mirror of
https://github.com/ikawrakow/ik_llama.cpp.git
synced 2026-05-03 04:41:47 +00:00
13 lines
335 B
Markdown
13 lines
335 B
Markdown
### 🔀 [#111](https://github.com/ikawrakow/ik_llama.cpp/pull/111) - Use fused mul - unary op also for MoE models
|
|
|
|
| **Author** | `ikawrakow` |
|
|
| :--- | :--- |
|
|
| **State** | ❌ **Closed** |
|
|
| **Created** | 2024-10-26 |
|
|
| **Updated** | 2024-10-26 |
|
|
|
|
---
|
|
|
|
#### Description
|
|
|
|
This gives us a ~1% speedup for MoE models on CUDA and Metal. |