enabled do_sample

app.py

@@ -5,8 +5,8 @@ import os
 import openai
 
 # Load the Vicuna 7B model and tokenizer
-vicuna_tokenizer = AutoTokenizer.from_pretrained("lmsys/vicuna-7b-v1.5")
-vicuna_model = AutoModelForCausalLM.from_pretrained("lmsys/vicuna-7b-v1.5")
+vicuna_tokenizer = AutoTokenizer.from_pretrained("lmsys/vicuna-7b-v1.3")
+vicuna_model = AutoModelForCausalLM.from_pretrained("lmsys/vicuna-7b-v1.3")
 
 # Load the LLaMA 7b model and tokenizer
 llama_tokenizer = AutoTokenizer.from_pretrained("daryl149/llama-2-7b-chat-hf")
@@ -60,7 +60,7 @@ def vicuna_respond(tab_name, message, chat_history):
     print('Prompt + Context:')
     print(formatted_prompt)
     input_ids = vicuna_tokenizer.encode(formatted_prompt, return_tensors="pt")
-    output_ids = vicuna_model.generate(input_ids, max_length=149, num_beams=5, no_repeat_ngram_size=2)
+    output_ids = vicuna_model.generate(input_ids, do_sample=True, max_length=149, num_beams=5, no_repeat_ngram_size=2)
     bot_message = vicuna_tokenizer.decode(output_ids[0], skip_special_tokens=True)
     print(bot_message)
         
@@ -73,7 +73,7 @@ def llama_respond(tab_name, message, chat_history):
     print('Prompt + Context:')
     print(formatted_prompt)
     input_ids = llama_tokenizer.encode(formatted_prompt, return_tensors="pt")
-    output_ids = llama_model.generate(input_ids, max_length=149, num_beams=5, no_repeat_ngram_size=2)
+    output_ids = llama_model.generate(input_ids, do_sample=True, max_length=149, num_beams=5, no_repeat_ngram_size=2)
     bot_message = llama_tokenizer.decode(output_ids[0], skip_special_tokens=True)
     print(bot_message)