graphcore-research / unit-scaling

Related projects ⓘ

Alternatives and complementary repositories for unit-scaling

• srush / triton-autodiff
  Experiment of using Tangent to autodiff triton
  ☆72Updated 9 months ago
• pytorch-labs / float8_experimental
  This repository contains the experimental PyTorch native float8 training UX
  ☆211Updated 3 months ago
• mgmalek / efficient_cross_entropy
  ☆77Updated 5 months ago
• cloneofsimo / min-fsdp
  ☆73Updated 4 months ago
• proger / accelerated-scan
  Accelerated First Order Parallel Associative Scan
  ☆163Updated 3 months ago
• young-geng / scalax
  A simple library for scaling up JAX programs
  ☆127Updated 2 weeks ago
• cchan / tccl
  extensible collectives library in triton
  ☆71Updated last month
• gpu-mode / triton-index
  Cataloging released Triton kernels.
  ☆134Updated 2 months ago
• epfml / dynamic-sparse-flash-attention
  ☆132Updated last year
• jundaf2 / INT8-Flash-Attention-FMHA-Quantization
  ☆156Updated last year
• MatX-inc / seqax
  seqax = sequence modeling + JAX
  ☆133Updated 4 months ago
• nikhilvyas / SOAP
  ☆128Updated this week
• google-research / jaxpruner
  ☆207Updated 6 months ago
• pytorch-labs / applied-ai
  Applied AI experiments and examples for PyTorch
  ☆166Updated 3 weeks ago
• Dao-AILab / fast-hadamard-transform
  Fast Hadamard transform in CUDA, with a PyTorch interface
  ☆111Updated 5 months ago
• google / aqt
  ☆267Updated this week
• lucidrains / triton-transformer
  Implementation of a Transformer, but completely in Triton
  ☆249Updated 2 years ago
• davisyoshida / lorax
  LoRA for arbitrary JAX models and functions
  ☆132Updated 8 months ago
• shawntan / scattermoe
  Triton-based implementation of Sparse Mixture of Experts.
  ☆185Updated last month
• graphcore-research / out-of-the-box-fp8-training
  Demo of the unit_scaling library, showing how a model can be easily adapted to train in FP8.
  ☆35Updated 4 months ago
• jxbz / modula
  Scalable neural net training via automatic normalization in the modular norm.
  ☆121Updated 3 months ago
• mobiusml / gemlite
  Simple and fast low-bit matmul kernels in CUDA / Triton
  ☆143Updated this week
• softmax1 / Flash-Attention-Softmax-N
  CUDA and Triton implementations of Flash Attention with SoftmaxN.
  ☆66Updated 5 months ago
• nshepperd / flash_attn_jax
  JAX bindings for Flash Attention v2
  ☆79Updated 4 months ago
• cloneofsimo / min-max-gpt
  Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training
  ☆113Updated 7 months ago
• foundation-model-stack / fms-fsdp
  🚀 Efficiently (pre)training foundation models with native PyTorch features, including FSDP for training and SDPA implementation of Flash…
  ☆193Updated this week
• lucidrains / autoregressive-linear-attention-cuda
  CUDA implementation of autoregressive linear attention, with all the latest research findings
  ☆43Updated last year
• Deep-Learning-Profiling-Tools / triton-viz
  ☆152Updated this week
• albanD / subclass_zoo
  ☆148Updated 5 months ago
• insuhan / hyper-attn
  ☆74Updated 11 months ago