Ecosystem & Stack: deepspeed

A next-generation training engine optimized for ultra-large Mixture-of-Experts (MoE) models, offering superior efficiency and scalability.

An easy-to-use, scalable, and high-performance open-source framework for Reinforcement Learning from Human Feedback (RLHF), leveraging Ray and vLLM for distributed training of LLMs and VLMs.

A Kubernetes-native platform for scalable distributed AI model training and LLM fine-tuning across various frameworks.

An open-source, cloud-native, all-in-one MLOps platform designed for the full lifecycle management of machine learning, deep learning, and large language model development and deployment.

An end-to-end platform for fine-tuning, evaluating, and deploying open-source Large Language Models (LLMs) and Vision Language Models (VLMs).

A user-friendly, versatile fine-tuning kit for image, video, and audio diffusion models, emphasizing simplicity and cutting-edge features.

A deep learning toolkit for training custom multimodal large language models focused on speech, language, audio, and music processing.

BELLE is an open-source project dedicated to fostering the development of Chinese conversational large language models, aiming to make LLMs accessible to everyone.

YaYi is an open-source Chinese Large Language Model, built on LLaMA 2 & BLOOM, designed for secure, reliable, and domain-specific applications through extensive instruction tuning.

Provides robust training recipes and scripts to align large language models with human and AI preferences, enhancing helpfulness and safety.

An open-source framework providing code, models, and insights for stable Reinforcement Learning from Human Feedback (RLHF) training in Large Language Models, focusing on the PPO algorithm and reward modeling.

An open-source PyTorch implementation and replication of OpenAI's DALL-E, a text-to-image transformer, including CLIP for generation ranking.

A framework for embedding secret cryptographic fingerprints into Large Language Models (LLMs) via fine-tuning to verify ownership and prevent unauthorized use.

A toolkit for finetuning ChatGLM series models (ChatGLM-6B, ChatGLM2-6B, ChatGLM3-6B) using various methods like Freeze, Lora, P-tuning, and full parameter training for downstream NLP tasks.

A modularized multimodal large language model designed for OCR-free document understanding.