Update app.py

app.py

@@ -3,6 +3,9 @@ import torch
 import transformers
 import gradio as gr 
 from huggingface_hub import hf_hub_download
+
+from huggingface_hub import snapshot_download
+
 import safetensors
 
 # from transformer_engine.pytorch import fp8_autocast
@@ -59,10 +62,15 @@ files_to_download = [
 model_dir = f"./{model_id}"
 os.makedirs(model_dir, exist_ok=True)
 
+snapshot_download(repo_id=model_id, ignore_patterns=".bin", token=auth_token)
+
 
+'''
 # Download each file
 for file in files_to_download:
     hf_hub_download(repo_id=model_id, filename=file, local_dir=model_dir, token=auth_token)
+''' 
+    
 
 '''
 with fp8_autocast():  # Enables FP8 computations
@@ -79,7 +87,8 @@ with fp8_autocast():  # Enables FP8 computations
 model = transformers.AutoModelForCausalLM.from_pretrained(model_dir, quantization_config=quantization_config)
 tokenizer = transformers.AutoTokenizer.from_pretrained(model_dir)
 
-'''model.to(dtype=torch.float16)  # Load as FP16 first
+'''
+model.to(dtype=torch.float16)  # Load as FP16 first
 model = model.half()  # Convert to FP8-like (closest possible)
 '''