Update app/llm.py

app/llm.py

@@ -32,16 +32,16 @@ class ChatModel(BaseModel):
 llm_chat = llama_cpp.Llama.from_pretrained(
     repo_id="moriire/healthcare-ai-q8_0",
     filename="healthcare-ai-q8_0-unsloth.Q8_0.gguf",
-    #tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained("moriire/healthcare-ai-adapter-merged"),
+    tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained("moriire/healthcare-ai-q8_0"),
     verbose=False,
      n_ctx=1024,
      n_gpu_layers=0,
-    chat_format="llama-2"
+    #chat_format="llama-2"
 )
 llm_generate = llama_cpp.Llama.from_pretrained(
     repo_id="moriire/healthcare-ai-q8_0",
     filename="healthcare-ai-q8_0-unsloth.Q8_0.gguf",
-    #tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained("moriire/healthcare-ai-adapter-merged"),
+    #tokenizer=llama_cpp.llama_tokenizer.LlamaHFTokenizer.from_pretrained("moriire/healthcare-ai-q8_0"),
     verbose=False,
      n_ctx=4096,
      n_gpu_layers=0,