Commit Graph

  • 34e6284f95 Update README.md Phil Wang 2020-12-24 10:58:41 -08:00
  • aa9ed249a3 add knowledge distillation with distillation tokens, in light of new finding from facebook ai 0.6.0 Phil Wang 2020-12-24 10:39:15 -08:00
  • ea0924ec96 update readme Phil Wang 2020-12-23 19:06:48 -08:00
  • 59787a6b7e allow for mean pool with efficient version too 0.5.1 Phil Wang 2020-12-23 18:15:40 -08:00
  • 24339644ca offer a way to use mean pooling of last layer 0.5.0 Phil Wang 2020-12-23 17:23:58 -08:00
  • b786029e18 fix the dimension per head to be independent of dim and heads, to make sure users do not have it be too small to learn anything 0.4.0 Phil Wang 2020-12-17 07:43:52 -08:00
  • 9624181940 simplify mlp head 0.3.0 Phil Wang 2020-12-07 14:31:50 -08:00
  • a656a213e6 update diagram Phil Wang 2020-12-04 12:26:28 -08:00
  • f1deb5fb7e Merge pull request #31 from minhlong94/main Phil Wang 2020-11-21 08:05:38 -08:00
  • 3f50dd72cf Update README.md Long M. Lưu 2020-11-21 18:37:03 +07:00
  • ee5e4e9929 Update vit_pytorch.py Long M. Lưu 2020-11-21 18:23:04 +07:00
  • 6c8dfc185e remove float(-inf) as masking value 0.2.7 Phil Wang 2020-11-13 12:25:21 -08:00
  • 4f84ad7a64 authors are now known Phil Wang 2020-11-03 14:28:20 -08:00
  • c74bc781f0 cite Phil Wang 2020-11-03 11:59:05 -08:00
  • dc5b89c942 use einops repeat Phil Wang 2020-10-28 18:13:57 -07:00
  • c1043ab00c update readme Phil Wang 2020-10-26 19:01:03 -07:00
  • 7a214d7109 allow for training on different image sizes, provided images are smaller than what was passed as image_size keyword on init 0.2.6 Phil Wang 2020-10-25 13:17:42 -07:00
  • 6d1df1a970 more efficient 0.2.5 Phil Wang 2020-10-22 22:37:06 -07:00
  • d65a8c17a5 remove dropout from last linear to logits 0.2.4 Phil Wang 2020-10-16 13:58:23 -07:00
  • f7c164d910 assert minimum number of patches 0.2.3 Phil Wang 2020-10-16 12:19:50 -07:00
  • c7b74e0bc3 rename ipy notebook Phil Wang 2020-10-14 10:35:46 -07:00
  • 5b5d98a3a7 dropouts are more specific and aggressive in the paper, thanks for letting me know @hila-chefer Phil Wang 2020-10-14 05:48:27 -07:00
  • 35796104b0 dropouts are more specific and aggressive in the paper, thanks for letting me know @hilach70 0.2.2 Phil Wang 2020-10-14 05:48:27 -07:00
  • b0e4790c24 bump package 0.2.1 Phil Wang 2020-10-13 13:12:19 -07:00
  • 0b2b3fc20c add dropouts Phil Wang 2020-10-13 13:11:32 -07:00
  • ced464dcb4 Update setup.py Phil Wang 2020-10-11 00:06:26 -07:00
  • 5bf45a2d4d Merge pull request #4 from adimyth/main Phil Wang 2020-10-10 19:12:31 -07:00
  • fa32e22855 adds a classification example using 'cats & dogs' data adimyth 2020-10-11 03:15:19 +05:30
  • a0fa41070f norm cls token before sending to mlp head 0.2.0 Phil Wang 2020-10-10 12:08:42 -07:00
  • b298031c17 write up example for using efficient transformers Phil Wang 2020-10-07 19:15:21 -07:00
  • ee1cbbadec write up example for using efficient transformers 0.1.0 Phil Wang 2020-10-07 19:13:25 -07:00
  • d66b29e4cf cleanup stray print 0.0.5 Phil Wang 2020-10-07 11:22:45 -07:00
  • f7123720c3 add masking 0.0.4 Phil Wang 2020-10-07 11:21:03 -07:00
  • f5fffd9e2e remove extraneous line Phil Wang 2020-10-04 15:22:26 -07:00
  • 8fb261ca66 fix a bug and add suggestion for BYOL pre-training 0.0.3 Phil Wang 2020-10-04 14:55:29 -07:00
  • 112ba5c476 update with link to Yannics video Phil Wang 2020-10-04 13:53:47 -07:00
  • f899226d4f add diagram Phil Wang 2020-10-04 12:47:08 -07:00
  • 825a9484d1 small bug fix 0.0.2 Phil Wang 2020-10-04 12:39:51 -07:00
  • ee8088b3ea first commit 0.0.1 Phil Wang 2020-10-04 12:34:44 -07:00
  • 070469d0ac Create python-publish.yml Phil Wang 2020-10-04 12:34:30 -07:00
  • ea03db32f0 Update README.md Phil Wang 2020-10-03 15:49:27 -07:00
  • 30362d50dc Update README.md Phil Wang 2020-10-03 15:49:02 -07:00
  • efb40e0b01 Initial commit Phil Wang 2020-10-03 15:47:26 -07:00