chore: disable TQDM on server args

Signed-off-by: Aaron Pham <[email protected]>

Author: aarnphm

vllm/config.py

@@ -1277,13 +1277,15 @@ class LoadConfig:
         ignore_patterns: The list of patterns to ignore when loading the model.
             Default to "original/**/*" to avoid repeated loading of llama's
             checkpoints.
+        use_tqdm_on_load: Whether to enable tqdm for showing progress bar
     """
 
     load_format: Union[str, LoadFormat, "BaseModelLoader"] = LoadFormat.AUTO
     download_dir: Optional[str] = None
     model_loader_extra_config: Optional[Union[str, dict]] = field(
         default_factory=dict)
     ignore_patterns: Optional[Union[list[str], str]] = None
+    use_tqdm_on_load: bool = False
 
     def compute_hash(self) -> str:
         """

vllm/engine/arg_utils.py

@@ -217,6 +217,7 @@ class EngineArgs:
     additional_config: Optional[Dict[str, Any]] = None
     enable_reasoning: Optional[bool] = None
     reasoning_parser: Optional[str] = None
+    use_tqdm_on_load: bool = True
 
     def __post_init__(self):
         if not self.tokenizer:
@@ -751,6 +752,12 @@ def add_cli_args(parser: FlexibleArgumentParser) -> FlexibleArgumentParser:
                             default=1,
                             help=('Maximum number of forward steps per '
                                   'scheduler call.'))
+        parser.add_argument(
+            '--use-tqdm-on-load',
+            action=StoreBoolean,
+            default=EngineArgs.use_tqdm_on_load,
+            help='Whether to disable progress bar (using tqdm)',
+        )
 
         parser.add_argument(
             '--multi-step-stream-outputs',
@@ -1179,6 +1186,7 @@ def create_load_config(self) -> LoadConfig:
             download_dir=self.download_dir,
             model_loader_extra_config=self.model_loader_extra_config,
             ignore_patterns=self.ignore_patterns,
+            use_tqdm_on_load=not self.use_tqdm_on_load,
         )
 
     def create_engine_config(self,

vllm/model_executor/model_loader/loader.py

@@ -354,11 +354,18 @@ def _get_weights_iterator(
                 self.load_config.download_dir,
                 hf_folder,
                 hf_weights_files,
+                self.load_config.use_tqdm_on_load,
             )
         elif use_safetensors:
-            weights_iterator = safetensors_weights_iterator(hf_weights_files)
+            weights_iterator = safetensors_weights_iterator(
+                hf_weights_files,
+                self.load_config.use_tqdm_on_load,
+            )
         else:
-            weights_iterator = pt_weights_iterator(hf_weights_files)
+            weights_iterator = pt_weights_iterator(
+                hf_weights_files,
+                self.load_config.use_tqdm_on_load,
+            )
 
         if current_platform.is_tpu():
             # In PyTorch XLA, we should call `xm.mark_step` frequently so that
@@ -806,9 +813,15 @@ def _prepare_weights(self, model_name_or_path: str,
 
     def _hf_weight_iter(self, hf_weights_files, use_safetensors: bool):
         if use_safetensors:
-            iterator = safetensors_weights_iterator(hf_weights_files)
+            iterator = safetensors_weights_iterator(
+                hf_weights_files,
+                self.load_config.use_tqdm_on_load,
+            )
         else:
-            iterator = pt_weights_iterator(hf_weights_files)
+            iterator = pt_weights_iterator(
+                hf_weights_files,
+                self.load_config.use_tqdm_on_load,
+            )
         for org_name, param in iterator:
             # mapping weight names from transformers to vllm while preserving
             # original names.
@@ -1396,7 +1409,10 @@ def _get_weights_iterator(
             revision: str) -> Generator[Tuple[str, torch.Tensor], None, None]:
         """Get an iterator for the model weights based on the load format."""
         hf_weights_files = self._prepare_weights(model_or_path, revision)
-        return runai_safetensors_weights_iterator(hf_weights_files)
+        return runai_safetensors_weights_iterator(
+            hf_weights_files,
+            self.load_config.use_tqdm_on_load,
+        )
 
     def download_model(self, model_config: ModelConfig) -> None:
         """Download model if necessary"""

vllm/model_executor/model_loader/weight_utils.py

@@ -366,16 +366,23 @@ def filter_files_not_needed_for_inference(
 _BAR_FORMAT = "{desc}: {percentage:3.0f}% Completed | {n_fmt}/{total_fmt} [{elapsed}<{remaining}, {rate_fmt}]\n"  # noqa: E501
 
 
+def enable_tqdm(disable_progress_bar: bool):
+    return not disable_progress_bar and (
+        not torch.distributed.is_initialized()
+        or torch.distributed.get_rank() == 0)
+
+
 def np_cache_weights_iterator(
-    model_name_or_path: str, cache_dir: Optional[str], hf_folder: str,
-    hf_weights_files: List[str]
+    model_name_or_path: str,
+    cache_dir: Optional[str],
+    hf_folder: str,
+    hf_weights_files: List[str],
+    disable_progress_bar: bool,
 ) -> Generator[Tuple[str, torch.Tensor], None, None]:
     """Iterate over the weights in the model np files.
 
     Will dump the model weights to numpy files if they are not already dumped.
     """
-    enable_tqdm = not torch.distributed.is_initialized(
-    ) or torch.distributed.get_rank() == 0
     # Convert the model weights from torch tensors to numpy arrays for
     # faster loading.
     np_folder = os.path.join(hf_folder, "np")
@@ -389,7 +396,7 @@ def np_cache_weights_iterator(
             for bin_file in tqdm(
                     hf_weights_files,
                     desc="Loading np_cache checkpoint shards",
-                    disable=not enable_tqdm,
+                    disable=not enable_tqdm(disable_progress_bar),
                     bar_format=_BAR_FORMAT,
             ):
                 state = torch.load(bin_file,
@@ -414,15 +421,13 @@ def np_cache_weights_iterator(
 
 
 def safetensors_weights_iterator(
-    hf_weights_files: List[str]
+    hf_weights_files: List[str], disable_progress_bar: bool
 ) -> Generator[Tuple[str, torch.Tensor], None, None]:
     """Iterate over the weights in the model safetensor files."""
-    enable_tqdm = not torch.distributed.is_initialized(
-    ) or torch.distributed.get_rank() == 0
     for st_file in tqdm(
             hf_weights_files,
             desc="Loading safetensors checkpoint shards",
-            disable=not enable_tqdm,
+            disable=not enable_tqdm(disable_progress_bar),
             bar_format=_BAR_FORMAT,
     ):
         with safe_open(st_file, framework="pt") as f:
@@ -432,32 +437,28 @@ def safetensors_weights_iterator(
 
 
 def runai_safetensors_weights_iterator(
-    hf_weights_files: List[str]
+    hf_weights_files: List[str], disable_progress_bar: bool
 ) -> Generator[Tuple[str, torch.Tensor], None, None]:
     """Iterate over the weights in the model safetensor files."""
-    enable_tqdm = not torch.distributed.is_initialized(
-    ) or torch.distributed.get_rank() == 0
     with SafetensorsStreamer() as streamer:
         for st_file in tqdm(
                 hf_weights_files,
                 desc="Loading safetensors using Runai Model Streamer",
-                disable=not enable_tqdm,
+                disable=not enable_tqdm(disable_progress_bar),
                 bar_format=_BAR_FORMAT,
         ):
             streamer.stream_file(st_file)
             yield from streamer.get_tensors()
 
 
 def pt_weights_iterator(
-    hf_weights_files: List[str]
+    hf_weights_files: List[str], disable_progress_bar: bool
 ) -> Generator[Tuple[str, torch.Tensor], None, None]:
     """Iterate over the weights in the model bin/pt files."""
-    enable_tqdm = not torch.distributed.is_initialized(
-    ) or torch.distributed.get_rank() == 0
     for bin_file in tqdm(
             hf_weights_files,
             desc="Loading pt checkpoint shards",
-            disable=not enable_tqdm,
+            disable=not enable_tqdm(disable_progress_bar),
             bar_format=_BAR_FORMAT,
     ):
         state = torch.load(bin_file, map_location="cpu", weights_only=True)