super_agent / README.md
lezaf
Specify langfuse version in requirements.txt
fb5f7c4
---
title: Template Final Assignment
emoji: ๐Ÿ•ต๐Ÿปโ€โ™‚๏ธ
colorFrom: indigo
colorTo: indigo
sdk: gradio
sdk_version: 5.25.2
app_file: app.py
pinned: false
hf_oauth: true
# optional, default duration is 8 hours/480 minutes. Max duration is 30 days/43200 minutes.
hf_oauth_expiration_minutes: 480
---
# ๐Ÿ•ต๏ธโ€โ™‚๏ธ Super Agent: The AI Answer Machine ๐Ÿš€
Welcome to **Super Agent** โ€“ the AI agent built to be evaluated on [GAAI](https://huggingface.co/papers/2311.12983) benchmark to complete the [AI agents](https://huggingface.co/learn/agents-course/unit0/introduction) course by [Hugging Face](https://huggingface.co/)!
---
## โœจ Features
- ๐Ÿ” **Web Search**: Finds up-to-date info using DuckDuckGo or Tavily.
- ๐Ÿ“Š **Excel & Data Extraction**: Reads and analyzes Excel files.
- ๐ŸŽง **Audio & YouTube Transcripts**: Extracts transcripts from audio files and YouTube videos.
- ๐Ÿงฎ **Math Tools**: Adds numbers, checks commutativity, and more.
- ๐Ÿง  **RAG-Ready**: Retrieval-Augmented Generation for smarter, more accurate answers.
- ๐Ÿค– **LLM Integration**: Works with Google Gemini, HuggingFace, and OpenAI models.
- ๐Ÿ›ก๏ธ **Exact Match Mode**: Answers are evaluated for exactness.
- ๐Ÿ› ๏ธ **Extensible Tooling**: Easily add new tools for more superpowers.
- โณ **Gemini Rate Limiter**: Includes a rate limiter implementation for Gemini, perfect for handling free tier API limits.
---
## ๐Ÿงฉ Supported Tools
- `web_search` โ€“ Search the web for up-to-date info
- `extract_data_from_excel` โ€“ Download and parse Excel files
- `extract_transcript_from_youtube` โ€“ Get YouTube video transcripts
- `extract_transcript_from_audio` โ€“ Get transcripts from audio files
- `add_numbers_in_list` โ€“ Add up numbers
- `check_commutativity` โ€“ Test if a binary operation is commutative