cloneofsimo / min-fsdp

Related projects: ⓘ

• cloneofsimo / min-max-gpt
  Minimal (400 LOC) implementation Maximum (multi-node, FSDP) GPT training
  ☆110Updated 5 months ago
• srush / triton-autodiff
  Experiment of using Tangent to autodiff triton
  ☆66Updated 7 months ago
• mgmalek / efficient_cross_entropy
  ☆66Updated 3 months ago
• dvruette / barrel-rec-pytorch
  ☆53Updated 8 months ago
• jxbz / modula
  Scalable neural net training via automatic normalization in the modular norm.
  ☆108Updated 3 weeks ago
• softmax1 / Flash-Attention-Softmax-N
  CUDA and Triton implementations of Flash Attention with SoftmaxN.
  ☆66Updated 3 months ago
• mlfoundations / scaling
  Language models scale reliably with over-training and on downstream tasks
  ☆91Updated 5 months ago
• Zyphra / tree_attention
  Tree Attention: Topology-aware Decoding for Long-Context Attention on GPU clusters
  ☆94Updated 2 weeks ago
• amirzandieh / HyperAttention
  Triton Implementation of HyperAttention Algorithm
  ☆46Updated 9 months ago
• athms / mad-lab
  A MAD laboratory to improve AI architecture designs 🧪
  ☆84Updated 4 months ago
• cat-state / tinypar
  ☆20Updated last year
• joey00072 / ohara
  Collection of autoregressive model implementation
  ☆62Updated 2 weeks ago
• cloneofsimo / scaling-guide
  WIP
  ☆76Updated last month
• MatX-inc / seqax
  seqax = sequence modeling + JAX
  ☆129Updated 2 months ago
• HazyResearch / zoology
  Understand and test language model architectures on synthetic tasks.
  ☆156Updated 4 months ago
• aredden / torch-cublas-hgemm
  PyTorch half precision gemm lib w/ fused optional bias + optional relu/gelu
  ☆25Updated 3 weeks ago
• young-geng / scalax
  A simple library for scaling up JAX programs
  ☆116Updated last month
• cuda-mode / ring-attention
  ring-attention experiments
  ☆89Updated 5 months ago
• lucidrains / autoregressive-linear-attention-cuda
  CUDA implementation of autoregressive linear attention, with all the latest research findings
  ☆43Updated last year
• lucidrains / PEER-pytorch
  Pytorch implementation of the PEER block from the paper, Mixture of A Million Experts, by Xu Owen He at Deepmind
  ☆105Updated 3 weeks ago
• schwartz-lab-NLP / TOVA
  Token Omission Via Attention
  ☆118Updated 7 months ago
• ayaka14732 / llama-2-jax
  JAX implementation of the Llama 2 model
  ☆205Updated 7 months ago
• nshepperd / flash_attn_jax
  JAX bindings for Flash Attention v2
  ☆76Updated 2 months ago
• shikaiqiu / compute-better-spent
  ☆42Updated 3 months ago
• vdesai2014 / inference-optimization-blog-post
  ☆82Updated 6 months ago
• siyan-zhao / prepacking
  The source code of our work "Prepacking: A Simple Method for Fast Prefilling and Increased Throughput in Large Language Models"
  ☆57Updated 5 months ago
• epfml / schedules-and-scaling
  ☆47Updated 3 months ago
• RobertCsordas / moeut
  ☆50Updated last month
• andyehrenberg / flaxlm
  ☆27Updated this week
• lucidrains / taylor-series-linear-attention
  Explorations into the recently proposed Taylor Series Linear Attention
  ☆85Updated last month