[PR #207] [MERGED] Optimize int8 attention with mask. #222

New issue

Closed

opened 2026-10-26 22:21:44 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-26 22:21:44 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/Comfy-Org/comfy-kitchen/pull/207
Author: @comfyanonymous
Created: 9/26/2026
Status: ✅ Merged
Merged: 9/27/2026
Merged by: @comfyanonymous

Base: main ← Head: dev/attn_bias

📝 Commits (3)

5b1c690 Optimize int8 attention with mask.
69662e2 Small fix.
0bb107c Hip fixes.

📊 Changes

8 files changed (+969 additions, -221 deletions)

View changed files

📝 comfy_kitchen/backends/cuda/dlpack_bindings.cpp (+126 -90)
📝 comfy_kitchen/backends/cuda/sage_attention/attn_utils.cuh (+103 -3)
📝 comfy_kitchen/backends/cuda/sage_attention/qk_int_sv_i8_cuda.cuh (+125 -45)
📝 comfy_kitchen/backends/cuda/sage_attention/sage_attn_launcher.cu (+132 -10)
📝 comfy_kitchen/backends/hip/dlpack_bindings.cpp (+56 -3)
📝 comfy_kitchen/backends/hip/sage_attention/int8_attn.hip (+85 -18)
📝 comfy_kitchen/sage_attention.py (+75 -51)
📝 tests/test_int8_attention.py (+267 -1)

📄 Description

No description provided

_{🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.}

## 📋 Pull Request Information **Original PR:** https://github.com/Comfy-Org/comfy-kitchen/pull/207 **Author:** [@comfyanonymous](https://github.com/comfyanonymous) **Created:** 9/26/2026 **Status:** ✅ Merged **Merged:** 9/27/2026 **Merged by:** [@comfyanonymous](https://github.com/comfyanonymous) **Base:** `main` ← **Head:** `dev/attn_bias` --- ### 📝 Commits (3) - [`5b1c690`](https://github.com/Comfy-Org/comfy-kitchen/commit/5b1c690dc807ac8bd00099831ede8adfda06d560) Optimize int8 attention with mask. - [`69662e2`](https://github.com/Comfy-Org/comfy-kitchen/commit/69662e288473bc5add2ed08ccf8561b7e1f370ae) Small fix. - [`0bb107c`](https://github.com/Comfy-Org/comfy-kitchen/commit/0bb107c8b24f266e67da0971d44fe4dffe471344) Hip fixes. ### 📊 Changes **8 files changed** (+969 additions, -221 deletions) <details> <summary>View changed files</summary> 📝 `comfy_kitchen/backends/cuda/dlpack_bindings.cpp` (+126 -90) 📝 `comfy_kitchen/backends/cuda/sage_attention/attn_utils.cuh` (+103 -3) 📝 `comfy_kitchen/backends/cuda/sage_attention/qk_int_sv_i8_cuda.cuh` (+125 -45) 📝 `comfy_kitchen/backends/cuda/sage_attention/sage_attn_launcher.cu` (+132 -10) 📝 `comfy_kitchen/backends/hip/dlpack_bindings.cpp` (+56 -3) 📝 `comfy_kitchen/backends/hip/sage_attention/int8_attn.hip` (+85 -18) 📝 `comfy_kitchen/sage_attention.py` (+75 -51) 📝 `tests/test_int8_attention.py` (+267 -1) </details> ### 📄 Description _No description provided_ --- <sub>🔄 This issue represents a GitHub Pull Request. It cannot be merged through Gitea due to API limitations.</sub>

bsakaguchi

2026-10-26 22:21:44 +00:00