Commit Graph

  • 0529dcf6d1 Implement L1/L2 regularization Kevin Wu 2022-07-19 00:51:00 +00:00
  • a6909e50cb Increase default early stop patience, add logging Kevin Wu 2022-07-18 18:00:53 +00:00
  • 89893001db Allow for specifying modulo to dataset Kevin Wu 2022-07-18 02:02:25 +00:00
  • bd40c51a7d Properly implement radian L1 loss Kevin Wu 2022-07-18 01:33:02 +00:00
  • 57fee99bc7 Add Python gitignore Kevin Wu 2022-07-17 18:44:13 +00:00
  • edff1c52f1 Fix output of loss Kevin Wu 2022-07-17 18:41:07 +00:00
  • ba5cd22081 Add configurable loss Kevin Wu 2022-07-17 18:15:07 +00:00
  • 47197a7a14 Angle loss Kevin Wu 2022-07-17 18:05:39 +00:00
  • 4c1c71eada Cath dataloader to load between 0 and 2pi Kevin Wu 2022-07-17 15:20:54 +00:00
  • 76dd0eb5a8 Add lines denoting pi Kevin Wu 2022-07-15 23:57:19 +00:00
  • 47c8273240 Add some logging Kevin Wu 2022-07-15 23:56:42 +00:00
  • fcb0f62c0b Notebook looking at priors + variance schedule Kevin Wu 2022-07-14 03:08:29 +00:00
  • e6200e3975 Add plots of variance schedules Kevin Wu 2022-07-13 22:40:02 +00:00
  • 7b6361c759 Add type annotations Kevin Wu 2022-07-13 21:15:21 +00:00
  • 8b6077e93a Implement ability to query specific time points Kevin Wu 2022-07-13 21:14:22 +00:00
  • dd57a73573 Update bond angle notebook with more samples Kevin Wu 2022-07-13 21:14:04 +00:00
  • 656cd303f9 Notebook comparing generated/real proteins Kevin Wu 2022-07-12 20:43:04 +00:00
  • 5c47826015 Add dependency on seaborn Kevin Wu 2022-07-12 20:32:48 +00:00
  • 268afe26a6 Misc. fixes Kevin Wu 2022-07-12 19:29:32 +00:00
  • 50f35608ae Implement sampling multiple sequences Kevin Wu 2022-07-12 19:02:43 +00:00
  • e42b277692 Helper function to split target num to batches Kevin Wu 2022-07-12 18:57:26 +00:00
  • 5ba439e371 Shape fixes Kevin Wu 2022-07-12 17:57:40 +00:00
  • edbc419a33 Initial sampling code Kevin Wu 2022-07-12 17:57:35 +00:00
  • 4ff2427ea3 Centralize code for getting beta schedules Kevin Wu 2022-07-12 17:40:54 +00:00
  • 7966b7249b Centralize extract code Kevin Wu 2022-07-12 17:01:21 +00:00
  • 510df0935b Support for gradient clipping Kevin Wu 2022-07-08 19:31:37 +00:00
  • b4f9033e78 Code cleanup Kevin Wu 2022-07-08 19:08:58 +00:00
  • 4f3fce271b Minor fixes ihyperparam scripts Kevin Wu 2022-07-08 19:03:13 +00:00
  • 496c8ea3d1 Handle json config in train script Kevin Wu 2022-07-08 19:02:50 +00:00
  • e570d359ae Hyperparam search params and generator Kevin Wu 2022-07-08 18:46:30 +00:00
  • 9a67996816 Add early stopping Kevin Wu 2022-07-08 18:04:15 +00:00
  • b26a61c227 Fuller lightning integration Kevin Wu 2022-07-08 17:49:40 +00:00
  • 28b09059a0 Incorporate pytorch lightning Kevin Wu 2022-07-08 17:27:10 +00:00
  • 8290b7d802 Record all params to training loop Kevin Wu 2022-07-08 16:14:04 +00:00
  • 247e4f6baf Report validation loss Kevin Wu 2022-07-07 20:50:18 +00:00
  • 9af65a8284 Fix shape error when batch is size 1 Kevin Wu 2022-07-07 20:50:00 +00:00
  • 3377a3e8cc Add data splits to cath training script Kevin Wu 2022-07-07 18:49:16 +00:00
  • 080a81c96e Support for specifying data splits in Cath dataset Kevin Wu 2022-07-07 18:26:06 +00:00
  • 923a9080c0 Doc updates Kevin Wu 2022-07-07 18:11:24 +00:00
  • 175ba7adfc More configurability for variance schedules Kevin Wu 2022-07-07 18:06:11 +00:00
  • 19ce311895 reproducibility Kevin Wu 2022-07-07 17:58:39 +00:00
  • 39ae64b696 Add matplotlib dependency Kevin Wu 2022-07-07 16:50:26 +00:00
  • 71d974db4f Add plot of losses Kevin Wu 2022-07-07 16:34:23 +00:00
  • 93b2be6edf Implement attention mask Kevin Wu 2022-07-07 02:11:50 +00:00
  • e598d54c99 Fix reporting Kevin Wu 2022-07-06 23:17:00 +00:00
  • 31900261b6 Update dataset return values to always be dict Kevin Wu 2022-07-06 23:11:03 +00:00
  • 97dbe9824e Up the batch size! Kevin Wu 2022-07-06 22:54:45 +00:00
  • e7a4d8c12b Configurable training loop Kevin Wu 2022-07-06 22:41:46 +00:00
  • b120009d5e Initial training script Kevin Wu 2022-07-06 22:28:26 +00:00
  • abcf0a8ed7 Fix shapes Kevin Wu 2022-07-06 22:28:19 +00:00
  • 2932858ea5 Properly include time in embeddings Kevin Wu 2022-07-06 22:00:52 +00:00
  • 33dd1e2890 Initial transformer model Kevin Wu 2022-07-06 01:25:36 +00:00
  • 8601dcb27e Fix types Kevin Wu 2022-07-06 01:25:24 +00:00
  • 5d17fe7f67 More shape fixes Kevin Wu 2022-07-06 01:12:07 +00:00
  • 951bf7070d More shape fixes Kevin Wu 2022-07-06 01:09:47 +00:00
  • 35fa28a161 Fix datasets shape to better match 'embeddings' Kevin Wu 2022-07-06 01:09:05 +00:00
  • 1855a54a3c Padding support Kevin Wu 2022-07-06 00:49:46 +00:00
  • d9a40f45c0 Support for loading in a toy dataset Kevin Wu 2022-07-06 00:42:48 +00:00
  • 1fb95639f3 Report lengths of angle vectors Kevin Wu 2022-07-05 21:44:48 +00:00
  • 9a7a3035d2 Adding transformers dependency Kevin Wu 2022-07-05 21:33:56 +00:00
  • 84fda84eb6 Initial implementation of noising inputs Kevin Wu 2022-07-05 21:30:33 +00:00
  • 025029f70f Add beta schedules Kevin Wu 2022-07-05 21:05:32 +00:00
  • 4a78f26fd3 Misc. code cleanup Kevin Wu 2022-07-05 20:48:07 +00:00
  • 006d15fc16 Parallelize computing angles Kevin Wu 2022-07-05 20:41:46 +00:00
  • 242632b765 Isolate code for coords to angles Kevin Wu 2022-07-05 19:11:59 +00:00
  • a90e8bb507 Count errors Kevin Wu 2022-07-05 18:50:19 +00:00
  • 41d224c4de Trim trialing nans from coords Kevin Wu 2022-07-05 17:24:47 +00:00
  • 4208a695f5 Make return a single tensor Kevin Wu 2022-07-05 16:44:48 +00:00
  • 57603e6959 Make cath dataset to consecutive amino acids Kevin Wu 2022-07-05 16:38:02 +00:00
  • d762bfcae1 Initial implementation of Cath dataset Kevin Wu 2022-07-02 07:07:37 +00:00
  • 7a908098d7 remove submodule Kevin Wu 2022-07-02 05:46:50 +00:00
  • e5d9219392 Add dependency on sequence-models Kevin Wu 2022-07-02 05:33:49 +00:00
  • 54560a74d5 Initial env Kevin Wu 2022-07-02 05:31:05 +00:00
  • 1ce3f31da0 Add submodule for protein sequence models Kevin Wu 2022-07-02 01:17:26 +00:00
  • 307ae3b584 Add script to download cath data Kevin Wu 2022-07-01 22:52:05 +00:00
  • 2a1585be98 Add empty datasets file Kevin Wu 2022-06-30 18:22:21 +00:00
  • 5281578399 LICENSE updated to template Microsoft Open Source 2022-06-30 08:27:16 -07:00
  • 88746c89cc README.md updated to template Microsoft Open Source 2022-06-30 08:27:15 -07:00
  • ae24d65adf SUPPORT.md committed Microsoft Open Source 2022-06-30 08:27:14 -07:00
  • f04a68e835 SECURITY.md committed Microsoft Open Source 2022-06-30 08:27:13 -07:00
  • 56f0cf379b CODE_OF_CONDUCT.md committed Microsoft Open Source 2022-06-30 08:27:12 -07:00
  • 2bf637e433 Initial commit Kevin Kaichuang Yang 2022-06-30 11:24:39 -04:00