Update app.py

app.py

@@ -7,13 +7,7 @@ import os
 READ_HF=os.environ["read_hf"]
 
 from unsloth import FastLanguageModel
-model, tokenizer = FastLanguageModel.from_pretrained(
-    model_name = "VanguardAI/CoT_multi_llama_LoRA_4bit", # YOUR MODEL YOU USED FOR TRAINING
-    max_seq_length = 2048,
-    dtype = None,
-    load_in_4bit = True,
-    token= READ_HF    
-)
+
 
 
 
@@ -77,6 +71,13 @@ Category List : ["Dairy & Eggs", "Beverages & Snacks", "Cleaning & Hygiene", "Gr
 
 @spaces.GPU()
 def chunk_it(inventory_list, user_input_text):
+    model, tokenizer = FastLanguageModel.from_pretrained(
+    model_name = "VanguardAI/CoT_multi_llama_LoRA_4bit", # YOUR MODEL YOU USED FOR TRAINING
+    max_seq_length = 2048,
+    dtype = None,
+    load_in_4bit = True,
+    token= READ_HF    
+    )
     FastLanguageModel.for_inference(model) # Enable native 2x faster inference
     inputs = tokenizer(
         [