Update app.py

app.py

@@ -34,8 +34,8 @@ data.add_faiss_index("embeddings", custom_index=index)
 #question = "How can I reverse Diabetes?"
 
 SYS_PROMPT = """You are an assistant for answering questions.
-You are given the extracted parts of documents and a question. Provide a conversational answer.
-If you don't know the answer, just say "I do not know." Don't make up an answer. Don't repeat the SYS_PROMPT."""
+You are given the extracted parts of document, a question and history of questions and answers . Provide a conversational answer.
+If you do not know the answer, just say "I do not know." Do not make up an answer. Don't repeat the SYS_PROMPT."""
 # Provides context of how to answer the question
 
 #llm_model = "TheBloke/TinyLlama-1.1B-Chat-v1.0-GGUF", tinyllama-1.1b-chat-v1.0.Q5_K_M.gguf
@@ -47,7 +47,6 @@ model = Llama(
     n_gpu_layers = 0,
     temperature=0.75,
     n_ctx = 4096,
-    max_tokens=500,
     top_p=0.95 #,
  #   eos_tokens=terminators
     # callback_manager=callback_manager,
@@ -118,9 +117,10 @@ def talk(prompt, history):
     # the chat template structure should be based on text generation model format
 
 # indicates the end of a sequence
-    stream = model.create_chat_completion(messages = messages, max_tokens=1000, stop=["</s>"], stream=False)
+    stream = model.create_chat_completion(messages = messages, max_tokens=600, stop=["</s>"], stream=False)
 #    print(f"{stream}")
     print("check 7")
+    global historylog
     print(historylog)
     print(stream['choices'][0]['message']['content'])
     return(stream['choices'][0]['message']['content'])