update demo

Dockerfile

@@ -67,5 +67,5 @@ RUN pip install -r requirements.txt
 
 RUN ls -lh
 
-CMD ["bash", "start.sh"]
-# CMD ["python", 'start.py']
+# CMD ["bash", "start.sh"]
+CMD ["python", 'start.py']

start.py

@@ -1,7 +1,7 @@
 import subprocess
 
 if __name__ == '__main__':
-    backend_comand = ['python3', 'gradio_demo/seed_llama_flask.py', '--image_transform', 'configs/transform/clip_transform.yaml', '--tokenizer', 'configs/tokenizer/seed_llama_tokenizer.yaml', '--model', 'configs/llm/seed_llama_14b_8bit.yaml', '--port', '7890', '--llm_device', 'cuda:0', '--tokenizer_device', 'cuda:0', '--offload_encoder', '--offload_decoder']
+    backend_comand = ['python3', 'gradio_demo/seed_llama_flask.py', '--image_transform', 'configs/transform/clip_transform.yaml', '--tokenizer', 'configs/tokenizer/seed_llama_tokenizer_hf.yaml', '--model', 'configs/llm/seed_llama_14b_8bit.yaml', '--port', '7890', '--llm_device', 'cuda:0', '--tokenizer_device', 'cuda:0', '--offload_encoder', '--offload_decoder']
     
     frontend_comand = ['python3', 'gradio_demo/seed_llama_gradio.py', '--server_port', '7860', '--request_address', 'http://127.0.0.1:7890/generate', '--model_type', 'seed-llama-14b']