OpenAdapt’s open-source ML toolkit for training and evaluating general multimodal GUI-action models.
active 2025-12-09 → 2026-07-28 (UTC)
Activity over time
Daily event counts in the loaded window
Line chart, 232 days from 2025-12-09 to 2026-07-28. Pushes: 115 total, peak 11 in a day. Pull requests: 15 total, peak 2 in a day. Issues: 16 total, peak 13 in a day. Comments: 11 total, peak 5 in a day. Stars: 1 total, peak 1 in a day.
- Pushes
- Pull requests
- Issues
- Comments
- Stars
Stars, PRs, issues and forks are under-captured in the later part of this window. GH Archive progressively stopped capturing non-push events during 2026 — −95% or worse by the end of the window. Every series here except Pushes fades for that reason, so a decline above reflects the archive, not this repository. Pushes stay reliable throughout, so read them, and the contributor counts derived from them, as the real signal. Data health has the measurements.
Top contributors
Pushes, PRs, issues, reviews and comments — stars and forks excluded, so this is contribution rather than popularity
| Contributor | Contributions | Pushes | PRs | Comments |
|---|---|---|---|---|
| abrichr | 156 | 113 | 15 | 11 |
| github-actions[bot] | 2 | 2 | 0 | 0 |
Recent activity
Latest issues, pull requests and releases
- Pull request#59abrichr2026-03-22 03:13
- Pull request#56abrichr2026-03-18 06:35
- Pull request#54abrichr2026-03-16 22:57
- Releaseabrichr2026-03-04 04:33v0.14.1
- Pull request#52abrichr2026-03-03 23:28
- Releaseabrichr2026-03-03 02:06v0.12.0
- Issue comment#47abrichr2026-03-03 00:51perf(grpo): use JSONL append-only format for training log
- Issue#46abrichr2026-03-03 00:51fix(grpo): inner tokenizer extraction bypasses processor preprocessing
- Issue comment#46abrichr2026-03-03 00:51fix(grpo): inner tokenizer extraction bypasses processor preprocessing
- Issue#44abrichr2026-03-03 00:51refactor(grpo): replace _grpo_raw_text monkey-patch with proper data structure
- Issue comment#44abrichr2026-03-03 00:51refactor(grpo): replace _grpo_raw_text monkey-patch with proper data structure
- Issue comment#41abrichr2026-03-03 00:51perf(grpo): use PEFT named adapters for reference policy
- Issue#35abrichr2026-03-03 00:51feat(grpo): replace custom training loop with TRL GRPOTrainer
- Issue comment#35abrichr2026-03-03 00:51feat(grpo): replace custom training loop with TRL GRPOTrainer
- Issue#48abrichr2026-03-03 00:25feat(grpo): add gradient accumulation across multiple groups
- Issue#45abrichr2026-03-03 00:25fix(grpo): VLM processor metadata misalignment after token concatenation
- Issue#43abrichr2026-03-03 00:24fix(grpo): align prompt format across SFT, GRPO, and CoT warmup
- Issue#40abrichr2026-03-03 00:20feat(grpo): add checkpoint resume
- Issue#39abrichr2026-03-03 00:20test(grpo): add training loop tests with mocked environment
- Issue#39abrichr2026-03-03 00:20test(grpo): add training loop tests with mocked environment
- Issue#40abrichr2026-03-03 00:20feat(grpo): add checkpoint resume
- Issue#37abrichr2026-03-03 00:20feat(grpo): extend action types (scroll, key, noop)
- Issue#38abrichr2026-03-03 00:20feat(grpo): add pluggable reward functions / verifier registry
- Issue#38abrichr2026-03-03 00:20feat(grpo): add pluggable reward functions / verifier registry
- Releaseabrichr2026-02-24 20:18v0.11.1
Totals cover only the window loaded into ClickHouse and count events, not GitHub's lifetime totals — 1 stars here means stars gained during the window, not the repo's star count.