Update ME_LLAMA_MODEL to use the 13b variant for improved performance

medbot/config.py

@@ -1,2 +1,2 @@
-ME_LLAMA_MODEL = "clinicalnlplab/me-llama"
+ME_LLAMA_MODEL = "clinicalnlplab/me-llama-13b"
 FALLBACK_MODEL = "meta-llama/Llama-2-7b-chat-hf" 

medbot/model.py

@@ -1,6 +1,7 @@
 import torch
 from transformers import AutoModelForCausalLM, AutoTokenizer
 from .config import ME_LLAMA_MODEL, FALLBACK_MODEL
+import logging
 
 class ModelManager:
     def __init__(self):
@@ -11,6 +12,7 @@ class ModelManager:
         if self.model is not None and self.tokenizer is not None:
             return
         try:
+            logging.info(f"ME_LLAMA_MODEL type: {type(ME_LLAMA_MODEL)}, value: {ME_LLAMA_MODEL}")
             self.tokenizer = AutoTokenizer.from_pretrained(ME_LLAMA_MODEL, trust_remote_code=True)
             self.model = AutoModelForCausalLM.from_pretrained(
                 ME_LLAMA_MODEL,