jax-jumpy

library1.0.0pypypi✓ verified 79d ago

Jumpy provides a common backend for JAX or NumPy, allowing code to run seamlessly on either. Currently at version 1.0.0 with irregular releases. It wraps array operations, random functions, and device management.

ai-ml

pip install jax-jumpy

github ↗homepage ↗

INSTALL

IMPORT

SIG · JAX-JUMPY

jax-jumpy

ai-mlpythonv1.0.0

harness data pending

Install & Compatibility

Where this runs

No compatibility data collected yet for this library.

Code

Verified usage

Verified import paths — ran on the pinned version, not inferred.

jnp

✓ from jumpy import numpy as jnp

✗ from jumpy import jnp

core

✓ from jumpy import core

lax

✓ from jumpy import lax

Basic usage: import jumpy and use jp as a drop-in replacement for numpy or jax.numpy.

import jumpy as jp # Use jnp for array operations x = jp.ones((3, 4)) y = jp.zeros((3, 4)) z = jp.dot(x, y.T) print(z) # Use random module key = jp.random.PRNGKey(0) rand_vals = jp.random.uniform(key, (5,)) print(rand_vals)

Debug

Known issues

breakingjax-jumpy 1.0.0 changed the backend detection; code relying on `jumpy.backend` might break.

fix

Use `jumpy.which_backend()` or check `jumpy.backend_name` instead.

affects: >=1.0.0

deprecated`jumpy.random.RandomState` is deprecated in favor of explicit PRNGKey usage.

fix

Use `key = jumpy.random.PRNGKey(seed)` and pass key to functions.

affects: >=1.0.0

gotchaJumpy does not support all functions from both numpy and jax; missing functions raise AttributeError.

fix

Check the documentation for supported functions; contributions welcome.

affects: all

Upgrade

Version history

1.0.0latest on PyPI · released Mar 17, 2023

Audit

Dependencies

jaxoptionalOptional backend for JAX acceleration

numpyrequiredMain backend used when JAX is not available

Agent activity

11 hits · last 30 days

node

Resources

githubhttps://github.com/Farama-Foundation/Jumpy ↗

homepagehttps://github.com/Farama-Foundation/Jumpy ↗