Skip to main content

flash-attn quick installer

Project description

flashattn-helper

A small command-line utility that detects your Python, PyTorch, CUDA, and C++11 ABI. It then picks the matching prebuilt flash-attn wheel from the official GitHub Releases and installs it.

  • Detects: Python tag, PyTorch version and CUDA tag, C++11 ABI.
  • Filters matching wheels from Releases. Prefers cxx11abiFALSE by default.
  • One command install. Or use --dry-run to print the selected wheel URL.

Note: Currently optimized for Linux x86_64 official wheels. Windows or ROCm will print guidance but will not auto-install.

Installation

pip install flashattn

Usage

flashattn plan            # show environment and the wheel to be used
flashattn install         # download and install
flashattn install --version 2.5.8
flashattn install --abi FALSE --dry-run
flashattn uninstall       # remove existing flash-attn
flashattn doctor          # quick compatibility checks

Project details


Download files

Download the file for your platform. If you're not sure which to choose, learn more about installing packages.

Source Distribution

qifa-0.1.0.tar.gz (5.0 kB view details)

Uploaded Source

Built Distribution

If you're not sure about the file name format, learn more about wheel file names.

qifa-0.1.0-py3-none-any.whl (5.3 kB view details)

Uploaded Python 3

File details

Details for the file qifa-0.1.0.tar.gz.

File metadata

  • Download URL: qifa-0.1.0.tar.gz
  • Upload date:
  • Size: 5.0 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.9

File hashes

Hashes for qifa-0.1.0.tar.gz
Algorithm Hash digest
SHA256 899ccc020bf273c3755004ed89093b764b0cc44075599ef5aa8df694cb037a15
MD5 b1c747786e401ebf947c999030343efd
BLAKE2b-256 97964e9a77130f8ee490608a8f30d0d34054d056f7320bd879c92febae329d68

See more details on using hashes here.

File details

Details for the file qifa-0.1.0-py3-none-any.whl.

File metadata

  • Download URL: qifa-0.1.0-py3-none-any.whl
  • Upload date:
  • Size: 5.3 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/6.2.0 CPython/3.12.9

File hashes

Hashes for qifa-0.1.0-py3-none-any.whl
Algorithm Hash digest
SHA256 1503d5e15f50c58db3c5993eccfe3cabad748a510b54fd20318562ce3ba8747b
MD5 4b1755c09ba4efc40bade65b6f0a16eb
BLAKE2b-256 e520763694ba57540431bfad7c55ef576244df706f08a33c6df75aa53812da80

See more details on using hashes here.

Supported by

AWS Cloud computing and Security Sponsor Datadog Monitoring Depot Continuous Integration Fastly CDN Google Download Analytics Pingdom Monitoring Sentry Error logging StatusPage Status page