Update app.py

app.py

@@ -49,7 +49,6 @@ If you don't know the answer, just say "I do not know." Don't make up an answer.
 # Provides context of how to answer the question
 
 llm_model = "TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF"
-# TheBloke/Llama-2-7B-Chat-GGML , TinyLlama/TinyLlama-1.1B-Chat-v1.0 , microsoft/Phi-3-mini-4k-instruct, health360/Healix-1.1B-V1-Chat-dDPO
 # TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF and tinyllama-1.1b-chat-v1.0.Q4_K_M.gguf not working, TinyLlama/TinyLlama-1.1B-Chat-v0.6, andrijdavid/TinyLlama-1.1B-Chat-v1.0-GGUF"
 
 tokenizer = AutoTokenizer.from_pretrained("TinyLlama/TinyLlama-1.1B-Chat-v1.0")
@@ -138,19 +137,17 @@ def talk(prompt, history):
     # the chat template structure should be based on text generation model format
     print("check6")    
 
-    terminators = [
-      tokenizer.eos_token_id, # End-of-Sequence Token that indicates where the model should consider the text sequence to be complete
-      tokenizer.convert_tokens_to_ids("<|eot_id|>") # Converts a token strings in a single/ sequence of integer id using the vocabulary
-    ]
 # indicates the end of a sequence
     import pprint
     stream = model.create_chat_completion(messages = [{"role":"system","content":SYS_PROMPT},{"role":"user","content":formatted_prompt}], max_tokens=1000, stop=["</s>"], stream=True)
 #   print(output['choices'][0]['message']['content'])
+    print(f"{stream}")
     pprint.pprint(stream)
     text = []
     for output in stream:
 #       text += output['choices'][0]   
         text.append(output['choices'][0])
+        print(f"{text}")
         yield "".join(text)
         print(text)
         print("check3H")