Update app.py

app.py

@@ -8,6 +8,7 @@ import re
 import gc
 
 # 下載模型
+'''
 model_ids = [
     "OpenVINO/Qwen3-0.6B-int4-ov",
     "OpenVINO/Qwen3-1.7B-int4-ov",
@@ -26,14 +27,17 @@ for model_id in model_ids:
     except Exception as e:
       print(f"Error downloading {model_id}: {e}") # Handle download errors gracefully
 
+'''
+hf_hub.snapshot_download("hsuwill000/Llama-3.1-TAIDE-LX-8B-Chat_int4_ov", local_dir="ov", local_dir_use_symlinks=False)
+
 # 建立推理管線 (Initialize with a default model first)
 device = "CPU"
-InUsed_model_name = "Qwen3-0.6B-int4-ov"  # Choose a default model
+InUsed_model_name = "ov"  # Choose a default model
 pipe = ov_genai.LLMPipeline(InUsed_model_name, device)
 tokenizer = pipe.get_tokenizer()
 tokenizer.set_chat_template(tokenizer.chat_template)
 config = ov_genai.GenerationConfig()
-config.max_new_tokens = 1024
+config.max_new_tokens = 4096
 
 def streamer(subword):
     print(subword, end='', flush=True)