tenstorrent TT-Studio
TT-Studio is an easy-to-use web interface for running AI models on Tenstorrent hardware. It combines TT Inference Server's core packaging setup, containerization, and deployment automation with TT-Metal's model execution framework specifically optimized for Tenstorrent hardware.
Important notes:
- TT-Studio requires access to a Tenstorrent AI accelerator for full deployment features
- Alternatively, you can connect just the frontend to a remote API endpoint without direct hardware access
- The platform provides automatic hardware detection and seamless integration with Tenstorrent devices
- Uses containerized deployment through Docker for isolation and easy deployment
- The startup.sh script is deprecated - use
python run.pyfor all operations
Docs
- Main README: Complete overview, setup instructions, and quick start guide
- Setup Guide: Complete installation & configuration using run.py
- FAQ: Quick answers to common questions about TT-Studio
- Model Interface Guide: Using TT-Studio as AI playground (Chat, Vision, Speech, Images)
- Troubleshooting Guide: Solutions for common setup and runtime issues
- Contributing Guide: How to contribute code to the project
- Development Setup: Development environment configuration
Examples
- AI Model Interface: Complete examples of using Chat, Vision, Speech, and Image models
- vLLM Models Guide: Specific examples for running vLLM models
- AI Agent Setup: Setting up and using the AI assistant functionality
Key Components
- Frontend Interface: Modern React-based UI for model interaction and management
- Backend API: Django-based service for model management, deployment, and API endpoints
- TT Inference Server: FastAPI server for handling model inference requests
- Docker Containers: Complete containerization for isolation and easy deployment
- Automatic Hardware Detection: Seamless integration and auto-mounting of Tenstorrent devices (/dev/tenstorrent)
- Automated Setup: Complete environment configuration and model setup automation via run.py script
Supported AI Models
- Chat-based Language Models (LLMs): Text generation and conversational AI
- Computer Vision (YOLO): Object detection and image analysis
- Speech Recognition (Whisper): Audio-to-text transcription
- Image Generation (Stable Diffusion): AI-powered image creation