del

app.py

@@ -2,20 +2,17 @@ import gradio as gr
 from unsloth import FastLanguageModel
 import torch
 
+max_seq_length = 2048 # Choose any! We auto support RoPE Scaling internally!
+dtype = None # None for auto detection. Float16 for Tesla T4, V100, Bfloat16 for Ampere+
+load_in_4bit = True # Use 4bit quantization to reduce memory usage. Can be False.
 
-# Load your model and tokenizer (make sure to adjust the path to where your model is stored)
-max_seq_length = 2048  # Adjust as necessary
-load_in_4bit = True  # Enable 4-bit quantization for reduced memory usage
-model_path = "llama_lora_model_1"  # Path to your custom model
-
-# Load the model and tokenizer
 model, tokenizer = FastLanguageModel.from_pretrained(
-    model_name=model_path,
-    max_seq_length=max_seq_length,
-    load_in_4bit=load_in_4bit,
+    model_name = "llama_lora_model_1",
+    max_seq_length = max_seq_length,
+    dtype = dtype,
+    load_in_4bit = load_in_4bit,
 )
 
-
 # Respond function
 def respond(
     message,