slime is an LLM post-training framework for RL Scaling.

Fast and memory-efficient exact attention

NanoGPT (124M) in 3 minutes

A high-throughput and memory-efficient inference and serving engine for LLMs

🤗 Transformers: State-of-the-art Machine Learning for Pytorch, TensorFlow, and JAX.

PyTorch building blocks for the OLMo ecosystem

Efficient optimizers

AllenAI's post-training codebase

Train transformer language models with reinforcement learning.

veRL: Volcano Engine Reinforcement Learning for LLM

An Easy-to-use, Scalable and High-performance RLHF Framework (70B+ PPO Full Tuning & Iterative DPO & LoRA & RingAttention & RFT)

Efficient Triton Kernels for LLM Training

VLM Evaluation: Benchmark for VLMs, spanning text generation tasks from VQA to Captioning

A flexible and efficient codebase for training visually-conditioned language models (VLMs)

An ESP32 and 2.9", 4.2" or 7.5" ePaper Display reads Weather Underground data via their API and then displays the weather

[NeurIPS'23 Oral] Visual Instruction Tuning: LLaVA (Large Language-and-Vision Assistant) built towards GPT-4V level capabilities.

YOLOX is a high-performance anchor-free YOLO, exceeding yolov3~v5 with MegEngine, ONNX, TensorRT, ncnn, and OpenVINO supported. Documentation: https://yolox.readthedocs.io/

A very hacky set of functions for getting plotly to do what I want when doing mech interp research, designed to be compatible with PyTorch

Adaptive Experimentation Platform

BibTeX-js can parse a BibTeX-file and render it as part of an HTML file. This way, you can easily add a list of publications to your private homepage or display a list of recommended publications f…

MCTS for general game playing