferenceClient(model=self.url, token=self._token)
+
+    @property
+    def async_client(self) -> AsyncInferenceClient:
+        """Returns a client to make predictions on this Inference Endpoint.
+
+        Raises:
+            [`InferenceEndpointError`]: If the Inference Endpoint is not yet deployed.
+        """
+        if self.url is None:
+            raise InferenceEndpointError(
+                "Cannot create a client for this Inference Endpoint as it is not yet deployed. "
+                "Please wait for the Inference Endpoint to be deployed using `endpoint.wait()` and try again."
+            )
+        return AsyncInferenceClient(model=self.url, token=self._token)
+
+    def wait(self, timeout: Optional[int] = None, refresh_every: int = 5) -> None:
+        """Wait for the Inference Endpoint to be deployed.
+
+        Information from the server will be fetched every 1s. If the Inference Endpoint is not deployed after `timeout`
+        seconds, a [`InferenceEndpointTimeoutError`] will be raised. The [`InferenceEndpoint`] will be mutated in place with the latest
+        data.
+
+        Args:
+            timeout (`int`, *optional*):
+                The maximum time to wait for the Inference Endpoint to be deployed, in seconds. If `None`, will wait
+                indefinitely.
+            refresh_every (`int`, *optional*):
+                The time to wait between each fetch of the Inference Endpoint status, in seconds. Defaults to 5s.
+        """
+        if self.url is not None:  # Means the endpoint is deployed
+            logger.info("Inference Endpoint is ready to be used.")
+            return
+
+        if timeout is not None and timeout < 0:
+            raise ValueError("`timeout` cannot be negative.")
+        if refresh_every <= 0:
+            raise ValueError("`refresh_every` must be positive.")
+
+        start = time.time()
+        while True:
+            self.fetch()
+            if self.url is not None:  # Means the endpoint is deployed
+                logger.info("Inference Endpoint is ready to be used.")
+                return
+            if timeout is not None:
+                if time.time() - start > timeout:
+                    raise InferenceEndpointTimeoutError("Timeout while waiting for Inference Endpoint to be deployed.")
+            logger.info(f"Inference Endpoint is not deployed yet ({self.status}). Waiting {refresh_every}s...")
+            time.sleep(refresh_every)
+
+    def fetch(self) -> "InferenceEndpoint":
+        """Fetch latest information about the Inference Endpoint."""
+        obj = self._api.get_inference_endpoint(name=self.name, namespace=self.namespace, token=self._token)
+        self.raw = obj.raw
+        self._populate_from_raw()
+        return self
+
+    def update(
+        self,
+        *,
+        # Compute update
+        accelerator: Optional[str] = None,
+        instance_size: Optional[str] = None,
+        instance_type: Optional[str] = None,
+        min_replica: Optional[int] = None,
+        max_replica: Optional[int] = None,
+        # Model update
+        repository: Optional[str] = None,
+        framework: Optional[str] = None,
+        revision: Optional[str] = None,
+        task: Optional[str] = None,
+    ) -> "InferenceEndpoint":
+        """Update the Inference Endpoint.
+
+        This method allows the update of either the compute configuration, the deployed model, or both. All arguments are
+        optional but at least one must be provided.
+
+        This is an alias for [`HfApi.update_inference_endpoint`]. The current object is mutated in place with the
+        latest data from the server.
+
+        Args:
+            accelerator (`str`, *optional*):
+                The hardware accelerator to be used for inference (e.g. `"cpu"`).
+            instance_size (`str`, *optional*):
+                The size or type of the instance to be used for hosting the model (e.g. `"large"`).
+            instance_type (`str`, *optional*):
+                The cloud instance type where the Inference Endpoint will be deployed (e.g. `"c6i"`).
+            min_replica (`int`, *optional*):
+                The minimum number of replicas (instances) to keep running for the Inference Endpoint.
+            max_replica (`int`, *optional*):
+                The maximum number of replicas (instances) to scale to for the Inference Endpoint.
+
+            repository (`str`, *optional*):
+                The name of the model repository associated with the Inference Endpoint (e.g. `"gpt2"`).
+            framework (`str`, *optional*):
+                The machine learning framework used for the model (e.g. `"custom"`).
+            revision (`str`, *optional*):
+                The specific model revision to deploy on the Inference Endpoint (e.g. `"6c0e6080953db56375760c0471a8c5f2929baf11"`).
+            task (`str`, *optional*):
+                The task on which to deploy the model (e.g. `"text-classification"`).
+        """
+        # Make API call
+        obj = self._api.update_inference_endpoint(
+            name=self.name,
+            namespace=self.namespace,
+            accelerator=accelerator,
+            instance_size=instance_size,
+            instance_type=instance_type,
+            min_replica=min_replica,
+            max_replica=max_replica,
+            repository=repository,
+            framework=framework,
+            revision=revision,
+            task=task,
+            token=self._token,
+        )
+
+        # Mutate current object
+        self.raw = obj.raw
+        self._populate_from_raw()
+        return self
+
+    def pause(self) -> "InferenceEndpoint":
+        """Pause the Inference Endpoint.
+
+        A paused Inference Endpoint will not be charged. It can be resumed at any time using [`InferenceEndpoint.resume`].
+        This is different than scaling the Inference Endpoint to zero with [`InferenceEndpoint.scale_to_zero`], which
+        would be automatically restarted when a request is made to it.
+
+        This is an alias for [`HfApi.pause_inference_endpoint`]. The current object is mutated in place with the
+        latest data from the server.
+        """
+        obj = self._api.pause_inference_endpoint(name=self.name, namespace=self.namespace, token=self._token)
+        self.raw = obj.raw
+        self._populate_from_raw()
+        return self
+
+    def resume(self) -> "InferenceEndpoint":
+        """Resume the Inference Endpoint.
+
+        This is an alias for [`HfApi.resume_inference_endpoint`]. The current object is mutated in place with the
+        latest data from the server.
+        """
+        obj = self._api.resume_inference_endpoint(name=self.name, namespace=self.namespace, token=self._token)
+        self.raw = obj.raw
+        self._populate_from_raw()
+        return self
+
+    def scale_to_zero(self) -> "InferenceEndpoint":
+        """Scale Inference Endpoint to zero.
+
+        An Inference Endpoint scaled to zero will not be charged. It will be resume on the next request to it, with a
+        cold start delay. This is different than pausing the Inference Endpoint with [`InferenceEndpoint.pause`], which
+        would require a manual resume with [`InferenceEndpoint.resume`].
+
+        This is an alias for [`HfApi.scale_to_zero_inference_endpoint`]. The current object is mutated in place with the
+        latest data from the server.
+        """
+        obj = self._api.scale_to_zero_inference_endpoint(name=self.name, namespace=self.namespace, token=self._token)
+        self.raw = obj.raw
+        self._populate_from_raw()
+        return self
+
+    def delete(self) -> None:
+        """Delete the Inference Endpoint.
+
+        This operation is not reversible. If you don't want to be charged for an Inference Endpoint, it is preferable
+        to pause it with [`InferenceEndpoint.pause`] or scale it to zero with [`InferenceEndpoint.scale_to_zero`].
+
+        This is an alias for [`HfApi.delete_inference_endpoint`].
+        """
+        self._api.delete_inference_endpoint(name=self.name, namespace=self.namespace, token=self._token)
+
+    def _populate_from_raw(self) -> None:
+        """Populate fields from raw dictionary.
+
+        Called in __post_init__ + each time the Inference Endpoint is updated.
+        """
+        # Repr fields
+        self.name = self.raw["name"]
+        self.repository = self.raw["model"]["repository"]
+        self.status = self.raw["status"]["state"]
+        self.url = self.raw["status"].get("url")
+
+        # Other fields
+        self.framework = self.raw["model"]["framework"]
+        self.revision = self.raw["model"]["revision"]
+        self.task = self.raw["model"]["task"]
+        self.created_at = parse_datetime(self.raw["status"]["createdAt"])
+        self.updated_at = parse_datetime(self.raw["status"]["updatedAt"])
+        self.type = self.raw["type"]

huggingface_hub/_login.py

@@ -36,6 +36,14 @@ from .utils import (
 
 logger = logging.get_logger(__name__)
 
+_HF_LOGO_ASCII = """
+    _|    _|  _|    _|    _|_|_|    _|_|_|  _|_|_|  _|      _|    _|_|_|      _|_|_|_|    _|_|      _|_|_|  _|_|_|_|
+    _|    _|  _|    _|  _|        _|          _|    _|_|    _|  _|            _|        _|    _|  _|        _|
+    _|_|_|_|  _|    _|  _|  _|_|  _|  _|_|    _|    _|  _|  _|  _|  _|_|      _|_|_|    _|_|_|_|  _|        _|_|_|
+    _|    _|  _|    _|  _|    _|  _|    _|    _|    _|    _|_|  _|    _|      _|        _|    _|  _|        _|
+    _|    _|    _|_|      _|_|_|    _|_|_|  _|_|_|  _|      _|    _|_|_|      _|        _|    _|    _|_|_|  _|_|_|_|
+"""
+
 
 def login(
     token: Optional[str] = None,
@@ -144,13 +152,7 @@ def interpreter_login(new_session: bool = True, write_permission: bool = False)
         print("User is already logged in.")
         return
 
-    print("""
-    _|    _|  _|    _|    _|_|_|    _|_|_|  _|_|_|  _|      _|    _|_|_|      _|_|_|_|    _|_|      _|_|_|  _|_|_|_|
-    _|    _|  _|    _|  _|        _|          _|    _|_|    _|  _|            _|        _|    _|  _|        _|
-    _|_|_|_|  _|    _|  _|  _|_|  _|  _|_|    _|    _|  _|  _|  _|  _|_|      _|_|_|    _|_|_|_|  _|        _|_|_|
-    _|    _|  _|    _|  _|    _|  _|    _|    _|    _|    _|_|  _|    _|      _|        _|    _|  _|        _|
-    _|    _|    _|_|      _|_|_|    _|_|_|  _|_|_|  _|      _|    _|_|_|      _|        _|    _|    _|_|_|  _|_|_|_|
-    """)
+    print(_HF_LOGO_ASCII)
     if HfFolder.get_token() is not None:
         print(
             "    A token is already saved on your machine. Run `huggingface-cli"

huggingface_hub/_multi_commits.py

@@ -15,13 +15,13 @@
 """Contains utilities to multi-commits (i.e. push changes iteratively on a PR)."""
 import re
 from dataclasses import dataclass, field
-from hashlib import sha256
 from typing import TYPE_CHECKING, Iterable, List, Optional, Set, Tuple, Union
 
 from ._commit_api import CommitOperationAdd, CommitOperationDelete
 from .community import DiscussionWithDetails
 from .utils import experimental
 from .utils._cache_manager import _format_size
+from .utils.insecure_hashlib import sha256
 
 
 if TYPE_CHECKING:

huggingface_hub/_snapshot_download.py

@@ -6,9 +6,10 @@ from tqdm.auto import tqdm as base_tqdm
 from tqdm.contrib.concurrent import thread_map
 
 from .constants import (
+    DEFAULT_ETAG_TIMEOUT,
     DEFAULT_REVISION,
+    HF_HUB_CACHE,
     HF_HUB_ENABLE_HF_TRANSFER,
-    HUGGINGFACE_HUB_CACHE,
     REPO_TYPES,
 )
 from .file_download import REGEX_COMMIT_HASH, hf_hub_download, repo_folder_name
@@ -26,7 +27,6 @@ def snapshot_download(
     *,
     repo_type: Optional[str] = None,
     revision: Optional[str] = None,
-    endpoint: Optional[str] = None,
     cache_dir: Union[str, Path, None] = None,
     local_dir: Union[str, Path, None] = None,
     local_dir_use_symlinks: Union[bool, Literal["auto"]] = "auto",
@@ -34,7 +34,7 @@ def snapshot_download(
     library_version: Optional[str] = None,
     user_agent: Optional[Union[Dict, str]] = None,
     proxies: Optional[Dict] = None,
-    etag_timeout: float = 10,
+    etag_timeout: float = DEFAULT_ETAG_TIMEOUT,
     resume_download: bool = False,
     force_download: bool = False,
     token: Optional[Union[bool, str]] = None,
@@ -43,6 +43,7 @@ def snapshot_download(
     ignore_patterns: Optional[Union[List[str], str]] = None,
     max_workers: int = 8,
     tqdm_class: Optional[base_tqdm] = None,
+    endpoint: Optional[str] = None,
 ) -> str:
     """Download repo files.
 
@@ -78,9 +79,6 @@ def snapshot_download(
         revision (`str`, *optional*):
             An optional Git revision id which can be a branch name, a tag, or a
             commit hash.
-        endpoint (`str`, *optional*):
-            Hugging Face Hub base url. Will default to https://huggingface.co/. Otherwise, one can set the `HF_ENDPOINT`
-            environment variable.
         cache_dir (`str`, `Path`, *optional*):
             Path to the folder where cached files are stored.
         local_dir (`str` or `Path`, *optional*):
@@ -146,7 +144,7 @@ def snapshot_download(
     </Tip>
     """
     if cache_dir is None:
-        cache_dir = HUGGINGFACE_HUB_CACHE
+        cache_dir = HF_HUB_CACHE
     if revision is None:
         revision = DEFAULT_REVISION
     if isinstance(cache_dir, Path):
@@ -188,6 +186,7 @@ def snapshot_download(
     api = HfApi(library_name=library_name, library_version=library_version, user_agent=user_agent, endpoint=endpoint)
     repo_info = api.repo_info(repo_id=repo_id, repo_type=repo_type, revision=revision, token=token)
     assert repo_info.sha is not None, "Repo info returned from server must have a revision sha."
+    assert repo_info.siblings is not None, "Repo info returned from server must have a siblings list."
 
     filtered_repo_files = list(
         filter_repo_objects(

huggingface_hub/_space_api.py

@@ -61,8 +61,11 @@ class SpaceHardware(str, Enum):
     CPU_UPGRADE = "cpu-upgrade"
     T4_SMALL = "t4-small"
     T4_MEDIUM = "t4-medium"
+    ZERO_A10G = "zero-a10g"
     A10G_SMALL = "a10g-small"
     A10G_LARGE = "a10g-large"
+    A10G_LARGEX2 = "a10g-largex2"
+    A10G_LARGEX4 = "a10g-largex4"
     A100_LARGE = "a100-large"
 
 
@@ -116,10 +119,10 @@ class SpaceRuntime:
 
     def __init__(self, data: Dict) -> None:
         self.stage = data["stage"]
-        self.hardware = data["hardware"]["current"]
-        self.requested_hardware = data["hardware"]["requested"]
-        self.sleep_time = data["gcTimeout"]
-        self.storage = data["storage"]
+        self.hardware = data.get("hardware", {}).get("current")
+        self.requested_hardware = data.get("hardware", {}).get("requested")
+        self.sleep_time = data.get("gcTimeout")
+        self.storage = data.get("storage")
         self.raw = data
 
 

huggingface_hub/_tensorboard_logger.py

@@ -24,6 +24,7 @@ if is_tensorboard_available():
     from tensorboardX import SummaryWriter
 
     # TODO: clarify: should we import from torch.utils.tensorboard ?
+
 else:
     SummaryWriter = object  # Dummy class to avoid failing at import. Will raise on instance creation.