Releases: open-mmlab/mmengine
Releases · open-mmlab/mmengine
v0.11.0rc2
Update version to 0.11.0rc2 The deployment of pypi landing of 0.11.0rc0 and 0.11.0.rc1 are both failed, we should update the version to 0.11.0.rc2 to deploy the wheel of latest change.
v0.11.0rc1
Just update the index in PyPI, the deploy of v0.11.0rc0 failed...
v0.11.0rc0
What's Changed
- [Enhance] Move
import yapfto function by @HAOCHENYE in #1652 - [Enhance] Suppress import errors for optional optimizers by @HAOCHENYE in #1660
- [Enhance] Use opencv-python-headless by default by @HAOCHENYE in #1663
- [Fix] Fix config bug in python 3.12 by @HAOCHENYE in #1669
- [Fix] Load checkpoint with
weights_only=Flaseby @HAOCHENYE in #1670 - [Test] Fix unittest by @HAOCHENYE in #1671
- [Test] Removed
numpy.compactinnumpy>=2.0.0by @MGAMZ in #1672 - [Fix] Correctly process
compile.disablein config by @MGAMZ in #1674 - [Fix] Replace deprecated
pkg_resourceswithimportlibby @MGAMZ in #1676
New Contributors
Full Changelog: v0.10.7...v0.11.0rc0
MMEngine Release V0.10.7
What's Changed
- Update deploy.yml by @C1rN09 in #1625
- [Enhance] Support trigger ci manually by @HAOCHENYE in #1626
- [Enhance] Support trigger deploy CI manually by @HAOCHENYE in #1627
- [Fix] Fix deploy ci by @HAOCHENYE in #1628
- [Fix] remove torch dependencies in
build_function.pyby @HAOCHENYE in #1632 - fix ci by @tenacioustommy in #1636
Full Changelog: v0.10.6...v0.10.7
MMEngine Release V0.10.6
What's Changed
- [Fix] Fix lint by @HAOCHENYE in #1598
- Fix: return in finally by @tibor-reiss in #1596
- fix error when pytest>=8.2 by @tenacioustommy in #1601
- fix Adafactor optim on torch2.5 and fix compatibility by @tenacioustommy in #1600
- ensure type in cfg by @tenacioustommy in #1602
- Bump version to v0.10.6 by @HAOCHENYE in #1623
New Contributors
- @tibor-reiss made their first contribution in #1596
- @tenacioustommy made their first contribution in #1601
v0.10.5
What's Changed
- [Feature] Support calculating loss during validation by @fanqiNO1 in #1503
- [Fix] Fix check builtin module by @HAOCHENYE in #1571
New Contributors
- @ClowDragon made their first contribution in #1537
Full Changelog: v0.10.4...v0.10.5
MMEngine Release V0.10.4
v0.10.4 (23/4/2024)
New Features & Enhancements
- Support custom
artifact_locationin MLflowVisBackend. by @daavoo in #1505 - Add the supported pytorch versions in README by @zhouzaida in #1512
- Perform evaluation upon training completion by @LZHgrla in #1529
- Enable
exclude_frozen_parametersforDeepSpeedEngine._zero3_consolidated_16bit_state_dictby @LZHgrla in #1517
Bug Fixes
- Fix warning capture by @fanqiNO1 in #1494
- Remove codeowners file by @zhouzaida in #1496
- Fix config of readthedocs by @zhouzaida in #1511
- Delete frozen parameters when using
paramwise_cfgby @LZHgrla in #1441
Docs
- Refine mmengine intro by @zhouzaida in #1479
- Fix typo by @zhouzaida in #1481
- Fix typos and remove fullwidth unicode chars by @evdcush in #1488
- Fix docstring of Config by @MambaWong in #1506
- Fix typo by @hiramf in #1532
New Contributors
Full Changelog: v0.10.3...v0.10.4
MMEngine Releases v0.10.3
New Features & Enhancements
- Add the support for musa device support by @hanhaowen-mt in #1453
- Support
save_optimizer=Falsefor DeepSpeed by @LZHgrla in #1474 - Update the docstring of visualizer by @Anm-pinellia in #1476
Bug Fixes
- Fix
Config.to_dictby @HAOCHENYE in #1465 - Fix the resume of iteration by @LZHgrla in #1471
- Fix
dist.collect_resultsto keep all ranks' elements by @LZHgrla in #1469
Docs
- Add the usage of ProfilerHook by @zhouzaida in #1466
- Fix the nnodes in the doc of ddp training by @XiwuChen in #1462
New Contributors
- @XiwuChen made their first contribution in #1462
- @hanhaowen-mt made their first contribution in #1453
- @Anm-pinellia made their first contribution in #1476
Full Changelog: v0.10.2...v0.10.3
MMEngine Release V0.10.2
v0.10.2 (26/12/2023)
New Features & Enhancements
- Support multi-node distributed training with NPU backend by @shun001 in #1459
- Use
ImportErrorto coverModuleNotFoundErrorby @del-zhenwu in #1438
Bug Fixes
- Fix bug in
load_model_state_dictofBaseStrategyby @SCZwangxiao in #1447 - Fix placement policy in ColossalAIStrategy by @fanqiNO1 in #1440
New Contributors
- @del-zhenwu made their first contribution in #1438
- @SCZwangxiao made their first contribution in #1447
- @shun001 made their first contribution in #1459
Full Changelog: v0.10.1...v0.10.2