No project description provided

These details have not been verified by PyPI

Project links

Project description

ml_dtypes

This is not an officially supported Google product.

ml_dtypes is a stand-alone implementation of several NumPy dtype extensions used in machine learning libraries, including:

bfloat16: an alternative to the standard float16 format
float8_*: several experimental 8-bit floating point representations including:
- float8_e4m3b11
- float8_e4m3fn
- float8_e5m2

Installation

The ml_dtypes package is tested with Python versions 3.9-3.11, and can be installed with the following command:

pip install ml_dtypes

To test your installation, you can run the following:

pip install absl-py pytest
pytest --pyargs ml_dtypes

To build from source, clone the repository and run:

git submodule init
git submodule update
pip install .

Example Usage

>>> from ml_dtypes import bfloat16
>>> import numpy as np
>>> np.zeros(4, dtype=bfloat16)
array([0, 0, 0, 0], dtype=bfloat16)

Importing ml_dtypes also registers the data types with numpy, so that they may be referred to by their string name:

>>> np.dtype('bfloat16')
dtype(bfloat16)
>>> np.dtype('float8_e5m2')
dtype(float8_e5m2)

Project details

These details have not been verified by PyPI

Project links

Release history Release notifications | RSS feed

0.5.0

Sep 13, 2024

0.4.1

Sep 13, 2024

0.4.0

Apr 1, 2024

0.4.0b1 pre-release

Mar 12, 2024

0.3.2

Jan 3, 2024

0.3.1

Sep 22, 2023

0.3.0

Sep 19, 2023

0.2.0

Jun 6, 2023

0.1.0

Apr 11, 2023

0.0.4

Mar 27, 2023

0.0.3

Mar 22, 2023

This version

0.0.2

Mar 21, 2023

0.0.2rc5 pre-release

Mar 20, 2023

0.0.2rc4 pre-release

Mar 20, 2023

0.0.1

Mar 10, 2023

Download files

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

Source Distribution

ml_dtypes-0.0.2.tar.gz (682.9 kB view hashes)

Uploaded Mar 21, 2023 Source

Built Distributions

ml_dtypes-0.0.2-cp311-cp311-win_amd64.whl (135.2 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.11 Windows x86-64

ml_dtypes-0.0.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (230.9 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.11 manylinux: glibc 2.17+ x86-64

ml_dtypes-0.0.2-cp311-cp311-macosx_10_9_universal2.whl (311.9 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.11 macOS 10.9+ universal2 (ARM64, x86-64)

ml_dtypes-0.0.2-cp310-cp310-win_amd64.whl (135.3 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.10 Windows x86-64

ml_dtypes-0.0.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (230.9 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.10 manylinux: glibc 2.17+ x86-64

ml_dtypes-0.0.2-cp310-cp310-macosx_10_9_universal2.whl (311.9 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.10 macOS 10.9+ universal2 (ARM64, x86-64)

ml_dtypes-0.0.2-cp39-cp39-win_amd64.whl (135.3 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.9 Windows x86-64

ml_dtypes-0.0.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (231.0 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.9 manylinux: glibc 2.17+ x86-64

ml_dtypes-0.0.2-cp39-cp39-macosx_10_9_universal2.whl (312.2 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.9 macOS 10.9+ universal2 (ARM64, x86-64)

ml_dtypes-0.0.2-cp38-cp38-win_amd64.whl (134.7 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.8 Windows x86-64

ml_dtypes-0.0.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl (230.9 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.8 manylinux: glibc 2.17+ x86-64

ml_dtypes-0.0.2-cp38-cp38-macosx_10_9_universal2.whl (311.7 kB view hashes)

Uploaded Mar 21, 2023 CPython 3.8 macOS 10.9+ universal2 (ARM64, x86-64)

Hashes for ml_dtypes-0.0.2.tar.gz

Hashes for ml_dtypes-0.0.2.tar.gz
Algorithm	Hash digest
SHA256	`14a0795ba84095dd012c2f7c387fdf813301c1e6be51e4c9b56809ee542f9520`
MD5	`2e51cafbce59c415a1a1e7e7935dceb1`
BLAKE2b-256	`f6d99b106fd8768c644fd0e54408fd2aab168a164cf6356817d5770589c47e06`

Hashes for ml_dtypes-0.0.2-cp311-cp311-win_amd64.whl

Hashes for ml_dtypes-0.0.2-cp311-cp311-win_amd64.whl
Algorithm	Hash digest
SHA256	`7893db4a13023aec612fb5e81c20845e5f5c0e5be7200ce6a7025272fb4b4abd`
MD5	`0409903fb93d5be8fcc7e049006b3046`
BLAKE2b-256	`77b0f6f26b173e40dd283b79f0951fba9e265b1f73886baf381492ba6129868b`

Hashes for ml_dtypes-0.0.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for ml_dtypes-0.0.2-cp311-cp311-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`e817ad95e8344258d0d63ea6f12e11966cfad980c6a5cd701a26e4fe1f3b2f43`
MD5	`f1e74c9b8bcff6f65152784797086696`
BLAKE2b-256	`396797953858927c542508d51225577d76c6ef46c3b00c672f83f7d0e99c9f0d`

Hashes for ml_dtypes-0.0.2-cp311-cp311-macosx_10_9_universal2.whl

Hashes for ml_dtypes-0.0.2-cp311-cp311-macosx_10_9_universal2.whl
Algorithm	Hash digest
SHA256	`da479bd7ff83c5e446f7c04eef57fcf5cd5c6f3221259e9faf52fba4d7e6bb55`
MD5	`b3314301c40a85f197390b83844933a1`
BLAKE2b-256	`9f1f1c5a02fd968c33212981e26f348cdf17279f5803a44dc8e3bce4f9061f7f`

Hashes for ml_dtypes-0.0.2-cp310-cp310-win_amd64.whl

Hashes for ml_dtypes-0.0.2-cp310-cp310-win_amd64.whl
Algorithm	Hash digest
SHA256	`c1a757cd21022ab1731753fc4f0c2985cc7455713900fd7478b5398d5be75a95`
MD5	`fc7d675e63d48f3a7fef754b016eced5`
BLAKE2b-256	`381e67676a18734ce4f500b40e80e45b4a4eb967c0e8164b25f0e7d51afa1add`

Hashes for ml_dtypes-0.0.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for ml_dtypes-0.0.2-cp310-cp310-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`e4ee04c62001923c4ee488ca3469561e34fe587c7ab6bee35ffce1ab79f114d0`
MD5	`46d3bd27ccadc90987fab631e7a92142`
BLAKE2b-256	`c144c119c1b8fe2685b123e63074a87d22e5f959b65125f8580ac508687a419a`

Hashes for ml_dtypes-0.0.2-cp310-cp310-macosx_10_9_universal2.whl

Hashes for ml_dtypes-0.0.2-cp310-cp310-macosx_10_9_universal2.whl
Algorithm	Hash digest
SHA256	`7964d7bf373eb28072d0aa33c011e047c06d946ec275e82e1a1ad47deadb038b`
MD5	`99b977db6e3ba6afb48cbdc112eee15a`
BLAKE2b-256	`e3ec3d65b01b392263c6a7470586ca10026d5c2d7e7a1b0335e9c0361e7c5527`

Hashes for ml_dtypes-0.0.2-cp39-cp39-win_amd64.whl

Hashes for ml_dtypes-0.0.2-cp39-cp39-win_amd64.whl
Algorithm	Hash digest
SHA256	`85cfff8caaaa6bb9f039e3826b7f77ad2d4420dc6940510449566270362fd448`
MD5	`d7bfdcc08ce87b9ca94e61ed1f1a4e3d`
BLAKE2b-256	`90958c565fa2c81ad7119716ba1435725797bc41c5b5a4fd259526ce51c9b6a3`

Hashes for ml_dtypes-0.0.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for ml_dtypes-0.0.2-cp39-cp39-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`54f63f549172c21315ce21d17a633d7de1c06b4d96bb4ec944c2972ef5896c36`
MD5	`b63d243b60d3a58914c2734bc59194f9`
BLAKE2b-256	`8760bbbfd7ad4dfeed88a7404d5101b5379000f73cf0f6e48ac31f828f9a42f8`

Hashes for ml_dtypes-0.0.2-cp39-cp39-macosx_10_9_universal2.whl

Hashes for ml_dtypes-0.0.2-cp39-cp39-macosx_10_9_universal2.whl
Algorithm	Hash digest
SHA256	`63d775489736c6a7d8dee0c48b8d724ca3ce4cbe3bb7d1192c2281ebbd879449`
MD5	`f09a49f1510f3a12ffa27e3ba9f88e6e`
BLAKE2b-256	`b1790b637adeef30ced27800520fdc30c43c868fcd3360b13127bb4ae564b6e1`

Hashes for ml_dtypes-0.0.2-cp38-cp38-win_amd64.whl

Hashes for ml_dtypes-0.0.2-cp38-cp38-win_amd64.whl
Algorithm	Hash digest
SHA256	`b167a2511c6a630a4a5f66b4780a3a1c526449f67e061aa6a78a551272c435b6`
MD5	`988800fa05453c54cfc07e333fb71027`
BLAKE2b-256	`0fbed87875c3009218e5f29766eff6017f978384d83771d8a586cc2a537bc7e8`

Hashes for ml_dtypes-0.0.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl

Hashes for ml_dtypes-0.0.2-cp38-cp38-manylinux_2_17_x86_64.manylinux2014_x86_64.whl
Algorithm	Hash digest
SHA256	`b98979a7084f392a5530d64bb46cc04ad2032b9ca68cccac1680f4b8805c274b`
MD5	`7d9d0f2e6392174431f4de40d000bd7f`
BLAKE2b-256	`f76cc401359e0137055fd03f3bd61b867aba08e22104dd728e4d762df20d28a1`

Hashes for ml_dtypes-0.0.2-cp38-cp38-macosx_10_9_universal2.whl

Hashes for ml_dtypes-0.0.2-cp38-cp38-macosx_10_9_universal2.whl
Algorithm	Hash digest
SHA256	`935097a4a4ea1305dbee97d19ded0d82168afe11f6431b1f2e9bd1477528fcc1`
MD5	`102ced1aa372e8d5fcac28f033e8881f`
BLAKE2b-256	`3a26c78b302a69bf487dc3fe2ae188e14f04de752e9469d5ff3e12c1edc53b64`