[PR #891] [MERGED] use compile for gate, slight performance improvement #1503

New issue

Closed

opened 2026-10-29 14:53:00 +00:00 by bsakaguchi · 0 comments

bsakaguchi commented

2026-10-29 14:53:00 +00:00

Owner

📋 Pull Request Information

Original PR: https://github.com/kvcache-ai/ktransformers/pull/891
Author: @Atream
Created: 3/14/2025
Status: ✅ Merged
Merged: 3/14/2025
Merged by: @Atream

Base: main ← Head: performance-optimize-gpu

📝 Commits (1)

a889288 use compile for gate, slight performance improvement

📊 Changes

9 files changed (+155 additions, -37 deletions)

View changed files

📝 ktransformers/operators/gate.py (+136 -17)
📝 ktransformers/operators/linear.py (+0 -1)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-fp8-linear-ggml-experts.yaml (+1 -1)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-4.yaml (+4 -4)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-8.yaml (+8 -8)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-fp8-linear-ggml-experts.yaml (+2 -2)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-marlin.yaml (+1 -1)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu.yaml (+2 -2)
📝 ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat.yaml (+1 -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/kvcache-ai/ktransformers/pull/891 **Author:** [@Atream](https://github.com/Atream) **Created:** 3/14/2025 **Status:** ✅ Merged **Merged:** 3/14/2025 **Merged by:** [@Atream](https://github.com/Atream) **Base:** `main` ← **Head:** `performance-optimize-gpu` --- ### 📝 Commits (1) - [`a889288`](https://github.com/kvcache-ai/ktransformers/commit/a889288fc1c3906b6aad6fe5852e7b1bc62a3c4a) use compile for gate, slight performance improvement ### 📊 Changes **9 files changed** (+155 additions, -37 deletions) <details> <summary>View changed files</summary> 📝 `ktransformers/operators/gate.py` (+136 -17) 📝 `ktransformers/operators/linear.py` (+0 -1) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-fp8-linear-ggml-experts.yaml` (+1 -1) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-4.yaml` (+4 -4) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-8.yaml` (+8 -8) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-fp8-linear-ggml-experts.yaml` (+2 -2) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu-marlin.yaml` (+1 -1) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat-multi-gpu.yaml` (+2 -2) 📝 `ktransformers/optimize/optimize_rules/DeepSeek-V3-Chat.yaml` (+1 -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

2026-10-29 14:53:00 +00:00