Train LLMs and SLMs while you code with Claude or other Bash agents. BashGym enables you to collect the data from your coding sessions, create synthetic data, train LoRA adapters, and deploy open source agents all in one platform.
Stars
5
Forks
0
Watchers
5
Open Issues
0
Overall repository health assessment
No package.json found
This might not be a Node.js project
222
commits
Merge pull request #15 from GhostPeony/feat/training-strategies-device-mgmt
8914f63View on GitHubfix: add include_failed param to auto-classify for reclassification
742f2e2View on GitHubfix: traces with verification_passed=None incorrectly classified as failed
3053e76View on GitHubMerge pull request #14 from GhostPeony/feat/training-strategies-device-mgmt
d70b1fdView on GitHubdocs: update README with AutoResearch schema evolution, Cascade RL, Nemotron models
c237df2View on GitHubMerge pull request #13 from GhostPeony/feat/training-strategies-device-mgmt
5b54f71View on GitHubfix: CI failures — remote trainer test signature + black formatting
02b7b64View on GitHubMerge remote-tracking branch 'origin/master' into feat/training-strategies-device-mgmt
eb25d48View on GitHubchore: stage uncommitted changes (gitignore, HF script adapter, test config)
e5d2958View on GitHubfix: strategy button overflow + add Nemotron models + roadmap in README
290a906View on GitHubfeat: AutoCurriculum Compiler Phase 3 — Cascade RL Scheduler + MOPD distillation
33d747eView on GitHubfeat: AutoCurriculum Compiler Phase 1+2 — Data Designer upgrade + SchemaResearcher
a5eab53View on GitHub