sgl-project/sglang

[Kernel] cuDNN attention backend

Open

#2.272 geöffnet am 30. Nov. 2024

Auf GitHub ansehen
 (5 Kommentare) (1 Reaktion) (2 zugewiesene Personen)Python (6.216 Forks)auto 404
enhancementgood first issuehelp wantedhigh priorityinactive

Repository-Metriken

Stars
 (28.442 Stars)
PR-Merge-Metriken
 (Durchschn. Merge 2T 1h) (1.000 gemergte PRs in 30 T)

Beschreibung

cuDNN provides very fast attention implementation and it is well maintained by NVIDIA. We would like to add a new attention backend based on cudnn.

Steps

  1. Learn this cudnn paged attention python api. https://github.com/NVIDIA/cudnn-frontend/blob/v1.8.0/samples/python/52_scaled_dot_product_attention_with_paged_caches.ipynb
  2. Add a new attention backend "cudnn" here https://github.com/sgl-project/sglang/tree/main/python/sglang/srt/layers/attention
  3. We should be able to use it with python3 -m sglang.launch_server --model meta-llama/Llama-3.1-8B-Instruct --attention-backend cudnn

Contributor Guide