Hi, this package is in development, not ready yet.

It is being tested and most stuff already works.

Needs more layers and archictectures.

Here is how it works:

flowchart LR
    A[You] -->|Want convenience| B(Pytorch)
    B -->|Works slowly because adds python overhead| C[CUDA]

    D[Me] --> |Want to do math| E(Rewrite popular hf models in JAX)
    E --> |Models convert to pytorch through a simple parser and prebuilt layers| F[huggingface]

JAX is fast (because of XLA compilation), works on CUDA and TPU. This repo implements DiLoCo to use any number of heterogenous compute (kaggle, molab, colab, lightning) without direct connection (only huggingface chekpointing).

Key dates

PyPI data
Data sourced directly from PyPI's database.
  • Released:
Newer release available (0.0.2.27)

1 maintainer

PyPI data
Data sourced directly from PyPI's database.
Avatar for Berrny from gravatar.com Berrny

License expression

MIT
View SPDX License List

Requires

Python >=3.11

Provides Extra

cuda cuda12 cuda13 test

Report project as malware

Download files

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

Source Distribution

adjaxt-0.0.2.15.tar.gz (41.1 kB view details)

Uploaded Source

Built Distribution

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

adjaxt-0.0.2.15-py3-none-any.whl (33.8 kB view details)

Uploaded Python 3

File details

Details for the file adjaxt-0.0.2.15.tar.gz.

File metadata

  • Download URL: adjaxt-0.0.2.15.tar.gz
  • Upload date:
  • Size: 41.1 kB
  • Tags: Source
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/7.0.0 CPython/3.14.4

File hashes

Hashes for adjaxt-0.0.2.15.tar.gz
Algorithm Hash digest
SHA256 bdb06a59786701a66909fcf4920c286fbdb976b5ffaadd073a21edaf32064018
MD5 a078afbfd8ecf6a415aa24b1e5c2e1ba
BLAKE2b-256 f95124fe3c9d734588ed2560b08ae2882afc5e60320e6bc1c4a1d48055513ab7

See more details on using hashes here.

File details

Details for the file adjaxt-0.0.2.15-py3-none-any.whl.

File metadata

  • Download URL: adjaxt-0.0.2.15-py3-none-any.whl
  • Upload date:
  • Size: 33.8 kB
  • Tags: Python 3
  • Uploaded using Trusted Publishing? No
  • Uploaded via: twine/7.0.0 CPython/3.14.4

File hashes

Hashes for adjaxt-0.0.2.15-py3-none-any.whl
Algorithm Hash digest
SHA256 d97b3255da98dffe6c5faf25c5f5a825cdaac708bf01bd9e06916d0f89ce48c2
MD5 90ccf38cd34febc51565951faac04da0
BLAKE2b-256 251341fc9cb2a048b5ac585a7190377fac8cb71c840e2f2c5a6bde7451341e76

See more details on using hashes here.

Release history Release notifications | RSS feed

0.0.2.27

2 files

0.0.2.26

2 files

0.0.2.25

2 files

0.0.2.24

2 files

0.0.2.23

2 files

0.0.2.22

2 files

0.0.2.21

2 files

0.0.2.20

2 files

0.0.2.19

2 files

0.0.2.18

2 files

0.0.2.17

2 files

0.0.2.16

2 files

This release

0.0.2.15 This release

2 files

0.0.2.14

2 files

0.0.2.13

2 files

0.0.2.12

2 files

0.0.2.11

2 files

0.0.2.10

2 files

0.0.2.9

2 files

0.0.2.8

2 files

0.0.2.7

2 files

0.0.2.6

2 files

0.0.2.5

2 files

0.0.2.4

2 files

0.0.2.3

2 files

0.0.2.2

2 files

0.0.2.1

2 files

0.0.2.0

2 files

0.0.1.5

2 files

0.0.1.4

2 files

0.0.1.3

2 files

0.0.1.2

2 files

0.0.1.1

2 files

0.0.1

2 files

Anthropic, PBC Visionary sponsor Bloomberg Visionary sponsor Hudson River Trading Visionary sponsor Meta Visionary sponsor NVIDIA Visionary sponsor Microsoft Sustainability sponsor Depot Continuous Integration AWS Cloud computing and Security Sponsor Datadog Monitoring Fastly CDN Google Download Analytics Sentry Error logging StatusPage Status page