[Bugfix] add hf_token to EngineArgs (#16093)

Signed-off-by: paolovic <paul-philipp.luley@uzh.ch>
Co-authored-by: paolovic <paul-philipp.luley@uzh.ch>

vllm/engine/arg_utils.py

@@ -138,6 +138,7 @@ class EngineArgs:
     code_revision: Optional[str] = None
     rope_scaling: Optional[Dict[str, Any]] = None
     rope_theta: Optional[float] = None
+    hf_token: Optional[Union[bool, str]] = None
     hf_overrides: Optional[HfOverrides] = None
     tokenizer_revision: Optional[str] = None
     quantization: Optional[str] = None
@@ -602,6 +603,16 @@ class EngineArgs:
                             help='RoPE theta. Use with `rope_scaling`. In '
                             'some cases, changing the RoPE theta improves the '
                             'performance of the scaled model.')
+        parser.add_argument(
+            '--hf-token',
+            type=str,
+            nargs='?',
+            const=True,
+            default=None,
+            help='The token to use as HTTP bearer authorization'
+            ' for remote files. If `True`, will use the token '
+            'generated when running `huggingface-cli login` '
+            '(stored in `~/.huggingface`).')
         parser.add_argument('--hf-overrides',
                             type=json.loads,
                             default=EngineArgs.hf_overrides,
@@ -1038,6 +1049,7 @@ class EngineArgs:
             code_revision=self.code_revision,
             rope_scaling=self.rope_scaling,
             rope_theta=self.rope_theta,
+            hf_token=self.hf_token,
             hf_overrides=self.hf_overrides,
             tokenizer_revision=self.tokenizer_revision,
             max_model_len=self.max_model_len,