Upload 2 files

app.py

@@ -0,0 +1,92 @@
+import os
+import gradio as gr
+from transformers.pipelines import pipeline
+from langchain.document_loaders import PyPDFLoader
+from langchain.text_splitter import RecursiveCharacterTextSplitter
+from langchain.embeddings import HuggingFaceEmbeddings
+from langchain.vectorstores import FAISS
+from langchain.chains import RetrievalQA
+from langchain_community.llms import HuggingFacePipeline
+
+# ✅ Load Hugging Face LLM (LLama 3 fine-tuned model)
+llm_pipeline = pipeline("text2text-generation", model="BivinSadler/llama3-finetuned-Statistics", max_length=512)
+llm = HuggingFacePipeline(pipeline=llm_pipeline)
+
+# ✅ Create embeddings for RAG
+embedding_model = HuggingFaceEmbeddings(model_name="sentence-transformers/all-MiniLM-L6-v2")
+
+# ✅ Build RAG Agent Function
+def build_rag_agent(pdf_path):
+    loader = PyPDFLoader(pdf_path)
+    docs = loader.load()
+    splitter = RecursiveCharacterTextSplitter(chunk_size=500, chunk_overlap=50)
+    chunks = splitter.split_documents(docs)
+    vectorstore = FAISS.from_documents(chunks, embedding_model)
+    retriever = vectorstore.as_retriever()
+    return RetrievalQA.from_chain_type(llm=llm, retriever=retriever, chain_type="stuff")
+
+# ✅ Create RAG agents for both syllabi
+stat6371_agent = build_rag_agent("PDFs/DS 6371 Syllabus Ver 6.pdf")
+ds7333_agent = build_rag_agent("PDFs/ds-7333_syllabus.pdf")
+
+# ✅ Writer Agent (makes answers easier)
+def writer_agent(raw_answer, audience="high school students"):
+    prompt = f"""
+You are a skilled teacher explaining to {audience}. Simplify the following answer in 2–3 short, clear sentences:
+
+Answer:
+{raw_answer}
+"""
+    result = llm_pipeline(prompt, max_length=200, do_sample=False)
+    return result[0]['generated_text']
+
+# ✅ Question Routing Agent (classifies the question)
+def route_question(question):
+    routing_prompt = f"""
+You are a routing agent. Classify the question into one of:
+A. Stat 6371
+B. DS 7333
+C. General statistics
+
+Question: "{question}"
+Answer with only A, B, or C.
+"""
+    result = llm_pipeline(routing_prompt, max_length=10, do_sample=False)
+    route = result[0]['generated_text'].strip().upper()
+    if route.startswith("A"):
+        return "stat6371"
+    elif route.startswith("B"):
+        return "ds7333"
+    else:
+        return "general"
+
+# ✅ Multi-Agent Pipeline
+def multiagent_system(question):
+    print(f"\n🧭 Routing: {question}")
+    route = route_question(question)
+
+    if route == "stat6371":
+        print("🔎 Stat 6371 Agent")
+        raw_answer = stat6371_agent.run(question)
+    elif route == "ds7333":
+        print("🔎 DS 7333 Agent")
+        raw_answer = ds7333_agent.run(question)
+    else:
+        print("🧠 General Stats HF Agent")
+        result = llm_pipeline(question, max_length=200, do_sample=False)
+        raw_answer = result[0]['generated_text']
+
+    print("✍️ Simplifying...")
+    return writer_agent(raw_answer)
+
+# ✅ Gradio UI
+iface = gr.Interface(
+    fn=multiagent_system,
+    inputs=gr.Textbox(lines=2, label="Ask a statistics question"),
+    outputs=gr.Textbox(label="Answer"),
+    title="📊 Multi-Agent Statistics Assistant (HuggingFace)",
+    description="Routes your stats question to the right syllabus (Stat 6371, DS 7333) or uses a general statistics model (LLama3)."
+)
+
+if __name__ == "__main__":
+    iface.launch()

requirements.txt

@@ -0,0 +1,11 @@
+gradio
+faiss-cpu
+PyPDF2
+pypdf
+transformers
+sentence-transformers
+huggingface-hub
+langchain
+langchain-community
+
+