Pre-release
This release is a pre-release and may not be stable for production use.
FlashAttention-4 (CuTeDSL)
FlashAttention-4 is a CuTeDSL-based implementation of FlashAttention for Hopper and Blackwell GPUs.
Installation
pip install flash-attn4
Usage
from flash_attn.cute import flash_attn_func, flash_attn_varlen_func
out = flash_attn_func(q, k, v, causal=True)
Development
git clone https://github.com/Dao-AILab/flash-attention.git
cd flash-attention
pip install -e "flash_attn/cute[dev]"
pytest tests/cute/
1 maintainer
Metadata
1 maintainer
Release files for fa4 4.0.0b3
For a detailed explanation of source distributions (sdists) and built distributions (wheels), please see the package formats documentation.
Source distribution (sdist)
| File | Size | Uploaded | |
|---|---|---|---|
| fa4-4.0.0b3.tar.gz | 243.7 kB | Details |
Built distribution (wheel)
| File | Interpreter | ABI | Platform | Reset |
|---|---|---|---|---|
| fa4-4.0.0b3-py3-none-any.whl | Python 3 | none | any | Details |
Total release size: 505.3 kB
Release files / fa4-4.0.0b3.tar.gz
| Download URL | fa4-4.0.0b3.tar.gz |
|---|---|
| Size | 243.7 kB |
| Tags | Source |
|
SHA-256 checksum How to use checksums |
72e368324767976a132fbcd0adeaaad08ccc76dabcb63709d2df94d002936d81
|
|
BLAKE2b-256 checksum How to use checksums |
d74f97e6ce858a9539bf859d02f26a1ac1a03a13801e1107b618bdcfc0a127d9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.11
|
Release files / fa4-4.0.0b3-py3-none-any.whl
| Download URL | fa4-4.0.0b3-py3-none-any.whl |
|---|---|
| Size | 261.6 kB |
| Tags | Python 3 |
|
SHA-256 checksum How to use checksums |
c5fac755d64366cd1052f4b9e674d27b5fb7a4b5c92700647c94c06074618547
|
|
BLAKE2b-256 checksum How to use checksums |
e920d6fada8b91f08bf4fd65ec55df5b18d1c804cde480eb97fdce63d0e989c9
|
| Upload date | |
|
Uploaded using Trusted Publishing? What is trusted publishing? |
No |
| Uploaded via |
twine/6.2.0 CPython/3.13.11
|
