updated rag

utils/rag_chain.py

@@ -1,4 +1,4 @@
-# utils/rag_chain.py
+import os
 from langchain.vectorstores import FAISS
 from langchain.embeddings import HuggingFaceEmbeddings
 from langchain.text_splitter import RecursiveCharacterTextSplitter
@@ -16,12 +16,15 @@ class RAGChatbot:
         self.llm = self._load_llm()
 
     def _load_llm(self):
-        model_id = "meta-llama/Llama-4-Scout-17B-16E"
-        tokenizer = AutoTokenizer.from_pretrained(model_id)
+        model_id = "meta-llama/Meta-Llama-3-8B-Instruct"
+        token = os.getenv("HF_TOKEN")  # read the HF token from environment
+
+        tokenizer = AutoTokenizer.from_pretrained(model_id, token=token)
         model = AutoModelForCausalLM.from_pretrained(
             model_id,
             device_map="auto",
-            torch_dtype=torch.float16
+            torch_dtype=torch.float16,
+            token=token
         )
         pipe = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=512)
         return HuggingFacePipeline(pipeline=pipe)