update/removed quantization

app.py

@@ -17,7 +17,7 @@ def load_model():
     
     model = AutoPeftModelForCausalLM.from_pretrained(
         lora_model_name, # YOUR MODEL YOU USED FOR TRAINING
-        load_in_4bit = True, # False
+        load_in_4bit = False, # False
     )
     tokenizer = AutoTokenizer.from_pretrained(lora_model_name)
     model.eval()

requirements.txt

@@ -3,4 +3,4 @@ gradio
 transformers
 peft
 torch
-bitsandbytes
+# bitsandbytes