Default Branch

66a952b49b · fix: report a save or upload before re-ablating the model (#488) · Updated 2026-10-12 05:23:19 -07:00

Branches

449b351735 · fix: skip non-metrics in benchmark output · Updated 2026-10-07 05:37:26 -07:00

1
1

d128973ca9 · fix: silence warning spam from PyTorch's C++ logging · Updated 2026-10-07 04:35:00 -07:00

2
1

dc1a569470 · chore: add yet more CI test hashes · Updated 2026-10-02 02:04:32 -07:00

7
3

15ad4261be · ci: add yet another alternative model hash · Updated 2026-10-01 01:37:30 -07:00

8
11

3599fba081 · Merge branch 'master' into benchmark-scorer · Updated 2026-09-03 05:10:40 -07:00

12
3

3444a0dd67 · Merge branch 'master' into version-2-dev · Updated 2026-08-17 03:50:28 -07:00

15
2

2fb773adc5 · chore: remove Gemini style guide · Updated 2026-08-17 02:42:34 -07:00

16
1
ara

edc3b12345 · fix(ara): free gradient buffers after optimization (#426) · Updated 2026-08-17 02:27:09 -07:00

85
19

26ea30e117 · fix: use binary mode for hashes everywhere · Updated 2026-06-27 00:52:32 -07:00

30
17

e9bcd86e66 · docs: add automatic reproduction instructions to reproduce README · Updated 2026-06-11 02:01:32 -07:00

37
10

a981ceb50c · Revert "Revert "fix: disable LoRA export for now" (#308)" · Updated 2026-05-02 18:51:36 -07:00

63
1

5c0f344760 · fix: fix remaining issues · Updated 2026-04-23 06:06:01 -07:00

67
7

49294d2abf · Revert "perf: optimize abliteration matrix op (#46)" · Updated 2025-12-06 16:58:19 -08:00    vh

143
1