microsoft · Ben Younes (ousamabenyounes) · Aug 10, 2026
diff --git a/README.md b/README.md
@@ -271,7 +271,7 @@ pip install -r requirements.txt
 3. Build the project
 ```bash
 # Manually download the model and run with local path
-huggingface-cli download microsoft/BitNet-b1.58-2B-4T-gguf --local-dir models/BitNet-b1.58-2B-4T
+hf download microsoft/BitNet-b1.58-2B-4T-gguf --local-dir models/BitNet-b1.58-2B-4T
 python setup_env.py -md models/BitNet-b1.58-2B-4T -q i2_s
 
 ```
@@ -382,7 +382,7 @@ python utils/e2e_benchmark.py -m models/dummy-bitnet-125m.tl1.gguf -p 512 -n 128
 
 ```sh
 # Prepare the .safetensors model file
-huggingface-cli download microsoft/bitnet-b1.58-2B-4T-bf16 --local-dir ./models/bitnet-b1.58-2B-4T-bf16
+hf download microsoft/bitnet-b1.58-2B-4T-bf16 --local-dir ./models/bitnet-b1.58-2B-4T-bf16
 
 # Convert to gguf model
 python ./utils/convert-helper-bitnet.py ./models/bitnet-b1.58-2B-4T-bf16

diff --git a/gpu/README.md b/gpu/README.md
@@ -34,7 +34,7 @@ End-to-end inference:
 ```bash
 # Download and convert the BitNet-b1.58-2B model
 mkdir checkpoints
-huggingface-cli download microsoft/bitnet-b1.58-2B-4T-bf16 --local-dir ./checkpoints/bitnet-b1.58-2B-4T-bf16
+hf download microsoft/bitnet-b1.58-2B-4T-bf16 --local-dir ./checkpoints/bitnet-b1.58-2B-4T-bf16
 python ./convert_safetensors.py --safetensors_file ./checkpoints/bitnet-b1.58-2B-4T-bf16/model.safetensors --output checkpoints/model_state.pt --model_name 2B
 python ./convert_checkpoint.py --input ./checkpoints/model_state.pt
 rm ./checkpoints/model_state.pt

diff --git a/setup_env.py b/setup_env.py
@@ -123,7 +123,7 @@ def prepare_model():
         model_dir = os.path.join(model_dir, SUPPORTED_HF_MODELS[hf_url]["model_name"])
         Path(model_dir).mkdir(parents=True, exist_ok=True)
         logging.info(f"Downloading model {hf_url} from HuggingFace to {model_dir}...")
-        run_command(["huggingface-cli", "download", hf_url, "--local-dir", model_dir], log_step="download_model")
+        run_command(["hf", "download", hf_url, "--local-dir", model_dir], log_step="download_model")
     elif not os.path.exists(model_dir):
         logging.error(f"Model directory {model_dir} does not exist.")
         sys.exit(1)

diff --git a/tests/test_setup_env_download.py b/tests/test_setup_env_download.py
@@ -0,0 +1,61 @@
+"""Regression test for issue #560.
+
+`setup_env.py` must download models with the current Hugging Face CLI entry
+point (`hf`). The legacy `huggingface-cli` command was removed from recent
+`huggingface_hub` releases, so the documented `-hr/--hf-repo` download flow
+fails with "command not found" when the script still shells out to it.
+"""
+
+import os
+import sys
+import types
+import unittest
+from pathlib import Path
+from unittest import mock
+
+REPO_ROOT = os.path.dirname(os.path.dirname(os.path.abspath(__file__)))
+if REPO_ROOT not in sys.path:
+    sys.path.insert(0, REPO_ROOT)
+
+import setup_env  # noqa: E402
+
+DOWNLOAD_LOG_STEP = "download_model"
+LEGACY_CLI = "huggingface-cli"
+CURRENT_CLI = "hf"
+
+
+class _StopAfterDownload(Exception):
+    """Sentinel to halt prepare_model() right after the download command."""
+
+
+class DownloadCommandTest(unittest.TestCase):
+    def test_download_uses_current_hf_cli(self):
+        captured = []
+
+        def fake_run_command(command, shell=False, log_step=None):
+            captured.append(command)
+            if log_step == DOWNLOAD_LOG_STEP:
+                raise _StopAfterDownload
+
+        hf_repo = next(iter(setup_env.SUPPORTED_HF_MODELS))
+        fake_args = types.SimpleNamespace(
+            hf_repo=hf_repo,
+            model_dir="models",
+            quant_type="i2_s",
+            quant_embd=False,
+        )
+
+        with mock.patch.object(setup_env, "run_command", fake_run_command), \
+                mock.patch.object(setup_env, "args", fake_args, create=True), \
+                mock.patch.object(Path, "mkdir", lambda *a, **k: None):
+            with self.assertRaises(_StopAfterDownload):
+                setup_env.prepare_model()
+
+        download_cmd = captured[-1]
+        self.assertEqual(download_cmd[0], CURRENT_CLI)
+        self.assertEqual(download_cmd[1], "download")
+        self.assertNotIn(LEGACY_CLI, download_cmd)
+
+
+if __name__ == "__main__":
+    unittest.main()