Minor xet changes: HF_HUB_DISABLE_XET flag, suppress logger.info (#3039)

* Add HF_HUB_DISABLE_XET env var

* Emit warning if hf-hub too old for hf-xet

* Add logger.info() for using hf-xet for upload

* Removing version check

* PR feedback

Author: rajatarya

docs/source/en/package_reference/environment_variables.md

@@ -86,6 +86,7 @@ Integer value to define the number of seconds to wait for server response when d
 ## Xet 
 
 ### Other Xet environment variables
+* [`HF_HUB_DISABLE_XET`](../package_reference/environment_variables#hfhubdisablexet)
 * [`HF_XET_CACHE`](../package_reference/environment_variables#hfxetcache)
 * [`HF_XET_HIGH_PERFORMANCE`](../package_reference/environment_variables#hfxethighperformance)
 * [`HF_XET_RECONSTRUCT_WRITE_SEQUENTIALLY`](../package_reference/environment_variables#hfxetreconstructwritesequentially)
@@ -164,6 +165,10 @@ Each library defines its own policy (i.e. which usage to monitor) but the core i
 
 You can set `HF_HUB_DISABLE_TELEMETRY=1` as environment variable to globally disable telemetry.
 
+### HF_HUB_DISABLE_XET
+
+Set to disable using `hf-xet`, even if it is available in your Python environment. This is since `hf-xet` will be used automatically if it is found, this allows explicitly disabling its usage.
+
 ### HF_HUB_ENABLE_HF_TRANSFER
 
 Set to `True` for faster uploads and downloads from the Hub using `hf_transfer`.

src/huggingface_hub/commands/upload.py

@@ -59,6 +59,7 @@
 from huggingface_hub.errors import RevisionNotFoundError
 from huggingface_hub.hf_api import HfApi
 from huggingface_hub.utils import disable_progress_bars, enable_progress_bars
+from huggingface_hub.utils._runtime import is_xet_available
 
 
 logger = logging.get_logger(__name__)
@@ -215,7 +216,7 @@ def _upload(self) -> str:
             if self.delete is not None and len(self.delete) > 0:
                 warnings.warn("Ignoring `--delete` since a single file is uploaded.")
 
-        if not HF_HUB_ENABLE_HF_TRANSFER:
+        if not is_xet_available() and not HF_HUB_ENABLE_HF_TRANSFER:
             logger.info(
                 "Consider using `hf_transfer` for faster uploads. This solution comes with some limitations. See"
                 " https://huggingface.co/docs/huggingface_hub/hf_transfer for more details."

src/huggingface_hub/utils/_runtime.py

@@ -154,6 +154,10 @@ def get_hf_transfer_version() -> str:
 
 # xet
 def is_xet_available() -> bool:
+    # since hf_xet is automatically used if available, allow explicit disabling via environment variable
+    if constants._is_true(os.environ.get("HF_HUB_DISABLE_XET")):  # type: ignore
+        return False
+
     return is_package_available("hf_xet")
 
 

tests/test_xet_utils.py

@@ -1,6 +1,7 @@
 from unittest.mock import MagicMock
 
 import pytest
+from _pytest.monkeypatch import MonkeyPatch
 
 from huggingface_hub import constants
 from huggingface_hub.utils._xet import (
@@ -252,3 +253,12 @@ def test_fetch_xet_metadata_with_url_invalid_response(mocker) -> None:
 
     with pytest.raises(ValueError, match="Xet headers have not been correctly set by the server."):
         _fetch_xet_connection_info_with_url(url=url, headers=headers)
+
+
+def test_env_var_hf_hub_disable_xet() -> None:
+    """Test that setting HF_HUB_DISABLE_XET results in is_xet_available() returning False."""
+    from huggingface_hub.utils._runtime import is_xet_available
+
+    monkeypatch = MonkeyPatch()
+    monkeypatch.setenv("HF_HUB_DISABLE_XET", "1")
+    assert not is_xet_available()