app.py

@@ -50,10 +50,11 @@ logger.debug("done load")
 # tokenizer = AutoTokenizer.from_pretrained("openchat/openchat_v2_w")
 # model = AutoModelForCausalLM.from_pretrained("openchat/openchat_v2_w", load_in_8bit_fp32_cpu_offload=True, load_in_8bit=True)
 
-model_path = model.config._dict["model_name_or_path"]
-logger.debug(f"{model.config=} {type(model.config)=} {model_path=}")
+# model_path = model.config._dict["model_name_or_path"]
+# logger.debug(f"{model.config=} {type(model.config)=} {model_path=}")
+logger.debug(f"{model.config=}, {type(model.config)=} ")
 
-model_size_gb = Path(model_path).stat().st_size / 2**30
+# model_size_gb = Path(model_path).stat().st_size / 2**30
 
 logger.info(f"{model_name=} {model_size_gb=:.2f} GB")