Update app.py

app.py

@@ -62,7 +62,17 @@ from transformers import AutoTokenizer, AutoModelForCausalLM, BitsAndBytesConfig
 # Load model directly
 
 tokenizer = AutoTokenizer.from_pretrained("FlawedLLM/Bhashini_00")
-model = AutoModelForCausalLM.from_pretrained("FlawedLLM/Bhashini_00", load_in_4bit=True)
+quantization_config = BitsAndBytesConfig(
+        load_in_4bit=True,
+        bnb_4bit_use_double_quant=True,
+        bnb_4bit_quant_type="nf4",  
+        bnb_4bit_compute_dtype=torch.float16)
+model = AutoModelForCausalLM.from_pretrained("FlawedLLM/Bhashini_00", 
+                                             device_map="auto",
+                                             quantization_config=quantization_config, 
+                                             torch_dtype =torch.float16, 
+                                             low_cpu_mem_usage=True, 
+                                             use_safetensors=True,)
 
 @spaces.GPU(duration=300)
 def chunk_it(input_command, item_list):
@@ -77,7 +87,7 @@ def chunk_it(input_command, item_list):
     ### Response:
     {}"""
     if item_list is not None:
-        item_list = f"The ItemName should be STRICTLY chosen from the given list of ItemNames, : {item_list} , except when adding item. Try to be as strict as possible, if item name not available, then write null."
+        item_list = f'''The ItemName should be STRICTLY chosen from the given list : {item_list} , except when adding item. Try to be as strict as possible, if ItemName does not find anything similar in the list, then the ItemName should be "Null" '''
     inputs = tokenizer(
     [
         alpaca_prompt.format(