[Refactor][Misc] Remove unused code in diffusion hooks/model_loader (#5270)

Signed-off-by: Alicia <115451386+congw729@users.noreply.github.com>

vllm_omni/diffusion/model_loader/diffusers_loader.py

@@ -12,7 +12,6 @@ from typing import cast
 
 import torch
 from torch import nn
-from vllm.config import ModelConfig
 from vllm.config.load import LoadConfig
 from vllm.logger import init_logger
 from vllm.model_executor.layers.quantization.base_config import QuantizeMethodBase
@@ -81,7 +80,6 @@ def _natural_sort_key(filepath: str) -> list:
     return [int(s) if s.isdigit() else s for s in re.split(r"(\d+)", os.path.basename(filepath))]
 
 
-MODEL_INDEX = "model_index.json"
 DIFFUSION_MODEL_WEIGHTS_INDEX = "diffusion_pytorch_model.safetensors.index.json"
 TRANSFORMER_WEIGHTS_INDEX = "model.safetensors.index.json"
 INDEX_FILES = [DIFFUSION_MODEL_WEIGHTS_INDEX, TRANSFORMER_WEIGHTS_INDEX]
@@ -326,15 +324,6 @@ class DiffusersPipelineLoader:
             return all_parameter_names
         return {name for name in all_parameter_names if name.startswith(source_prefixes)}
 
-    def download_model(self, model_config: ModelConfig) -> None:
-        self._prepare_weights(
-            model_name_or_path=model_config.model,
-            subfolder=None,
-            revision=model_config.revision,
-            fall_back_to_pt=True,
-            allow_patterns_overrides=None,
-        )
-
     def load_model(
         self,
         load_device: str,

vllm_omni/diffusion/model_loader/hub_prefetch.py

@@ -433,10 +433,7 @@ def from_pretrained_with_prefetch(
     ``factory`` is a bound ``SomeModel.from_pretrained`` (or any callable with
     the same ``(model, *, subfolder, local_files_only, **kwargs)`` signature).
 
-    This is a stronger sibling of :func:`retry_on_missing_shard`: that helper
-    only retries the missing-shard ``OSError`` and never re-prefetches, so it
-    cannot recover the second face of the same race. Two shapes of partial
-    -cache failure crash the diffusion server outright:
+    Two shapes of partial-cache failure crash the diffusion server outright:
 
     * ``OSError: <repo> does not appear to have a file named
       text_encoder/model-0000X-of-0000Y.safetensors`` - a shard is still under
@@ -488,31 +485,3 @@ def from_pretrained_with_prefetch(
 
     assert last_exc is not None  # loop only exits via return or a caught exc
     raise last_exc
-
-
-def retry_on_missing_shard(load_fn, *, max_retries: int = 3, base_delay: float = 5.0):
-    """Call *load_fn* with retry on the transformers v5 shard-resolution race.
-
-    When the prefetch lock cannot be acquired (e.g. flock unsupported on
-    the filesystem and dotfile lock times out), ``from_pretrained`` may
-    still hit the ``cached_files`` race. This wrapper retries with
-    exponential backoff when the OSError message matches the specific
-    "does not appear to have a file named" pattern.
-    """
-    for attempt in range(max_retries):
-        try:
-            return load_fn()
-        except OSError as exc:
-            if "does not appear to have a file" not in str(exc):
-                raise
-            if attempt == max_retries - 1:
-                raise
-            delay = base_delay * (attempt + 1)
-            logger.warning(
-                "from_pretrained failed with shard-resolution race (%s); retrying in %.1fs (attempt %d/%d)",
-                exc,
-                delay,
-                attempt + 1,
-                max_retries,
-            )
-            time.sleep(delay)