heretic

fix: report a save or upload before re-ablating the model (#488)

#107 -Commit 66a952b49b pushed by vrr

master

2026-10-13 15:37:11 +00:00

0s

fix: skip non-metrics in benchmark output (#486)

#106 -Commit 49b56e5af4 pushed by vrr

master

2026-10-08 21:36:20 +00:00

0s

fix: escape prompts and responses printed with `print_responses` (#482)

#105 -Commit 17f8dff09f pushed by vrr

master

2026-10-07 15:37:05 +00:00

0s

fix: stop `lora_A` from carrying over between trials (#476)

#104 -Commit 208c0ca35b pushed by vrr

master

2026-10-05 21:36:27 +00:00

0s

chore: add new Windows test hashes (#470)

#103 -Commit a09b4eae07 pushed by vrr

master

2026-10-03 15:35:37 +00:00

0s

feat: modifier plugins + ARA (#446)

#102 -Commit 662e4ba27e pushed by vrr

master

2026-10-02 09:45:44 +00:00

0s

feat: add base model to the model card (#459)

#101 -Commit 71e6d5eb38 pushed by vrr

master

2026-10-01 14:29:04 +00:00

0s

feat: A better logic for detecting the thinking prefix in the response. (#423)

#100 -Commit 3521f8648a pushed by vrr

master

2026-09-06 19:34:26 +00:00

0s

feat: add benchmark scorer (#444)

#99 -Commit 95dda4c4db pushed by vrr

master

2026-09-04 19:34:31 +00:00

0s

build(deps): bump nltk from 3.10.0 to 3.10.3 (#442)

#98 -Commit c7a44f0db7 pushed by vrr

master

2026-09-03 16:37:30 +00:00

0s

feat: Show a summary at end of tests (#425)

#97 -Commit bedb94ef11 pushed by vrr

master

2026-08-19 02:43:13 +00:00

0s

chore: bump version to 2.0.0.dev0 (#428)

#96 -Commit 638a583bd8 pushed by vrr

master

2026-08-17 20:55:25 +00:00

0s

docs: update model count in README (#424)

#95 -Commit 1a921522b2 pushed by vrr

master

2026-08-15 14:43:14 +00:00

0s

build(deps): bump aiohttp from 3.14.1 to 3.14.3 (#422)

#94 -Commit 9069d3c754 pushed by vrr

master

2026-08-07 02:43:15 +00:00

0s

build(deps): bump setuptools from 80.9.0 to 83.0.0 (#418)

#93 -Commit 7675b90d64 pushed by vrr

master

2026-07-29 14:03:29 +00:00

0s

build(deps): bump torch from 2.9.1 to 2.13.0 (#411)

#92 -Commit 18eaf9533f pushed by vrr

master

2026-07-26 02:03:45 +00:00

0s

build(deps): bump nltk from 3.9.4 to 3.10.0 (#415)

#91 -Commit 20b7bf7869 pushed by vrr

master

2026-07-23 20:03:41 +00:00

0s

docs: quote pip extras install examples (#395)

#90 -Commit 44228c5c42 pushed by vrr

master

2026-07-18 14:03:25 +00:00

0s

feat: Add alternate CI hash in the valid list (for qwen2.5) (#412)

#89 -Commit 02ebf7a751 pushed by vrr

master

2026-07-17 13:18:29 +00:00

0s

fix: validate scorer instance names in config (#407)

#88 -Commit e7b783ed85 pushed by vrr

master

2026-07-16 20:58:31 +00:00

0s

feat: generic plugin system (#53)

#87 -Commit c8a254b825 pushed by vrr

master

2026-07-08 21:18:22 +00:00

0s

fix: use `W_org` matrix only where needed (#398)

#86 -Commit 7470dfd7af pushed by vrr

master

2026-07-02 17:40:29 +00:00

0s

build(deps): bump pydantic-settings from 2.13.1 to 2.14.2 (#397)

#85 -Commit 680c43e1bf pushed by vrr

master

2026-06-28 23:40:29 +00:00

0s

feat: headless operation + end-to-end tests (#392)

#84 -Commit 0146b2760f pushed by vrr

master

2026-06-27 19:41:40 +00:00

0s

fix: resolve UnicodeEncodeError on Windows during model evaluation (#389)

#83 -Commit 3f68a0d4e5 pushed by vrr

master

2026-06-19 17:28:42 +00:00

0s

feat: let the optimizer disable MLP ablation via a 0 max_weight floor (#387)

#82 -Commit 00185db9fc pushed by vrr

master

2026-06-18 17:12:05 +00:00

0s

fix: correct total trial count when adding additional trials (#385)

#81 -Commit 554a58aa0f pushed by vrr

master

2026-06-18 08:40:36 +00:00

0s

build: bump version to 1.4.0

#80 -Commit 6ea3b8d778 pushed by vrr

master

2026-06-15 17:28:41 +00:00

0s

fix: minor cleanups and improvements

#79 -Commit 6757ada999 pushed by vrr

master

2026-06-14 14:50:05 +00:00

0s

feat: automatically reproduce model from reproduce.json (#326)

#78 -Commit 2fd163f5e4 pushed by vrr

master

2026-06-12 17:28:41 +00:00

0s