NeptuneVisBackend
, DVCLiveVisBackend
and AimVisBackend
. Refer to Visualization Backends for more details.NeptuneVisBackend
by @wangerlie in https://github.com/open-mmlab/mmengine/pull/1311DVCLiveVisBackend
by @RangeKing in https://github.com/open-mmlab/mmengine/pull/1336AimVisBackend
by @RangeKing in https://github.com/open-mmlab/mmengine/pull/1347Adafactor
Optimizer by @okotaku in https://github.com/open-mmlab/mmengine/pull/1361Runner
for each epoch by @ShuRaymond in https://github.com/open-mmlab/mmengine/pull/1292optimizer_cls
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1324A total of 21 developers contributed to this release. Thanks @LZHgrla, @wangerlie, @wangg12, @RangeKing, @hiyyg, @LRJKD, @KevinNuNu, @zeyuanyin, @Desjajja, @ShuRaymond, @okotaku, @crazysteeaam, @6Vvv, @NrealLzx, @YinAoXiong, @huaibovip, @xuuyangg, @Dominic23331, @fanqiNO1, @HAOCHENYE, @zhouzaida
collate_fn
for FlexibleRunner by @LZHgrla in https://github.com/open-mmlab/mmengine/pull/1284vis_backends
when save_dir
is not set by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1289A total of 3 developers contributed to this release. Thanks @HAOCHENYE, @zhouzaida, @LZHgrla
efficient_conv_bn_eval
for efficient convolution and batch normalization. See save memory on gpu for more detailsefficient_conv_bn_eval
for memory saving convolution and batch normalization by @youkaichao in https://github.com/open-mmlab/mmengine/pull/1202, https://github.com/open-mmlab/mmengine/pull/1251 and https://github.com/open-mmlab/mmengine/pull/1259BaseModule
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1263save_begin
to control when to save checkpoints by @KerwinKai in https://github.com/open-mmlab/mmengine/pull/1271ConfigDict.items
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1272, https://github.com/open-mmlab/mmengine/pull/1270 and https://github.com/open-mmlab/mmengine/pull/1088loop_stage
runtime information in message_hub
by @zhouzaida in https://github.com/open-mmlab/mmengine/pull/1277vis_backends
will not be used when save_dir
is None
by @Xinyu302 in https://github.com/open-mmlab/mmengine/pull/1275A total of 9 developers contributed to this release. Thanks @HAOCHENYE, @youkaichao, @josh6688, @i-aki-y, @mmeendez8, @zhouzaida, @gachiemchiep, @KerwinKai, @Li-Qingyun
lazy_import
by @Li-Qingyun in https://github.com/open-mmlab/mmengine/pull/1239Config.dump
and support converting Lazyxxx to string in ConfigDict.to_dict
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1232_get_ignored_modules
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1235Support training with FSDP and DeepSpeed. Refer to the example for more detailed usages.
Introduce the pure Python style configuration file:
Refer to the tutorial for more detailed usages.
FlexibleRunner
and Strategies
, and support training with DeepSpeed by @zhouzaida in https://github.com/open-mmlab/mmengine/pull/1183A total of 9 developers contributed to this release. Thanks @evdcush, @zhouzaida, @AkideLiu, @joihn, @HAOCHENYE, @edkair, @alexander-soare, @syo093c, @zgzhengSEU
ClearML
to record experiment dataSophia
optimizersann_file
, data_root
is None
for BaseDataset
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/850define_metric
arguments by @i-aki-y in https://github.com/open-mmlab/mmengine/pull/1099_metadata
of state_dict
when saving checkpoints by @Bomsw in https://github.com/open-mmlab/mmengine/pull/1131ProfileHook
cannot profile ddp-training by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1140cuda_visible_devices
by @adrianjoshua-strutt in https://github.com/open-mmlab/mmengine/pull/1164CUDA_VISIBLE_DEVICES
in logger @Xiangxu-0103 in https://github.com/open-mmlab/mmengine/pull/1172infer.md
by @Hongru-Xiao in https://github.com/open-mmlab/mmengine/pull/1121tutorials/runner.md
by @gy-7 in https://github.com/open-mmlab/mmengine/pull/1146data_element.md
by @xin-li-67 in https://github.com/open-mmlab/mmengine/pull/1067A total of 19 developers contributed to this release. Thanks @Hongru-Xiao @i-aki-y @Bomsw @KickCellarDoor @zhouzaida @YQisme @gachiemchiep @CescMessi @W-ZN @Ginray @adrianjoshua-strutt @CokeDong @xin-li-67 @Xiangxu-0103 @HAOCHENYE @Shiyang980713 @TankNee @zimonitrome @gy-7
MLflowVisBackend
by @sh0622-kim in https://github.com/open-mmlab/mmengine/pull/878worker_init_fn
in dataloader config by @shufanwu in https://github.com/open-mmlab/mmengine/pull/1038input_shape
for get_model_complexity_info
by @sjiang95 in https://github.com/open-mmlab/mmengine/pull/1065MessageHub
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1098_load_state_dict_post_hooks
in load_state_dict
by @mzr1996 in https://github.com/open-mmlab/mmengine/pull/1103log_with_hierarchy
is True
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1085collate_fn
could not be a function object by @zhouzaida in https://github.com/open-mmlab/mmengine/pull/1093optimizer.state
could be saved in cuda:0 by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/966A total of 17 developers contributed to this release. Thanks @enkilee, @JunweiZheng93, @sh0622-kim, @jsrdcht, @SheffieldCao, @josh6688, @mzr1996, @zhouzaida, @shufanwu, @Luo-Yihang, @C1rN09, @LEFTeyex, @zccjjj, @Ginray, @HAOCHENYE, @sjiang95, @luomaoling
repeat_factors
in ClassBalancedDataset
by @BIGWangYuDong in https://github.com/open-mmlab/mmengine/pull/1048DistributedDataParallel
that define init_weights
during initialization by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1045A total of 3 developers contributed to this release. Thanks @kitecats, @BIGWangYuDong, @HAOCHENYE
torch.compile
mmpretrain
to the MODULE2PACKAGE
. by @mzr1996 in https://github.com/open-mmlab/mmengine/pull/1002get_device
in the compiled model by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/1004torch.compile
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1021print_log
and logger.info(warning)
by @LEFTeyex in https://github.com/open-mmlab/mmengine/pull/997data_time
and metric in logs by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/1025A total of 7 developers contributed to this release. Thanks @LEFTeyex, @KerwinKai, @mzr1996, @evdcush, @C1rN09, @HAOCHENYE, @zhouzaida
EarlyStoppingHook
to stop training when the metric does not improvetorch.compile
in Runner by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/976EarlyStoppingHook
by @nijkah in https://github.com/open-mmlab/mmengine/pull/739FUNCTIONS
root Registry by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/983AmpOptimWrapper
by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/960vis_backend
with prefix by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/972ManagerMixin
built instance with duplicate name. by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/990pytorch < 1.10.0
by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/975matmul
with single dimension matrix by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/970A total of 10 developers contributed to this release. Thanks @xin-li-67, @acdart, @enkilee, @YuetianW, @luomaoling, @nijkah, @VoyagerXvoyagerx, @zhouzaida, @HAOCHENYE, @C1rN09
Apex
with ApexOptimWrapper
Lion
optimizer.Apex
with ApexOptimWrapper
by @xcnick in https://github.com/open-mmlab/mmengine/pull/742timeout
in dist configuration by @apacha in https://github.com/open-mmlab/mmengine/pull/877repr
of Registry by @zhouzaida in https://github.com/open-mmlab/mmengine/pull/942collect_env
raise errors and stops programs by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/944np.ScalarType
data in message_hub by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/898Visualizer
by @KevinNuNu in https://github.com/open-mmlab/mmengine/pull/887DefaultOptimWrapperConstructor
when the shared parameters do not require the grad by @HIT-cwh in https://github.com/open-mmlab/mmengine/pull/903shedule
by @Dai-Wenxun in https://github.com/open-mmlab/mmengine/pull/936A total of 15 developers contributed to this release. Thanks @Eiuyc, @xcnick, @KevinNuNu, @XWHtorrentx, @tonysy, @zhouzaida, @Xiangxu-0103, @Dai-Wenxun, @jbwang1997, @apacha, @C1rN09, @HIT-cwh, @vansin, @HAOCHENYE, @luomaoling.
BaseInferencer
to provide a general inference interfaceReduceOnPlateauParamScheduler
to adjust learning rate by metricReduceOnPlateauParamScheduler
by @LEFTeyex in https://github.com/open-mmlab/mmengine/pull/819BaseInferencer
to provide general interface by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/874Config
cannot parse base config when there is .
in tmp path, etc. tmp/a.b/c
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/856PretrainedIinit
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/764PKG2PROJECT
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/872advanced_tutorials/logging.md
by @RangeKing in https://github.com/open-mmlab/mmengine/pull/861train_a_gan
to EN by @yaqi0510 in https://github.com/open-mmlab/mmengine/pull/860fileio.md
by @Xiangxu-0103 in https://github.com/open-mmlab/mmengine/pull/869inferencer
. by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/884A total of 8 developers contributed to this release. Thanks @LEFTeyex, @RangeKing, @yaqi0510, @Xiangxu-0103, @wangjiangben-hw, @mzr1996, @zhouzaida, @HAOCHENYE.
ProfileHook
to profile the running processconf_path
in PetrelBackend by @sunyc11 in https://github.com/open-mmlab/mmengine/pull/774--cfg-options
. by @mzr1996 in https://github.com/open-mmlab/mmengine/pull/759OptimWrapper.update_params
by @twmht in https://github.com/open-mmlab/mmengine/pull/796get_torchvision_model
compatible with torch 1.13 by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/793flat_decay_mult
and fix bias_decay_mult
of depth-wise-conv in DefaultOptimWrapperConstructor
by @RangiLyu in https://github.com/open-mmlab/mmengine/pull/771hub.get_model
fails on some MMCls models by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/784BaseModel.to
and BaseDataPreprocessor.to
to make them consistent with torch.nn.Module
by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/783ZeroRedundancyOptimizer
ambiguous error with param groups when PyTorch < 1.12.0 by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/818progress
argument to load_from_http
by @austinmw in https://github.com/open-mmlab/mmengine/pull/770contributing.md
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/754contributing.md
by @RangeKing in https://github.com/open-mmlab/mmengine/pull/792Visualizer.draw_polygons
by @lyviva in https://github.com/open-mmlab/mmengine/pull/798save_gpu_memory.md
by @xin-li-67 in https://github.com/open-mmlab/mmengine/pull/803registry.md
by @xin-li-67 in https://github.com/open-mmlab/mmengine/pull/817BaseDataElement
by @Xiangxu-0103 in https://github.com/open-mmlab/mmengine/pull/836structures
by @Xiangxu-0103 in https://github.com/open-mmlab/mmengine/pull/840optim_wrapper.md
by @xin-li-67 in https://github.com/open-mmlab/mmengine/pull/833initialized.md
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/844A total of 16 developers contributed to this release. Thanks @BayMaxBHL, @RangeKing, @Xiangxu-0103, @xin-li-67, @twmht, @shanmo, @sunyc11, @lyviva, @austinmw, @xiexinch, @mzr1996, @RangiLyu, @MambaWong, @C1rN09, @zhouzaida, @HAOCHENYE
TestRunnerTestCase
to build a Runner instance. by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/631eta_min_ratio
in CosineAnnealingParamScheduler
by @cir7 in https://github.com/open-mmlab/mmengine/pull/725revert_sync_batchnorm
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/695distributed_training.py
in examples by @PingHGao in https://github.com/open-mmlab/mmengine/pull/700CheckpointLoader.load_checkpoint
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/685add_graph
is not called bug by @shenmishajing in https://github.com/open-mmlab/mmengine/pull/632BaseDataset
typo about lazy initialization by @MengzhangLI in https://github.com/open-mmlab/mmengine/pull/733print
with print_log
for those functions called by runner by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/686evaluator.md
by @sanbuphy in https://github.com/open-mmlab/mmengine/pull/675utils.md
@sanbuphy in https://github.com/open-mmlab/mmengine/pull/702ZeroRedundancyOptimizer
by @nijkah in https://github.com/open-mmlab/mmengine/pull/551BaseDataPreprocessor
by @shenmishajing in https://github.com/open-mmlab/mmengine/pull/618clip_grad
, and support clip grad by value. by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/513register_all_modules
in Registry.get()
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/541_save_to_state_dict
implemented in mmengine by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/610ignore_keys
in ConcatDataset by @BIGWangYuDong in https://github.com/open-mmlab/mmengine/pull/556changelog.md
in chinese documents. by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/606BaseDataset
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/656hub.get_model
by @zengyh1900 in https://github.com/open-mmlab/mmengine/pull/659docs/zh_cn/advanced_tutorials/visualization.md
by @MambaWong in https://github.com/open-mmlab/mmengine/pull/616DefaultOptimWrapperConstructor
by @triple-Mu in https://github.com/open-mmlab/mmengine/pull/644Config
docstring by @sanbuphy in https://github.com/open-mmlab/mmengine/pull/654docs/zh_cn/tutorials/config.md
by @Xiangxu-0103 in https://github.com/open-mmlab/mmengine/pull/596docs/zh_cn/tutorials/model.md
by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/598eta_min
in CosineRestartParamScheduler
by @Z-Fran in https://github.com/open-mmlab/mmengine/pull/639BaseDataPreprocessor.cast_data
could not handle string data by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/602autocast
compatible with mps by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/587is_model_wrapper
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/640VisBackend.add_config
is not called by @shenmishajing in https://github.com/open-mmlab/mmengine/pull/613strict_load
of EMAHook to False by default by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/642open
encoding problem of Config in Windows by @sanbuphy in https://github.com/open-mmlab/mmengine/pull/648pip upgrade
CI by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/622FileIO
but without breaking bc by @zhouzaida in https://github.com/open-mmlab/mmengine/pull/533torch.lerp\_()
to speed up EMA by @RangiLyu in https://github.com/open-mmlab/mmengine/pull/519BN
to SyncBN
by config by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/506instance_data.py
by @Dai-Wenxun in https://github.com/open-mmlab/mmengine/pull/530config.md
by @Zhengfei-0311 in https://github.com/open-mmlab/mmengine/pull/562LogProcessor
does not smooth loss if the name of loss doesn't start with loss
by @liuyanyi in
https://github.com/open-mmlab/mmengine/pull/539detect_anomalous_params
in MMSeparateDistributedDataParallel
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/588filename_tmpl
argument by @C1rN09 in https://github.com/open-mmlab/mmengine/pull/518FSDP
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/520EMAHook
by @okotaku in https://github.com/open-mmlab/mmengine/pull/507EMAHook
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/523MultiProcessTestCase
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/535Registry
by @MambaWong in https://github.com/open-mmlab/mmengine/pull/536_save_to_state_dict
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/542Runner.resume
by @HAOCHENYE in https://github.com/open-mmlab/mmengine/pull/511get
instead of pop
to dump runner_type
in build_runner_from_cfg
by @nijkah in https://github.com/open-mmlab/mmengine/pull/549registry.md
by @vansin in https://github.com/open-mmlab/mmengine/pull/514MMFullyShardedDataParallel
by @yhna940 in https://github.com/open-mmlab/mmengine/pull/569此处可能存在不合适展示的内容,页面不予展示。您可通过相关编辑功能自查并修改。
如您确认内容无涉及 不当用语 / 纯广告导流 / 暴力 / 低俗色情 / 侵权 / 盗版 / 虚假 / 无价值内容或违法国家有关法律法规的内容,可点击提交进行申诉,我们将尽快为您处理。