sci-ml/flash-attn (tatsh-overlay)

Search

Install

Install this package:

emerge -a sci-ml/flash-attn

Package Information

Description:
Flash Attention: Fast and Memory-Efficient Exact Attention (Python component).
Homepage:
https://github.com/Dao-AILab/flash-attention
License:
BSD

Versions

Version EAPI Keywords Slot
2.8.3_p1 8 ~amd64 0

Metadata

Maintainers

Upstream

Raw Metadata XML
<pkgmetadata>
	<maintainer type="person">
		<email>audvare@gmail.com</email>
		<name>Andrew Udvare</name>
	</maintainer>
	<use>
		<flag name="rocm">Enable ROCm support.</flag>
	</use>
	<upstream>
		<remote-id type="github">Dao-AILab/flash-attention</remote-id>
		<remote-id type="pypi">flash-attn</remote-id>
	</upstream>
</pkgmetadata>

Lint Warnings

USE Flags

Manage flags for this package: euse -i <flag> -p sci-ml/flash-attn | euse -E <flag> -p sci-ml/flash-attn | euse -D <flag> -p sci-ml/flash-attn

Flag Description 2.8.3_p1
cuda ⚠️
rocm Enable ROCm support.

Files

Manifest

Type File Size Versions
Unmatched Entries
Type File Size
DIST flash_attn-2.8.3.post1.tar.gz 8451657 bytes