SGLang is yet another fast serving framework for large language models and vision language models.

Materials for the Hugging Face Diffusion Models Course

LLM101n: Let's build a Storyteller

Go ahead and axolotl questions

A curated list of reinforcement learning with human feedback resources (continually updated)

Custom data types and layouts for training and inference

Self-Explore to avoid ️the p️️it! Improving the Reasoning Capabilities of Language Models with Fine-grained Rewards

A Native-PyTorch Library for LLM Fine-tuning

Implementation of RLHF (Reinforcement Learning with Human Feedback) on top of the PaLM architecture. Basically ChatGPT but with PaLM

Firefly: 大模型训练工具，支持训练Qwen2、Yi1.5、Phi-3、Llama3、Gemma、MiniCPM、Yi、Deepseek、Orion、Xverse、Mixtral-8x7B、Zephyr、Mistral、Baichuan2、Llma2、Llama、Qwen、Baichuan、ChatGLM2、InternLM、Ziya2、Vicuna、Bloom等大模型

A framework for few-shot evaluation of language models.

Simple and efficient pytorch-native transformer text generation in <1000 LOC of python.

Train transformer language models with reinforcement learning.

OpenCompass is an LLM evaluation platform, supporting a wide range of models (Llama3, Mistral, InternLM2,GPT-4,LLaMa2, Qwen,GLM, Claude, etc) over 100+ datasets.

Advantage Leftover Lunch Reinforcement Learning (A-LoL RL): Improving Language Models with Advantage-based Offline Policy Gradients

Safe RLHF: Constrained Value Alignment via Safe Reinforcement Learning from Human Feedback

The official implementation of Self-Play Fine-Tuning (SPIN)

A Toolkit for Distributional Control of Generative Models

A library with extensible implementations of DPO, KTO, PPO, ORPO, and other human-aware loss functions (HALOs).

Reference implementation for DPO (Direct Preference Optimization)

[Arxiv] Aligning Modalities in Vision Large Language Models via Preference Fine-tuning

High-quality single file implementation of Deep Reinforcement Learning algorithms with research-friendly features (PPO, DQN, C51, DDPG, TD3, SAC, PPG)

A framework for prompt tuning using Intent-based Prompt Calibration

Scenic: A Jax Library for Computer Vision Research and Beyond

Mixture-of-Experts for Large Vision-Language Models

Efficient implementations of state-of-the-art linear attention models in Pytorch and Triton

Implementation of various self-attention mechanisms focused on computer vision. Ongoing repository.

Implementation of the training framework proposed in Self-Rewarding Language Model, from MetaAI