Version 0.0.10

What's Changed

• New version by @chaoming0625 in #1
• Add etrace learning algorithm with O(n^2) complexity, and hybrid O(n) & O(n^2) algorithms by @chaoming0625 in #2
• for pre- and post-eligibility trace, add independent decay factors and time constants by @chaoming0625 in #3
• algorithm updates by @chaoming0625 in #4
• fix numpy array() compatability by @chaoming0625 in #5
• support vjp and jvp diagonal Jacobian by @chaoming0625 in #6
• New updates by @chaoming0625 in #7
• fix bugs by @chaoming0625 in #8
• upgrade to support brainstate by @chaoming0625 in #9
• New algorithm by @chaoming0625 in #10
• fix by @chaoming0625 in #11
• Updates by @chaoming0625 in #12
• Conditionally support control flows, including scan, while, and cond by @chaoming0625 in #13
• fix bug by @chaoming0625 in #14
• add .get_etrace_of() function to obtain the etrace data of the give… by @chaoming0625 in #15
• V 0.0.2 by @chaoming0625 in #16
• compatible with brainstate>=0.1.0 by @chaoming0625 in #17
• New version by @chaoming0625 in #18
• Enable the consistent training of SingleStepData and MultiStepData by @chaoming0625 in #19
• Add ElementWiseParam and update modules by @chaoming0625 in #20
• Upgraded version of brainscale by @chaoming0625 in #21
• Update doc, code, tests, and imports by @chaoming0625 in #22
• fix names by @chaoming0625 in #23
• Update docs by @chaoming0625 in #24
• Update by @chaoming0625 in #25
• Update brainscale codebase by @chaoming0625 in #26
• Refactor weight gradient functions for clarity and accuracy by @chaoming0625 in #27
• Enable all examples work by @chaoming0625 in #28
• Fix by @chaoming0625 in #29
• Update examples by @chaoming0625 in #30
• Fix documentation by @chaoming0625 in #31
• Support CSR sparse operations by @chaoming0625 in #32
• Update systematic documentation by @chaoming0625 in #33
• Compatiable JAX >= 0.7.0 by @chaoming0625 in #34
• Add VJP support for weight functions in ETrace operators by @chaoming0625 in #35
• ⬆️ Bump actions/checkout from 4 to 5 by @dependabot[bot] in #36
• ⬆️ Bump actions/setup-python from 5 to 6 by @dependabot[bot] in #37
• Add mode support for elementwise parameters and improve BatchNorm import by @chaoming0625 in #38
• Update publish workflow by @chaoming0625 in #39

New Contributors

• @chaoming0625 made their first contribution in #1
• @dependabot[bot] made their first contribution in #36

Full Changelog: https://github.com/chaobrain/brainscale/commits/v0.0.10