Factual package intelligence from PyPI
Memory-efficient fused linear cross-entropy loss for JAX. Computes cross_entropy(x @ w.T, labels) without materializing the full logits tensor.
pip install jax-cce
PyPI declares 1 unique dependency rule for this release. Environment markers are shown when supplied by the project.
jax-cce publishes 1 wheel and 1 source archive for version 0.1.0. Wheel platform tags: any.
No version-specific Python classifiers are declared.
PyPI lists 1 release with files. The first dated release is ; 1 release falls within the 365 days preceding the latest dated release. The current release files were uploaded on .