Alternatives and similar repositories for FlashAttention-PyTorch

Users that are interested in FlashAttention-PyTorch are comparing it to the libraries listed below

• mdy666 / mdy_triton
  ☆125Updated 2 weeks ago
• NVlabs / COAT
  [ICLR 2025] COAT: Compressing Optimizer States and Activation for Memory-Efficient FP8 Training
  ☆190Updated 3 weeks ago
• OpenNLPLab / lightning-attention
  Lightning Attention-2: A Free Lunch for Handling Unlimited Sequence Lengths in Large Language Models
  ☆285Updated 2 months ago
• 66RING / tiny-flash-attention
  flash attention tutorial written in python, triton, cuda, cutlass
  ☆349Updated this week
• kyegomez / FlashAttention20
  Get down and dirty with FlashAttention2.0 in pytorch, plug in and play no complex CUDA kernels
  ☆104Updated last year
• dhcode-cpp / NSA-pytorch
  DeepSeek Native Sparse Attention pytorch implementation
  ☆65Updated 2 months ago
• madsys-dev / deepseekv2-profile
  ☆132Updated 2 months ago
• mdy666 / Qwen-Native-Sparse-Attention
  qwen-nsa
  ☆60Updated last month
• FlagOpen / FlagAttention
  A collection of memory efficient attention operators implemented in the Triton language.
  ☆266Updated 11 months ago
• fla-org / flame
  🔥 A minimal training framework for scaling FLA models
  ☆128Updated last week
• fxmeng / TransMLA
  TransMLA: Multi-Head Latent Attention Is All You Need
  ☆247Updated this week
• smart-lty / ParallelSpeculativeDecoding
  [ICLR 2025] PEARL: Parallel Speculative Decoding with Adaptive Draft Length
  ☆81Updated last month
• fkodom / grouped-query-attention-pytorch
  (Unofficial) PyTorch implementation of grouped-query attention (GQA) from "GQA: Training Generalized Multi-Query Transformer Models from …
  ☆165Updated last year
• xlite-dev / ffpa-attn
  📚FFPA(Split-D): Extend FlashAttention with Split-D for large headdim, O(1) GPU SRAM complexity, 1.8x~3x↑🎉 faster than SDPA EA.
  ☆174Updated this week
• feifeibear / long-context-attention
  USP: Unified (a.k.a. Hybrid, 2D) Sequence Parallel Attention for Long Context Transformers Model Training and Inference
  ☆492Updated 3 weeks ago
• fanshiqing / grouped_gemm
  PyTorch bindings for CUTLASS grouped GEMM.
  ☆121Updated 4 months ago
• tspeterkim / paged-attention-minimal
  a minimal cache manager for PagedAttention, on top of llama3.
  ☆87Updated 8 months ago
• HandH1998 / QQQ
  QQQ is an innovative and hardware-optimized W4A8 quantization solution for LLMs.
  ☆121Updated last month
• luliyucoordinate / cute-flash-attention
  Implement Flash Attention using Cute.
  ☆82Updated 4 months ago
• ruikangliu / FlatQuant
  Official PyTorch implementation of "FlatQuant: Flatness Matters for LLM Quantization"
  ☆127Updated last week
• InternLM / Awesome-LLM-Training-System
  ☆35Updated 9 months ago
• harleyszhang / llm_counts
  llm theoretical performance analysis tools and support params, flops, memory and latency analysis.
  ☆88Updated 4 months ago
• InternLM / turbomind
  ☆84Updated last month
• ModelTC / EasyLLM
  Built upon Megatron-Deepspeed and HuggingFace Trainer, EasyLLM has reorganized the code logic with a focus on usability. While enhancing …
  ☆46Updated 7 months ago
• nbasyl / LLM-FP4
  The official implementation of the EMNLP 2023 paper LLM-FP4
  ☆198Updated last year
• jackfsuia / nanoRLHF
  RLHF experiments on a single A100 40G GPU. Support PPO, GRPO, REINFORCE, RAFT, RLOO, ReMax, DeepSeek R1-Zero reproducing.
  ☆58Updated 2 months ago
• AniZpZ / AutoSmoothQuant
  An easy-to-use package for implementing SmoothQuant for LLMs
  ☆97Updated last month
• mit-han-lab / tinychat-tutorial
  ☆65Updated 6 months ago
• tgale96 / grouped_gemm
  PyTorch bindings for CUTLASS grouped GEMM.
  ☆89Updated 2 weeks ago
• ailzhang / EfficientPyTorch
  Code release for book "Efficient Training in PyTorch"
  ☆64Updated last month