mirrors
/
mmselfsup
mirror of https://github.com/open-mmlab/mmselfsup.git
1
0
Fork 0
Code Issues Packages Projects Releases Wiki Activity
334 Commits
5 Branches
19 Tags
8 MiB
Commit Graph

53 Commits (dde842bb59aa75a75bd0dcade5202536e5e500f3)

Author SHA1 Message Date
Yixiao Fang 9ef4b9a3d8
[Enhance] enhance MIM function (#152) 
* [Docs] update mminstall.txt

* [Docs] add MANIFEST.in

* [Docs] add metafile

* [WIP]: temporal commit

* [Enhance] Try to create a symbolic link on windows.

* [Docs] add metefile of algorithms
 2021-12-17 16:39:03 +08:00
Yuan Liu a716eea5de
Revert "[WIP] Support installation and training by mim (#144)" (#151) 
This reverts commit 240dd7318e.
 2021-12-17 12:34:36 +08:00
Yixiao Fang 240dd7318e
[WIP] Support installation and training by mim (#144) 
* [Docs] update mminstall.txt

* [Docs] add MANIFEST.in

* [Docs] add metafile
 2021-12-17 12:22:59 +08:00
Yuan Liu febe511b67
[Refactor]: Refactor the directory structure of docs (#146) 
* fix links of docs

* [Docs]: Use shared menu from theme instead

* [Refactor]: Refactor the directory structure of docs

* [Fix]: Fix lint

* [Fix]: Fix __version__.py file link bug

Co-authored-by: fangyixiao18 <fangyx18@hotmail.com>
 2021-12-16 20:29:57 +08:00
Jiahao Xie 6320bba34a
[Docs] update benchmark results (#17) 
* [Docs] add partial results

* [Docs] update results
 2021-12-16 11:32:04 +08:00
Yixiao Fang 4912f52b0f
[Refactor] fix link in .md (#8) 2021-12-15 19:44:34 +08:00
Jiahao Xie f537bfa043
Merge pull request #4 from Jiahao000/fyx_dev 
[Refactor] refactor the codebase
 2021-12-15 19:18:54 +08:00
Jiahao000 67716e2011 [Refactor] add benchmark configs 2021-12-15 19:09:54 +08:00
fangyixiao18 371b5d5b05 [Refactor] refactor selfsup config files 2021-12-15 19:05:11 +08:00
Jiahao000 dfaa8215ae [Refactor] refactor msic files 2021-12-15 18:51:33 +08:00
Jiahao000 9325a85d9f [Refactor] remove old benchmark configs 2021-12-15 18:39:37 +08:00
Jiahao000 ef6b158db5 [Refactor] add benchmark configs 2021-12-15 18:38:19 +08:00
xieenze 3d968392af update model zoo 2020-12-25 10:58:39 +08:00
xieenze 34a5455f60 add byol r50_bs2048_accumulate2_ep200_fp16.py 2020-12-25 10:32:24 +08:00
cxx 15706cbed2 Support prefetching on the linear classification task (False by default). 2020-12-16 14:36:41 +08:00
cxx 796cd68297 Support prefetching data to speed up IO. 2020-12-01 18:26:27 +08:00
hezijian 8e996175b1 fix config 2020-10-21 14:16:18 +08:00
hezijian d542115530 fix the wrong config name 2020-10-21 14:12:31 +08:00
hezijian 0d185ce2cd add byol with sync-bn and accmulate gradient 2020-10-21 14:07:09 +08:00
Xiaohang Zhan d191601da6
Merge pull request #46 from guochengqian/patch-1 
Create r50_withoutsobel.py for deepclustering experiment
 2020-10-14 22:46:44 +08:00
xieenze d4105d7ece replace opencv blur to pil blur 2020-09-27 15:00:50 +08:00
Guocheng Qian afcf21c179
Create r50_withoutsobel.py 2020-09-24 22:34:22 +03:00
xieenze 5e782b3291 update 2020-09-09 14:41:56 +08:00
xiaohangzhan bc05268461 fix moco v2 hue 2020-09-06 14:59:57 +08:00
Jiahao000 d77849849f restore odc config to default 440 epochs 2020-09-03 13:34:22 +08:00
Jiahao000 c3f7d6adf6 update places205 linear classification config 2020-09-01 00:06:55 +08:00
Jiahao000 ab1b26deb6 update places205 linear classification config 2020-08-13 23:50:14 +08:00
xieenze 0cf60e1101 update modelzoo 2020-08-13 00:09:20 +08:00
xiaohangzhan c873fee832 byol default setting 2020-08-08 23:16:09 +08:00
xiaohangzhan 64e27eb9b7 fix BYOL 2020-08-04 13:33:16 +08:00
xiaohangzhan 7232ec557a rename eval config 2020-07-28 00:16:14 +08:00
xiaohangzhan b2f0c075fa linear eval syncbn 2020-07-25 12:14:47 +08:00
xiaohangzhan 6b4cd7a987 Merge branch 'master' of github.com:open-mmlab/OpenSelfSup 2020-07-23 14:33:57 +08:00
xiaohangzhan beeb228b63 base.py for semi 1percent 2020-07-23 14:33:51 +08:00
xiaohangzhan 4abdd5afe4 fix typo to fit with mmcv 2020-07-23 11:00:18 +08:00
xiaohangzhan 8d7094002f fix imgs_per_gpu in 8GPU multihead linear evaluation 2020-07-19 18:11:57 +08:00
Xiaohang Zhan c94991abe8
8GPUs evaluation for Places 2020-07-19 18:05:14 +08:00
Xiaohang Zhan c96df50269
8GPUs evaluation for Places 2020-07-19 18:03:45 +08:00
xieenze 79fdb8ce54 support MoCo 8GPU linear_cls 2020-07-19 17:06:51 +08:00
xiaohangzhan fde22a2362 NPID: ensure negatives 2020-07-10 14:57:39 +08:00
Jiahao000 972e5107ac relative patch location 2020-07-09 22:32:11 +08:00
Jiahao000 7a4de682a8 relative patch location 2020-07-09 15:34:22 +08:00
Jiahao000 b6f1f0f8c7 relative patch location 2020-07-08 22:30:48 +08:00
xiaohangzhan 8f39b95109 rewrite odc neck; semi 10percent 2020-07-06 22:21:54 +08:00
xiaohangzhan c89239a871 update 2020-06-29 00:10:34 +08:00
xiaohangzhan fd9c4d7e0f imagenet linear evalution: evaluate last layer 2020-06-20 15:56:23 +08:00
xiaohangzhan ffc56f5195 default memcached=True 2020-06-20 14:47:42 +08:00
xiaohangzhan 8455d606c8 simclr non-linear neck with syncbn 2020-06-18 00:37:23 +08:00
xiaohangzhan c30613dd69 fix frozen stages in linear classification 2020-06-17 21:46:34 +08:00
xiaohangzhan a1c322556a benchmark scripts 2020-06-17 01:31:59 +08:00