Downloads · 30 days
0
Olafangensan/sageattention-rtx5070ti-blackwell-windows
sageattention-rtx5070ti-blackwell-windows is a machine learning model from Olafangensan. Use it for the machine learning task on the model card, and read the license before you ship it in a product. The card lists the license as apache-2.0.
Precompiled Windows .whl for woct0rdho/sageattention-autotune, targeted specifically at NVIDIA's Blackwell architecture (sm120).
Downloads · 30 days
0
Access
Public
Updated Jun 20, 2026
Repo size
10.1 MB
Likes
1
Public
Click a slice to open those files.
.whl10.1 MB · 100%
From the Hugging Face model README
Precompiled Windows .whl for woct0rdho/sageattention-autotune, targeted specifically at NVIDIA's Blackwell architecture (sm_120).
abi3 compatible)Before installation, remove the previous version of SageAttention, if it exists:
# ComfyUI Portable example
.\python_embeded\python.exe -m pip uninstall sageattention
# Standard venv example
pip uninstall sageattention
Download the wheel and install it using your ComfyUI portable Python or venv:
# ComfyUI Portable example
.\python_embeded\python.exe -m pip install sageattention-2.2.0-cp310-abi3-win_amd64.whl
# Standard venv example
pip install sageattention-2.2.0-cp310-abi3-win_amd64.whl
| Resolution | SageAttention 2 (s/it) | PyTorch Attention (s/it) |
|---|---|---|
| 0.5MP | 1.15 | 1.09 |
| 1.0MP | 2.10 | 2.00 |
| 1.5MP | 2.85 | 3.23 |
| 2.0MP | 3.80 | 4.81 |
| 2.5MP | 3.98 | 6.55 |
| 3.0MP | 4.51 | 8.49 |
| 3.5MP | 5.55 | 11.00 |
| 4.0MP | 6.55 | 14.08 |
While standard PyTorch Attention is slightly faster at very low resolutions, SageAttention scales drastically better, cutting 4MP generation time by over 50%.
sageattn_qk_int8_pv_fp16_cuda or auto.Do NOT use the --use-sage-attention launch argument, as it may force the unstable Triton backend on Blackwell.
Compiled from the sageattention-autotune repo with TORCH_CUDA_ARCH_LIST=12.0 to natively target sm_120. The -DTORCH_STABLE_ONLY flag was removed to ensure compatibility with PyTorch nightly builds.