Upload inference.py

inference.py

@@ -0,0 +1,22 @@
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.vectorstores import FAISS
+from langchain.chains import RetrievalQA
+from transformers import pipeline
+import torch
+
+# Load model and tokenizer
+model_name = "PranavKeshav/upf-finetuned-model"
+hf_pipeline = pipeline("text-generation", model=model_name, device=0 if torch.cuda.is_available() else -1)
+llm = HuggingFacePipeline(pipeline=hf_pipeline)
+
+# Load FAISS store and embeddings
+vectorstore = FAISS.load_local("faiss_store", HuggingFaceEmbeddings())
+qa_chain = RetrievalQA.from_chain_type(
+    llm=llm,
+    chain_type="stuff",
+    retriever=vectorstore.as_retriever()
+)
+
+def handler(input_data):
+    query = input_data.get("query", "")
+    return qa_chain.run(query)