fix

app.py

@@ -42,13 +42,12 @@ if 'int4' in model_path:
         exit()
     model = AutoModel.from_pretrained(model_path, trust_remote_code=True)
 else:
-    if True: #args.multi_gpus:
+    if False: #args.multi_gpus:
         from accelerate import load_checkpoint_and_dispatch, init_empty_weights, infer_auto_device_map
         with init_empty_weights():
             #model = AutoModel.from_pretrained(model_path, trust_remote_code=True, attn_implementation='sdpa', torch_dtype=torch.bfloat16)
             model = AutoModel.from_pretrained(model_path, trust_remote_code=True, torch_dtype=torch.bfloat16)
-        #device_map = infer_auto_device_map(model, max_memory={0: "10GB", 1: "10GB"},
-        device_map = infer_auto_device_map(model,
+        device_map = infer_auto_device_map(model, max_memory={0: "10GB", 1: "10GB"},
             no_split_module_classes=['SiglipVisionTransformer', 'Qwen2DecoderLayer'])
         device_id = device_map["llm.model.embed_tokens"]
         device_map["llm.lm_head"] = device_id # firtt and last layer should be in same device