config_device=meta

app.py

@@ -14,6 +14,7 @@ from transformers import AutoModelForCausalLM, AutoTokenizer, StoppingCriteria,
 
 
 import config
+config.init_device="meta"
 
 INSTRUCTION_KEY = "### Instruction:"
 RESPONSE_KEY = "### Response:"
@@ -116,7 +117,7 @@ print (f"sys.path : {sys.path}")
 
 
 def complete(state="complete"):   
-    print(f"\nCell {state} @ {(datetime.fromtimestamp(time.time()).strftime('%Y-%m-%d %H:%M:%S'))}")
+    print(f"\nCell {state}")
     
 complete(state='imports done')
 
@@ -193,7 +194,6 @@ def tester(uPrompt, max_new_tokens, temperature, top_k, top_p):
     results = f"{salutation} max_new_tokens{max_new_tokens}; temperature{temperature}; top_k{top_k}; top_p{top_p};  "
     
     return response
-config.init_device="meta"
 demo = gr.Interface(
     fn=tester,
     inputs=[gr.Textbox(label="Prompt",info="Prompt",lines=3,value="Provide Prompt"),