Inference Llama 2 in one file of pure C

Explainability of Deep RL algorithms using graph networks and layer-wise relevance propagation.

LLM training in simple, raw C/CUDA

Language Quantized AutoEncoders

Mora: More like Sora for Generalist Video Generation

Code for the paper "Training Diffusion Models with Reinforcement Learning"

DDPO for finetuning diffusion models, implemented in PyTorch with LoRA support

Large Context Attention

The simplest, fastest repository for training/finetuning medium-sized GPTs.

Serve, optimize and scale PyTorch models in production

Supercharge huggingface transformers with model parallelism.

Text-to-3D & Image-to-3D & Mesh Exportation with NeRF + Diffusion.

GaLore: Memory-Efficient LLM Training by Gradient Low-Rank Projection

Query language for blending SQL logic and LLM reasoning across structured + unstructured data. [Findings of ACL 2024]

DSPy: The framework for programming—not prompting—language models

A benchmark to evaluate language models on questions I've previously asked them to solve.

Minimal, clean code for the Byte Pair Encoding (BPE) algorithm commonly used in LLM tokenization.

[NeurIPS 2020] Official code for the paper "DeepSVG: A Hierarchical Generative Network for Vector Graphics Animation". Includes a PyTorch library for deep learning with SVG data.

Manage scalable open LLM inference endpoints in Slurm clusters

Train transformer language models with reinforcement learning.

Official inference library for Mistral models

Solve puzzles. Learn CUDA.

Solve puzzles. Improve your pytorch.

Mamba SSM architecture