The State of Open Source AI
Clarity in the current fast-paced mess of Open Source innovation
The ebook is a comprehensive guide exploring everything from model evaluations to deployment. Chapters include:
| Chapter | Description |
|---|---|
| Licences | Weights vs Data, Commercial use, Fair use, Pending lawsuits |
| Evaluation & Datasets | Leaderboards & Benchmarks for Text/Visual/Audio models |
| Models | LLaMA 1 vs 2, Stable Diffusion, DALL-E, Persimmon, ... |
| Unaligned Models | FraudGPT, WormGPT, PoisonGPT, WizardLM, Falcon |
| Fine-tuning | LLMs, Visual, & Audio models |
| Model Formats | ONNX, GGML, TensorRT |
| MLOps Engines | vLLM, TGI, Triton, BentoML, ... |
| Vector Databases | Weaviate, Qdrant, Milvus, Redis, Chroma, ... |
| Software Development toolKits | LangChain, LLaMA Index, LiteLLM |
| Desktop Apps | LMStudio, GPT4All, Koboldcpp, ... |
| Hardware | NVIDIA CUDA, AMD ROCm, Apple Silicon, Intel, TPUs, ... |
You can help keep the book up-to-date! Contributions, issues, and comments are welcome!
This book is released under CC-BY-4.0 (text) and Apache-2.0 (code).