Spaces:
Sleeping
Sleeping
Create README.md
Browse files
README.md
ADDED
|
@@ -0,0 +1,35 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
title: GAIA LangGraph Agent
|
| 3 |
+
emoji: 🤖
|
| 4 |
+
colorFrom: blue
|
| 5 |
+
colorTo: green
|
| 6 |
+
sdk: gradio
|
| 7 |
+
sdk_version: "4.0.0"
|
| 8 |
+
app_file: main.py
|
| 9 |
+
pinned: false
|
| 10 |
+
---
|
| 11 |
+
|
| 12 |
+
# GAIA Benchmark Agent with LangGraph
|
| 13 |
+
|
| 14 |
+
This project implements an agent using the LangGraph framework to solve the GAIA benchmark tasks provided by the Hugging Face Agents Course (Unit 4).
|
| 15 |
+
|
| 16 |
+
✅ Uses LangChain + LangGraph
|
| 17 |
+
✅ Handles 20 benchmark questions
|
| 18 |
+
✅ Submits answers for scoring via Hugging Face leaderboard
|
| 19 |
+
|
| 20 |
+
## How to run
|
| 21 |
+
|
| 22 |
+
1. Clone this repo and push it as a Hugging Face Space (public).
|
| 23 |
+
2. Install the dependencies.
|
| 24 |
+
3. Launch the Space and log in to Hugging Face.
|
| 25 |
+
4. Click "Run Evaluation & Submit All Answers" to evaluate your agent.
|
| 26 |
+
|
| 27 |
+
## Project files
|
| 28 |
+
|
| 29 |
+
- `main.py` — Gradio app runner
|
| 30 |
+
- `graph_builder.py` — Builds the LangGraph pipeline
|
| 31 |
+
- `llm_node.py` — Defines LLM node logic
|
| 32 |
+
- `formatter_node.py` — Cleans up answers for exact-match submission
|
| 33 |
+
- `api_client.py` — Wraps GAIA API interactions
|
| 34 |
+
|
| 35 |
+
Good luck!
|