Upload infer_qwen2_vl.py with huggingface_hub

infer_qwen2_vl.py

@@ -71,9 +71,9 @@ for batch_idx in tqdm(range(begin, end, batch_size)):
     print(f"data index range : {batch_idx} ~ {min(batch_idx + batch_size, end)}")
     image_inputs_batch, video_inputs_batch = [], []
     for idx,i in enumerate(batch):
-        img = batch[i]
+        #img = batch[i]
         print('gain image successfully !')
-        messages[1]["content"][0]["image"] = img
+        messages[1]["content"][0]["image"] = i
         text = processor.apply_chat_template(
             messages, tokenize=False, add_generation_prompt=True
         )