changed the llama model version

app.py

@@ -42,7 +42,7 @@ h1 {
 """
 
 # Load the tokenizer and model with quantization
-model_id = "meta-llama/Meta-Llama-3-8B-Instruct"
+model_id = "meta-llama/Meta-Llama-3.1-8B-Instruct"
 bnb_config = BitsAndBytesConfig(
     load_in_4bit=True,
     bnb_4bit_use_double_quant=True,

requirements.txt

@@ -1,4 +1,5 @@
 accelerate
-transformers
+torch
+transformers==4.43.1
 SentencePiece
 bitsandbytes