You can deploy this model and start earning money today!

pseudobacon/
SageAttention2

Views

No views yet

SageAttention 2.2.0 — RTX 50 Series Blackwell (Linux)

Pre-built SageAttention 2.2.0 for NVIDIA RTX 5090 (sm_120 Blackwell) on Linux. INT8 quantized attention achieving 2-3x speedup over FlashAttention2 without accuracy loss.

Built from woct0rdho/SageAttention — a fork of thu-ml/SageAttention with ABI3 stable wheel support and additional fixes.

Wheel

sageattention-2.2.0+cu130torch2.11.0andhigher-cp39-abi3-linux_x86_64.whl

Requirements

Component	Version
GPU	NVIDIA RTX 50 series (sm_120 Blackwell)
OS	Linux x86_64
Python	3.9+ (ABI3 stable — compatible with Python 3.9 or higher)
PyTorch	2.11.0 or higher
CUDA	13.0 or higher
Triton	compatible with your PyTorch version

Installation

pip install sageattention-2.2.0+cu130torch2.11.0andhigher-cp39-abi3-linux_x86_64.whl --no-deps

Important: Always use --no-deps to prevent pip from overwriting your PyTorch installation.

Verify

python

import torch
print(torch.__version__)

from sageattention import sageattn_varlen
print("SageAttention 2.2.0 OK")

Features

sageattn_varlen: Variable-length attention (used by SeedVR2, WanVideo, etc.)
INT8 QK quantization with per-thread granularity
FP16 PV accumulation for accuracy
Supports sm_80 (Ampere), sm_89 (Ada), sm_120 (Blackwell) via included CUDA kernels + Triton fallback
Native varlen API — no reshaping overhead
ABI3 stable wheel — single wheel compatible with Python 3.9 and above

Source

Fork: woct0rdho/SageAttention
Upstream: thu-ml/SageAttention v2.2.0

License

Apache 2.0

pseudobacon/SageAttention2