fix: Use iterator as not to store all the file loads in memory at once (#36149)

Signed-off-by: Shaun Kotek - Nvidia <skotek@nvidia.com>

vllm/model_executor/model_loader/weight_utils.py

@@ -773,7 +773,9 @@ def multi_thread_safetensors_weights_iterator(
         return result
 
     with concurrent.futures.ThreadPoolExecutor(max_workers=max_workers) as executor:
-        futures = [executor.submit(_load_file, st_file) for st_file in hf_weights_files]
+        # Note to use generator here so we do not store all the loaded files in memory
+        # at the same time, which can cause OOM for large models.
+        futures = (executor.submit(_load_file, st_file) for st_file in hf_weights_files)
         futures_iter = tqdm(
             concurrent.futures.as_completed(futures),
             total=len(hf_weights_files),
@@ -784,7 +786,9 @@ def multi_thread_safetensors_weights_iterator(
 
         for future in futures_iter:
             state_dict = future.result()
-            yield from state_dict.items()
+            del future
+            for key in list(state_dict):
+                yield key, state_dict.pop(key)
 
 
 def runai_safetensors_weights_iterator(