piwheels

flash-attention-softmax-n

CUDA and Triton implementations of Flash Attention with SoftmaxN.

Installation

In a virtualenv (see these instructions if you need to create one):

pip3 install flash-attention-softmax-n

Dependencies

• einops
• torch

Releases

Version	Released	Bookworm Python 3.11	Trixie Python 3.13	Files
0.3.2	2023-11-21
flash_attention_softmax_n-0.3.2-py3-none-any.whl (34 KB)
0.3.1	2023-09-23
flash_attention_softmax_n-0.3.1-py3-none-any.whl (33 KB)
0.3.0	2023-09-05
flash_attention_softmax_n-0.3.0-py3-none-any.whl (33 KB)
0.2.1	2023-08-30
flash_attention_softmax_n-0.2.1-py3-none-any.whl (26 KB)
0.2.0	2023-08-29
flash_attention_softmax_n-0.2.0-py3-none-any.whl (26 KB)
0.1.4	2023-08-28
flash_attention_softmax_n-0.1.4-py3-none-any.whl (23 KB)
0.1.3	2023-08-28
flash_attention_softmax_n-0.1.3-py3-none-any.whl (23 KB)
0.1.2	2023-08-26
flash_attention_softmax_n-0.1.2-py3-none-any.whl (21 KB)
0.1.1	2023-08-26
flash_attention_softmax_n-0.1.1-py3-none-any.whl (21 KB)
0.1.0	2023-08-26
flash_attention_softmax_n-0.1.0-py3-none-any.whl (21 KB)
0.1.0rc6 pre-release	2023-08-26
flash_attention_softmax_n-0.1.0rc6-py3-none-any.whl (21 KB)

Issues with this package?

• Search issues for this package
• Package or version missing? Open a new issue
• Something else? Open a new issue