Update app.py

app.py

@@ -1,26 +1,22 @@
+
 from fastapi import FastAPI, File, UploadFile
 from fastapi.responses import JSONResponse
-from transformers import AutoTokenizer, AutoProcessor, AutoModelForCausalLM
+from transformers import AutoProcessor, AutoModelForCausalLM
 import torch
 from PIL import Image
 import io
+import os
 
 app = FastAPI()
 
-model_path = "lmms-lab/LLaVA-OneVision-1.5-8B-Instruct"
-os.environ["HF_HOME"] = "./huggingface_cache" 
-
-model = AutoModelForCausalLM.from_pretrained(model_path, torch_dtype="auto", device_map="auto", trust_remote_code=True)
-processor = AutoProcessor.from_pretrained(model_path, trust_remote_code=True)
-
-def process_vision_info(messages):
-    # Dummy implementation, replace with actual from qwen_vl_utils
-    image_inputs = [msg['content'][0]['image'] for msg in messages]
-    video_inputs = None
-    return image_inputs, video_inputs
+MODEL_NAME = os.getenv("MODEL_NAME", "lmms-lab/LLaVA-OneVision-1.5-8B-Instruct")
+model = AutoModelForCausalLM.from_pretrained(
+    MODEL_NAME, torch_dtype="auto", device_map="auto", trust_remote_code=True
+)
+processor = AutoProcessor.from_pretrained(MODEL_NAME, trust_remote_code=True)
 
 @app.post("/analyze-image")
-async def analyze_image(file: UploadFile = File(...)):
+async def analyze_image(file: UploadFile = File(...), prompt: str = "Describe this image."):
     image_bytes = await file.read()
     image = Image.open(io.BytesIO(image_bytes))
     messages = [
@@ -28,20 +24,18 @@ async def analyze_image(file: UploadFile = File(...)):
             "role": "user",
             "content": [
                 {"type": "image", "image": image},
-                {"type": "text", "text": "Describe this image."},
+                {"type": "text", "text": prompt},
             ],
         }
     ]
     text = processor.apply_chat_template(messages, tokenize=False, add_generation_prompt=True)
-    image_inputs, video_inputs = process_vision_info(messages)
     inputs = processor(
         text=[text],
-        images=image_inputs,
-        videos=video_inputs,
+        images=[image],
         padding=True,
         return_tensors="pt",
     )
-    inputs = inputs.to("cuda")
+    inputs = inputs.to(model.device)
     generated_ids = model.generate(**inputs, max_new_tokens=1024)
     generated_ids_trimmed = [
         out_ids[len(in_ids):] for in_ids, out_ids in zip(inputs.input_ids, generated_ids)
@@ -49,4 +43,4 @@ async def analyze_image(file: UploadFile = File(...)):
     output_text = processor.batch_decode(
         generated_ids_trimmed, skip_special_tokens=True, clean_up_tokenization_spaces=False
     )
-    return JSONResponse(content={"result": output_text})
+    return JSONResponse(content={"result": output_text[0]})