Download Latest Version
v0.13.0 - _dtypes_ module _ adds _AMP_F_ dtype source code.tar.gz (278.8 kB)
Get an email when there's a new version of dfdx
| Name | Modified | Size | Downloads / Week |
|---|---|---|---|
| Parent folder | |||
| README.md | 2023-07-27 | 908 Bytes | |
| v0.13.0 - _dtypes_ module _ adds _AMP_F_ dtype source code.tar.gz | 2023-07-27 | 278.8 kB | |
| v0.13.0 - _dtypes_ module _ adds _AMP_F_ dtype source code.zip | 2023-07-27 | 457.1 kB | |
| Totals: 3 Items | 736.9 kB | 0 | |
What's Changed
- Make
storage_traits::TensorToArraypub by @AndrejOrsula in https://github.com/coreylowman/dfdx/pull/817 - accurate-gelu by @jcrist1 in https://github.com/coreylowman/dfdx/pull/813
- Fixing examples/04-gradients.rs by @coreylowman in https://github.com/coreylowman/dfdx/pull/824
- Moving optim kernels to tensor ops by @coreylowman in https://github.com/coreylowman/dfdx/pull/828
- [Breaking] Adds
AMP<F>dtype by @coreylowman in https://github.com/coreylowman/dfdx/pull/811 - Adding documentation to dtypes module and amp by @coreylowman in https://github.com/coreylowman/dfdx/pull/834
New Contributors
- @AndrejOrsula made their first contribution in https://github.com/coreylowman/dfdx/pull/817
- @jcrist1 made their first contribution in https://github.com/coreylowman/dfdx/pull/813
Full Changelog: https://github.com/coreylowman/dfdx/compare/v0.12.1...v0.13.0