Update app.py

app.py

@@ -1,7 +1,7 @@
 import re
 import spaces
 import gradio as gr
-from transformers import AutoTokenizer, AutoModel
+from transformers import AutoTokenizer, AutoModel, BitsAndBytesConfig
 import torch
 # from peft import PeftModel, PeftConfig
 
@@ -36,13 +36,20 @@ import torch
 
 # model = AutoModel.from_pretrained("FlawedLLM/Bhashini", load_in_4bit=True, device_map='auto')
     # I highly do NOT suggest - use Unsloth if possible
-from peft import AutoPeftModelForCausalLM
-from transformers import AutoTokenizer
-model = AutoPeftModelForCausalLM.from_pretrained(
-        "FlawedLLM/Bhashini", # YOUR MODEL YOU USED FOR TRAINING
-        load_in_4bit = True,
-    )
-tokenizer = AutoTokenizer.from_pretrained("FlawedLLM/Bhashini")
+# from peft import AutoPeftModelForCausalLM
+# from transformers import AutoTokenizer
+# model = AutoPeftModelForCausalLM.from_pretrained(
+#         "FlawedLLM/Bhashini", # YOUR MODEL YOU USED FOR TRAINING
+#         load_in_4bit = True,
+#     )
+# tokenizer = AutoTokenizer.from_pretrained("FlawedLLM/Bhashini")
+from unsloth import FastLanguageModel
+model, tokenizer = FastLanguageModel.from_pretrained(
+        model_name = "BhashiniLLama", # YOUR MODEL YOU USED FOR TRAINING
+        max_seq_length = max_seq_length,
+        dtype = dtype,
+        load_in_4bit = load_in_4bit,)
+FastLanguageModel.for_inference(model)
 
 @spaces.GPU(duration=300)
 def chunk_it(input_command):