Init simple wiki RAG

README.md

@@ -1,14 +1,45 @@
+# Mini RAG Demo – Retrieval-Augmented Generation on Wikipedia
+
+This is a lightweight Retrieval-Augmented Generation (RAG) app built with Gradio. It combines semantic search over a mini Wikipedia (`rag-datasets/rag-mini-wikipedia`) corpus with reranking and language generation to answer natural language questions using real documents.
+
 ---
-title: RAGTesting
-emoji: 💬
-colorFrom: yellow
-colorTo: purple
-sdk: gradio
-sdk_version: 5.0.1
-app_file: app.py
-pinned: false
-license: mit
-short_description: A simple RAG demo
+
+## What It Does
+
+- Embeds a query using a SentenceTransformer (`all-MiniLM-L6-v2`)
+- Retrieves the top-5 most semantically similar Wikipedia passages using FAISS
+- Reranks them using a CrossEncoder model (`cross-encoder/ms-marco-MiniLM-L-6-v2`)
+- Generates an answer using a Hugging Face language model
+
 ---
 
-An example chatbot using [Gradio](https://gradio.app), [`huggingface_hub`](https://huggingface.co/docs/huggingface_hub/v0.22.2/en/index), and the [Hugging Face Inference API](https://huggingface.co/docs/api-inference/index).
+## Tech Stack
+
+- **Gradio** – Web interface
+- **FAISS** – Fast dense vector retrieval
+- **Sentence-Transformers** – Embedding & reranking
+- **Transformers (Hugging Face)** – Language model for generation
+- **Hugging Face Datasets** – Mini Wikipedia corpus (`rag-datasets/rag-mini-wikipedia`)
+
+---
+
+## Models Used
+
+| Purpose       | Model                                       |
+|---------------|---------------------------------------------|
+| Embedding     | `all-MiniLM-L6-v2`                          |
+| Reranking     | `cross-encoder/ms-marco-MiniLM-L-6-v2`      |
+| Generation    | `mistralai/Mistral-7B-Instruct-v0.2` *(optional)* or a smaller model |
+
+---
+
+## 📦 Running Locally
+
+To run the app locally:
+
+```bash
+git clone https://huggingface.co/spaces/YOUR_USERNAME/mini-rag-demo
+cd mini-rag-demo
+pip install -r requirements.txt
+python app.py
+```

app.py

@@ -1,64 +1,60 @@
 import gradio as gr
-from huggingface_hub import InferenceClient
-
-"""
-For more information on `huggingface_hub` Inference API support, please check the docs: https://huggingface.co/docs/huggingface_hub/v0.22.2/en/guides/inference
-"""
-client = InferenceClient("HuggingFaceH4/zephyr-7b-beta")
-
-
-def respond(
-    message,
-    history: list[tuple[str, str]],
-    system_message,
-    max_tokens,
-    temperature,
-    top_p,
-):
-    messages = [{"role": "system", "content": system_message}]
-
-    for val in history:
-        if val[0]:
-            messages.append({"role": "user", "content": val[0]})
-        if val[1]:
-            messages.append({"role": "assistant", "content": val[1]})
-
-    messages.append({"role": "user", "content": message})
-
-    response = ""
-
-    for message in client.chat_completion(
-        messages,
-        max_tokens=max_tokens,
-        stream=True,
-        temperature=temperature,
-        top_p=top_p,
-    ):
-        token = message.choices[0].delta.content
-
-        response += token
-        yield response
-
-
-"""
-For information on how to customize the ChatInterface, peruse the gradio docs: https://www.gradio.app/docs/chatinterface
-"""
-demo = gr.ChatInterface(
-    respond,
-    additional_inputs=[
-        gr.Textbox(value="You are a friendly Chatbot.", label="System message"),
-        gr.Slider(minimum=1, maximum=2048, value=512, step=1, label="Max new tokens"),
-        gr.Slider(minimum=0.1, maximum=4.0, value=0.7, step=0.1, label="Temperature"),
-        gr.Slider(
-            minimum=0.1,
-            maximum=1.0,
-            value=0.95,
-            step=0.05,
-            label="Top-p (nucleus sampling)",
-        ),
-    ],
-)
-
-
-if __name__ == "__main__":
-    demo.launch()
+from datasets import load_dataset
+from sentence_transformers import SentenceTransformer, CrossEncoder
+import faiss
+import torch
+import numpy as np
+from transformers import AutoTokenizer, AutoModelForCausalLM, pipeline
+
+# Load corpus
+print("Loading dataset...")
+dataset = load_dataset("rag-datasets/rag-mini-wikipedia", "text-corpus", split="train[:1000]")
+corpus = [f"{item['title']}\n{item['text']}" for item in dataset]
+
+# Embedding model
+print("Encoding corpus...")
+embedder = SentenceTransformer("all-MiniLM-L6-v2")
+corpus_embeddings = embedder.encode(corpus, convert_to_tensor=True)
+corpus_embeddings_np = corpus_embeddings.cpu().numpy()
+
+# FAISS index
+index = faiss.IndexFlatL2(corpus_embeddings_np.shape[1])
+index.add(corpus_embeddings_np)
+
+# Reranker model
+reranker = CrossEncoder("cross-encoder/ms-marco-MiniLM-L-6-v2")
+
+# Generator (choose one: local HF model or OpenAI)
+tokenizer = AutoTokenizer.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2")
+model = AutoModelForCausalLM.from_pretrained("mistralai/Mistral-7B-Instruct-v0.2", device_map="auto", torch_dtype=torch.float16)
+generator = pipeline("text-generation", model=model, tokenizer=tokenizer, max_new_tokens=150)
+
+def rag_pipeline(query):
+    # Embed query
+    query_embedding = embedder.encode([query], convert_to_tensor=True).cpu().numpy()
+
+    # Retrieve top-k from FAISS
+    D, I = index.search(query_embedding, k=5)
+    retrieved_docs = [corpus[idx] for idx in I[0]]
+
+    # Rerank
+    rerank_pairs = [[query, doc] for doc in retrieved_docs]
+    scores = reranker.predict(rerank_pairs)
+    reranked_docs = [doc for _, doc in sorted(zip(scores, retrieved_docs), reverse=True)]
+
+    # Combine for context
+    context = "\n\n".join(reranked_docs[:2])
+    prompt = f"""Answer the following question using the provided context.\n\nContext:\n{context}\n\nQuestion: {query}\nAnswer:"""
+
+    # Generate
+    response = generator(prompt)[0]["generated_text"]
+    return response.split("Answer:")[-1].strip()
+
+# Gradio UI
+iface = gr.Interface(fn=rag_pipeline,
+                     inputs=gr.Textbox(lines=2, placeholder="Ask something..."),
+                     outputs="text",
+                     title="Mini RAG Wikipedia Demo",
+                     description="Retrieval-Augmented Generation on a small Wikipedia subset.")
+
+iface.launch()

requirements.txt

@@ -1 +1,6 @@
-huggingface_hub==0.25.2
+torch
+transformers
+datasets
+sentence-transformers
+gradio
+faiss-gpu