Update app.py

app.py

@@ -1,5 +1,5 @@
 import gradio as gr
-import torch
+import torchdo_s
 import os
 import time
 
@@ -110,7 +110,7 @@ def predict_chat(message: str, history: list):
                 temperature=TEMPERATURE,
                 top_k=TOP_K,
                 top_p=TOP_P,
-                do_sample=DO_SAMPLE, # Corrected parameter passing
+                #do_sample=DO_SAMPLE, # Corrected parameter passing
                 repetition_penalty=1.1,
                 stop=["User:", "\nUser", "\n#", "\n##", "<|endoftext|>"],
                 stream=True
@@ -127,7 +127,7 @@ def predict_chat(message: str, history: list):
                 temperature=TEMPERATURE,
                 top_k=TOP_K,
                 top_p=TOP_P,
-                do_sample=DO_SAMPLE, # Corrected parameter passing
+                #do_sample=DO_SAMPLE, # Corrected parameter passing
                 repetition_penalty=1.1,
                 stop=["User:", "\nUser", "\n#", "\n##", "<|endoftext|>"]
             )
@@ -153,7 +153,7 @@ def predict_chat(message: str, history: list):
             temperature=TEMPERATURE,
             top_k=TOP_K,
             top_p=TOP_P,
-            do_sample=DO_SAMPLE, # Uncommented for use
+            #do_sample=DO_SAMPLE, # Uncommented for use
             pad_token_id=tokenizer.pad_token_id
         )
         generated_text = tokenizer.decode(outputs[0][inputs.shape[-1]:], skip_special_tokens=True).strip()