removed gpu flag

utils.py

@@ -39,7 +39,7 @@ sys.path.append('/mount/src/gen_ai_dev')
 DATA_PATH = "./Infy financial report/"
 DATA_FILES = ["INFY_2022_2023.pdf", "INFY_2023_2024.pdf"]
 EMBEDDING_MODEL = "sentence-transformers/all-MiniLM-L6-v2"
-LLM_MODEL = "microsoft/phi-1_5" #"gpt2"  # Or "distilgpt2"  # Or "HuggingFaceH4/zephyr-7b-beta" or "microsoft/phi-2"
+LLM_MODEL = "gpt2"  # Or "distilgpt2"  # Or "HuggingFaceH4/zephyr-7b-beta" or "microsoft/phi-2"
 
 # Environment settings
 os.environ["TOKENIZERS_PARALLELISM"] = "false"
@@ -220,9 +220,8 @@ try:
         tokenizer = AutoTokenizer.from_pretrained(LLM_MODEL)
         model = AutoModelForCausalLM.from_pretrained(
             LLM_MODEL,
-            device_map="auto",
+            device_map="cpu",
             torch_dtype=torch.float16,
-            load_in_4bit=True
         )
         return pipeline(
             "text-generation",
@@ -232,7 +231,7 @@ try:
             do_sample=False,
             temperature=0.7,
             top_k=0,
-            top_p=1,
+            top_p=1
         )
     generator = load_generator()
 except Exception as e: