wip

app.py

@@ -2,15 +2,16 @@ import gradio as gr
 from transformers import AutoProcessor, Idefics3ForConditionalGeneration, image_utils
 import torch
 device = torch.device('cuda') if torch.cuda.is_available() else torch.device('cpu')
+print(f"Using device: {device}")
 model_id="eltorio/IDEFICS3_ROCO"
 # model = AutoModelForImageTextToText.from_pretrained(model_id).to(device)
 base_model_path="HuggingFaceM4/Idefics3-8B-Llama3" #or change to local path
-processor = AutoProcessor.from_pretrained(base_model_path)
+processor = AutoProcessor.from_pretrained(base_model_path, trust_remote_code=True)
 model = Idefics3ForConditionalGeneration.from_pretrained(
         base_model_path, torch_dtype=torch.bfloat16
     ).to(device)
 
-model.load_adapter(model_id)
+model.load_adapter(model_id,device_map="auto")
 
 def infere(image):
     messages = [
@@ -24,10 +25,11 @@ def infere(image):
     ]
     prompt = processor.apply_chat_template(messages, add_generation_prompt=True)
     inputs = processor(text=prompt, images=[image], return_tensors="pt")
+    print(f"inputs: {inputs}")
     inputs = {k: v.to(device) for k, v in inputs.items()}
     generated_ids = model.generate(**inputs, max_new_tokens=8192)
     generated_texts = processor.batch_decode(generated_ids, skip_special_tokens=True)
     return generated_texts
 
-demo = gr.Interface(fn=infere, inputs="image", outputs="text")
-demo.launch()
+radiotest = gr.Interface(fn=infere, inputs="image", outputs="text")
+radiotest.launch()