Update app.py

app.py

@@ -2,9 +2,12 @@ import gradio as gr
 import torch
 import re
 from transformers import AutoTokenizer, AutoModelForCausalLM
+from peft import PeftModel
+
 
 # Model configuration
-MODEL_NAME = "jmcinern/qomhra-8B-awq-dpo-beta-0.5-checkpoint-checkpoint-100"
+MODEL_NAME = "jmcinern/qwen3-8B-cpt-sft-awq"
+DPO_ADAPTER = "jmcinern/qomhra-8B-awq-dpo-beta-0.5-checkpoint-checkpoint-100"
 THINK_TAG_PATTERN = re.compile(r"<think>.*?</think>\s*", flags=re.DOTALL)
 
 
@@ -25,13 +28,17 @@ class ChatBot:
             print("Tokenizer loaded!")
 
             print("Loading model...")
-            self.model = AutoModelForCausalLM.from_pretrained(
+            base_model = AutoModelForCausalLM.from_pretrained(
                 MODEL_NAME,
                 trust_remote_code=True,
                 device_map="auto",
                 torch_dtype="auto",
                 low_cpu_mem_usage=True,
             )
+            self.model = PeftModel.from_pretrained(
+                base_model,
+                DPO_ADAPTER
+            )
             print("Model loaded!")
         except Exception as e:
             print(f"Error loading model: {e}")