feat: Support VLM in reference_hf (#2726)

Signed-off-by: Ce Gao <gaocegege@hotmail.com>

scripts/playground/reference_hf.py

@@ -25,12 +25,89 @@ I'm going to the
 
 import argparse
 
+import requests
+from PIL import Image
+
 import torch
-from transformers import AutoModelForCausalLM
+from transformers import (
+    AutoModelForCausalLM, AutoProcessor, AutoModelForImageTextToText
+)
 
 from sglang.srt.hf_transformers_utils import get_tokenizer
 
 
+@torch.no_grad()
+def vlm_text_with_image(args):
+    # Load the processor and model for ImageTextToText tasks
+    processor = AutoProcessor.from_pretrained(
+        args.model_path, trust_remote_code=True)
+    model = AutoModelForImageTextToText.from_pretrained(
+        args.model_path,
+        torch_dtype=args.dtype,
+        low_cpu_mem_usage=True,
+        device_map="auto",
+        trust_remote_code=True,
+    )
+
+    torch.cuda.set_device(0)
+
+    # List of image URLs to process
+    image_urls = [
+        "https://github.com/haotian-liu/LLaVA/blob/1a91fc274d7c35a9b50b3cb29c4247ae5837ce39/images/llava_v1_5_radar.jpg?raw=true"
+    ]
+
+    # Conversation template for the processor
+    conversation = [
+        {
+            "role": "user",
+            "content": [
+                {
+                    "type": "image",
+                },
+                {
+                    "type": "text",
+                    "text": "Describe this image."
+                }
+            ]
+        }
+    ]
+
+    max_new_tokens = args.max_new_tokens
+
+    for i, url in enumerate(image_urls):
+        # Load the image from the URL
+        image = Image.open(requests.get(url, stream=True).raw)
+
+        # Apply the chat template to the text prompt
+        # Notice that not all processors support chat templates.
+        # LLaVA and QWen are two processors that support chat templates.
+        if not hasattr(processor, "apply_chat_template"):
+            raise ValueError("The processor does not support chat templates.")
+        text_prompt = processor.apply_chat_template(
+            conversation, add_generation_prompt=True)
+
+        # Prepare inputs for the model
+        inputs = processor(text=[text_prompt], images=[image],
+                           return_tensors="pt").to("cuda:0")
+
+        # Generate output from the model
+        output_ids = model.generate(
+            **inputs, do_sample=False, max_new_tokens=max_new_tokens
+        )
+        output_str = processor.decode(output_ids[0])
+
+        # Get the logits from the model's forward pass
+        outputs = model.forward(**inputs)
+        logits = outputs.logits[0, -1, :]
+
+        print(f"\n========== Image {i} ==========")
+        print("prefill logits (final)", logits)
+        # TODO(gaocegege): The output contains numerous <|image_pad|> tokens,
+        # making it cluttered and difficult to read.
+        # These tokens should be removed or cleaned up for better readability.
+        print(output_str)
+
+
 @torch.no_grad()
 def normal_text(args):
     t = get_tokenizer(args.model_path, trust_remote_code=True)
@@ -108,7 +185,11 @@ if __name__ == "__main__":
 
     parser.add_argument("--dtype", type=str, default="float16")
 
+    parser.add_argument("--model-type", type=str, default="text")
+
     args = parser.parse_args()
 
-    normal_text(args)
-    # synthetic_tokens(args)
+    if args.model_type == "vlm":
+        vlm_text_with_image(args)
+    else:
+        normal_text(args)