v0.3
版本发布时间: 2022-11-22 00:27:21
CarperAI/trlx最新发布版本:v0.7.0(2023-06-24 06:21:52)
What's Changed
- Download simulacra by @reciprocated in https://github.com/CarperAI/trlx/pull/62
- Update documentation (first review) by @simoninithomas in https://github.com/CarperAI/trlx/pull/64
- Add ckpt/ to gitignore by @ayulockin in https://github.com/CarperAI/trlx/pull/70
- change version in package to match lib by @cat-state in https://github.com/CarperAI/trlx/pull/73
- Docs by @shahbuland in https://github.com/CarperAI/trlx/pull/71
- [fix] Remove stale options from
ppo_gptj.yml
by @jon-tow in https://github.com/CarperAI/trlx/pull/77 - Add
entity
name config forwandb
logging by @jon-tow in https://github.com/CarperAI/trlx/pull/78 - EXAMPLE : Interpreter grounded Neural Program Synthesis [WIP] by @reshinthadithyan in https://github.com/CarperAI/trlx/pull/81
- Update
TrainConfig
optimizer hyperparameters by @jon-tow in https://github.com/CarperAI/trlx/pull/82 - Add examples tip to contribution guide by @jon-tow in https://github.com/CarperAI/trlx/pull/84
- Fix pipeline's context overflow by @reciprocated in https://github.com/CarperAI/trlx/pull/87
- Refactor PPO objective function by @jon-tow in https://github.com/CarperAI/trlx/pull/88
- Fix slow ilql eval by @reciprocated in https://github.com/CarperAI/trlx/pull/91
- rerun https://github.com/CarperAI/trlx/pull/89 by @cat-state in https://github.com/CarperAI/trlx/pull/92
- Hyperparameter Optimization with Ray Tune and Weights and Biases by @ayulockin in https://github.com/CarperAI/trlx/pull/76
- Update readme instructions by @reciprocated in https://github.com/CarperAI/trlx/pull/93
- Update README to align nomenclature correctness by @ayulockin in https://github.com/CarperAI/trlx/pull/97
- Add optional reward scaling by @reciprocated in https://github.com/CarperAI/trlx/pull/95
- Force class registry via imports by @jon-tow in https://github.com/CarperAI/trlx/pull/100
- Add optional normalization (cont.) by @reciprocated in https://github.com/CarperAI/trlx/pull/98
- Restructure sweeps for reuse by @reciprocated in https://github.com/CarperAI/trlx/pull/102
New Contributors
- @simoninithomas made their first contribution in https://github.com/CarperAI/trlx/pull/64
- @ayulockin made their first contribution in https://github.com/CarperAI/trlx/pull/70
- @reshinthadithyan made their first contribution in https://github.com/CarperAI/trlx/pull/81
Full Changelog: https://github.com/CarperAI/trlx/compare/v0.2...v0.3