Initial Qwen-VL demo

app.py

@@ -0,0 +1,18 @@
+import gradio as gr
+from transformers import AutoProcessor, AutoModelForVision2Seq
+from PIL import Image
+import torch
+
+model_id = "Qwen/Qwen-VL"
+processor = AutoProcessor.from_pretrained(model_id)
+model = AutoModelForVision2Seq.from_pretrained(model_id, torch_dtype=torch.float16, device_map="auto")
+
+def answer_question(image, question):
+    inputs = processor(images=image, text=question, return_tensors="pt").to("cuda")
+    generated_ids = model.generate(**inputs, max_new_tokens=50)
+    generated_text = processor.batch_decode(generated_ids, skip_special_tokens=True)[0]
+    return generated_text
+
+iface = gr.Interface(fn=answer_question, inputs=["image", "text"], outputs="text", title="Qwen-VL Demo")
+iface.launch()
+

requirements.txt

@@ -0,0 +1,5 @@
+transformers
+torch
+gradio
+accelerate
+