open-source repositories
A list of open source generative AI projects for developers who want to contribute to building AI, and their portfolio. All of these repositories are actively being developed and maintained.
Frameworks, runtimes, and systems for training, inference, and scaling.
| Repo | Description | Stars |
|---|---|---|
| huggingface/transformers | Core NLP/LLM models (training & inference) |  |
| pytorch/pytorch | Deep learning engine, foundation of most LLMs |  |
| Lightning-AI/pytorch-lightning | High-level training framework |  |
| ggerganov/llama.cpp | Lightweight LLaMA/LLM inference in pure C++ |  |
| lm-sys/FastChat | Serving platform for open LLMs (Vicuna, Chatbot Arena) |  |
| unslothai/unsloth | 2-5x faster LLM fine-tuning with 80% less memory usage |  |
| huggingface/trl | Transformer Reinforcement Learning - train language models with RLHF |  |
| microsoft/DeepSpeed | Distributed training and optimization for large models |  |
| mlflow/mlflow | End-to-end ML lifecycle management (tracking, serving) |  |
|
Multi-agent frameworks and components (e.g. memory).
| Repo | Description | Stars |
|---|---|---|
| camel-ai/camel | Multi-agent agentic framework. |  |
| mem0ai/mem0 | Universal memory layer for AI Agents |  |
| langchain-ai/langchain | Framework for LLM-powered agents & tool use |  |
| jerryjliu/llama_index | Data framework for retrieval-augmented generation (RAG) |  |
| Significant-Gravitas/Auto-GPT | Autonomous goal-driven GPT-4 agent |  |
| TransformerOptimus/SuperAGI | Framework for building & managing autonomous agents |  |
| microsoft/semantic-kernel | SDK for LLM orchestration (skills, memory) |  |
| FlowiseAI/Flowise | Low-code visual builder for LLM workflows |  |
| FoundationAgents/MetaGPT | Multi-Agent Framework |  |
| langchain-ai/langgraph | Build resilient language agents as graphs. |  |
Open datasets & dataset tools for training or fine-tuning models.
| Repo | Description | Stars |
|---|---|---|
| (This section will be expanded) |
Frameworks for testing and benchmarking LLMs & agents.
| Repo | Description | Stars |
|---|---|---|
| EleutherAI/lm-evaluation-harness | A framework for few-shot evaluation of language models |  |
| confident-ai/deepeval | The LLM Evaluation Framework |  |
Libraries, plugins, and utilities that extend LLM/GenAI usage.
| Repo | Description | Stars |
|---|---|---|
| huggingface/diffusers | Diffusion models for image/audio/video generation |  |