Update app_v3.py

app_v3.py

@@ -1,8 +1,21 @@
 from transformers import AutoTokenizer, TextStreamer, pipeline
 from auto_gptq import AutoGPTQForCausalLM, BaseQuantizeConfig
 import time
+from huggingface_hub import snapshot_download
+import os
 
-model_name_or_path = "FPHam/Jackson_The_Formalizer_V2_13b_GPTQ"
+# Define pretrained and quantized model directories
+pretrained_model_dir = "FPHam/Jackson_The_Formalizer_V2_13b_GPTQ"
+cwd = os.getcwd()
+
+quantized_model_dir = cwd + "/Jackson2-4bit-128g-GPTQ"
+
+# Create the cache directory if it doesn't exist
+os.makedirs(quantized_model_dir, exist_ok=True)
+
+snapshot_download(repo_id=pretrained_model_dir, local_dir=quantized_model_dir, local_dir_use_symlinks=True)
+
+model_name_or_path = quantized_model_dir
 model_basename = "Jackson2-4bit-128g-GPTQg"
 
 use_triton = False