czzhangheng
/
DCRNN
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Commit Graph

  • d92490b808 Moved comparison to the top main Chintan Shah 2019-11-03 20:17:58 -0500
  • a3b76a56c6 Added PyTorch vs TF MAE comparison Chintan Shah 2019-11-03 20:14:40 -0500
  • bb32eb0f46 Changed README to reflect PyTorch implementation Chintan Shah 2019-10-30 12:30:45 -0400
  • 073f1d4a6e updated epeoch num Chintan Shah 2019-10-08 17:32:16 -0400
  • b2d2b21dbd logging MAE Chintan Shah 2019-10-08 13:46:34 -0400
  • f720529ac9 demoing with test data Chintan Shah 2019-10-08 13:11:43 -0400
  • d2913fd6f1 converting to CPU Chintan Shah 2019-10-08 13:09:02 -0400
  • f92e7295a0 added run_demo_pytorch Chintan Shah 2019-10-08 13:05:49 -0400
  • dda7013f07 returning predictions from the model during eval at every timestep Chintan Shah 2019-10-08 12:56:20 -0400
  • 46b552e075 updated eps value Chintan Shah 2019-10-08 02:44:13 -0400
  • 02fb2430f0 removed logging of every horizon Chintan Shah 2019-10-08 02:10:59 -0400
  • 765142de00 refactored Chintan Shah 2019-10-07 20:55:26 -0400
  • 5d7694e293 logging to 4 decimals Chintan Shah 2019-10-07 20:54:58 -0400
  • f6e6713f74 fixed range bug Chintan Shah 2019-10-07 20:40:54 -0400
  • 2560e1d954 Added per timestep loss Chintan Shah 2019-10-07 20:03:00 -0400
  • 3d93008a3e improved saving and restoring of model Chintan Shah 2019-10-07 11:56:14 -0400
  • a5a1063160 fixed docstring Chintan Shah 2019-10-07 10:48:48 -0400
  • 5509e9aae5 Ensured all parameters are added to the optimizer Chintan Shah 2019-10-07 09:47:38 -0400
  • de42a67391 added logging statement Chintan Shah 2019-10-07 07:59:41 -0400
  • 941675d6a7 Added kwargs Chintan Shah 2019-10-06 18:57:13 -0400
  • 96d8dc4417 handling nans in loss tensor Chintan Shah 2019-10-06 18:55:35 -0400
  • 5dd0f1dd3a implemented masked mae loss, added tensorflow writer, changed % logic Chintan Shah 2019-10-06 18:08:13 -0400
  • a8814d5d93 Added docstrings Chintan Shah 2019-10-06 17:12:06 -0400
  • ad8ac8ff2f Merge branch 'pytorch_integration' into pytorch_scratch Chintan Shah 2019-10-06 17:07:54 -0400
  • 5b93f3c778 Merge branch 'pytorch_integration' of github.com:chnsh/DCRNN into pytorch_integration Chintan Shah 2019-10-06 17:02:35 -0400
  • 9fb999c3bb squash! Added dcrnn_cell Chintan Shah 2019-10-06 17:01:49 -0400
  • d1964672c2 Added dcrnn_cell Chintan Shah 2019-10-06 11:55:02 -0400
  • d46b605a65 ensured row major ordering Chintan Shah 2019-10-06 15:53:14 -0400
  • 6331173f44 logging and refactor Chintan Shah 2019-10-06 15:22:57 -0400
  • ec5d9555a5 logging and refactor Chintan Shah 2019-10-06 15:15:11 -0400
  • 55a087ac9f logging and refactor Chintan Shah 2019-10-06 14:40:53 -0400
  • 02c4681ad9 logging and refactor Chintan Shah 2019-10-06 14:38:44 -0400
  • b6a2b3fe8e logging and refactor Chintan Shah 2019-10-06 14:34:58 -0400
  • 036e552bf6 logging and refactor Chintan Shah 2019-10-06 14:31:46 -0400
  • 31acadedce logging and refactor Chintan Shah 2019-10-06 14:29:28 -0400
  • e563e1bf37 moving tensors to GPU [v3] Chintan Shah 2019-10-06 14:13:02 -0400
  • 017ec70783 moving tensors to GPU [v2] Chintan Shah 2019-10-06 14:10:20 -0400
  • ba304e9f04 moving tensors to GPU Chintan Shah 2019-10-06 14:00:54 -0400
  • 9454fd91a2 Ensured sparse mm for readability, logging sparsely as well Chintan Shah 2019-10-06 13:44:55 -0400
  • 2e1836df40 Rough implementation complete - could forward pass it through the network Chintan Shah 2019-10-06 13:24:37 -0400
  • b65df994e4 Added dcrnn_cell Chintan Shah 2019-10-06 11:55:02 -0400
  • e80c47390d Merge branch 'pytorch_implementation' into pytorch_scratch Chintan Shah 2019-10-06 11:49:49 -0400
  • 5a790d5586 cuda no grad Chintan Shah 2019-10-04 23:30:10 -0400
  • 593e3db1bf Using model.cuda() if cuda is available Chintan Shah 2019-10-04 22:45:08 -0400
  • 8d3b1d0d66 Implemented lr annealing schedule Chintan Shah 2019-10-04 21:18:05 -0400
  • ba880b8230 Implementing load and save models and early stopping Chintan Shah 2019-10-04 17:25:03 -0400
  • d9f41172dc Implemented eval and function Chintan Shah 2019-10-04 17:07:38 -0400
  • 20c6aa5862 Fixed bugs with refactoring Chintan Shah 2019-10-04 16:05:52 -0400
  • 2b8d5e6b31 Refactored code and moved everything into a DCRNN forward pass Chintan Shah 2019-10-04 13:02:50 -0400
  • f41dc442b0 Implemented gradient clipping and returning output from training one batch Chintan Shah 2019-10-03 19:35:54 -0400
  • 9834b12d5a Cleaned up code, fixed bugs in implementation, seems like it starts training with GRU Chintan Shah 2019-10-02 22:20:43 -0400
  • f96a8c0d59 Dirty commit - setup model but [GRUCell] not working, tried ParameterList, did not work Chintan Shah 2019-10-02 18:09:33 -0400
  • c876cbfba3 Setup training loop and logging Chintan Shah 2019-10-02 17:34:07 -0400
  • 86c4c5704d Implemented single batch forward pass Chintan Shah 2019-10-01 22:47:59 -0400
  • 69288460b1 Simplified encoder decoder model and moved curriculum learning outside Chintan Shah 2019-10-01 10:07:34 -0400
  • a1c9af2bad Setup curriculum learning framework Chintan Shah 2019-09-30 21:58:55 -0400
  • bdce241a8f Implemented abstract method and changed scheme to do all layers first for each timestep Chintan Shah 2019-09-30 20:32:31 -0400
  • 0769a3b2e2 Implemented seq2seq without DCGRU and curriculum learning Chintan Shah 2019-09-30 19:14:24 -0400
  • 7349f2ed67 Figured out decoder shapes Chintan Shah 2019-09-30 16:04:03 -0400
  • 6386ac7eb4 Implemented encoder using GRUCell instead so that it's easier to swap that with DCGRUCell Chintan Shah 2019-09-29 17:40:52 -0400
  • 66fb202d21 Implemented Encoder with GRU - should swap GRU with DCGRU Chintan Shah 2019-09-29 12:51:49 -0400
  • 7a5e3c0216 model partially implemented - does not work yet Chintan Shah 2019-09-29 11:13:08 -0400
  • adbfa19146 Implemented DCGRUCell in pytorch (untested) Chintan Shah 2019-09-08 19:28:20 -0400
  • 00c70b3a27 Implemented fc layer and changed docker image to use pytorch Chintan Shah 2019-09-08 18:47:19 -0400
  • 7ba7fa320d Using pytorch image Chintan Shah 2019-09-07 17:53:46 -0400
  • 69d6c0e053 Add Dockerfile Chintan Shah 2019-09-07 17:18:31 -0400
  • 8c20ca1a9c Adds baseline methods for evaluation. Yaguang 2019-06-18 13:00:24 -0700
  • 37f0734cb4
         Update README.md Yaguang 2019-03-18 21:32:57 -0700
  • 3caf887484 Add data example. Yaguang 2019-03-17 11:56:55 -0700
  • 4344dddc6d Add instructions on using HDF5 with python. Yaguang 2019-03-17 11:43:30 -0700
  • 81b4626193 Remove unused code. Yaguang 2019-01-18 19:00:34 -0800
  • 734ecbc138
         Update README.md Yaguang 2019-01-12 12:04:28 -0800
  • 763eb7af69 Add pretrained model on PEMS-BAY. Yaguang 2019-01-10 17:44:07 -0800
  • ad36deb794
         Update README.md Yaguang 2019-01-08 11:33:20 -0800
  • c8a676604b Add PEMS-BAY configuration. liyaguang 2019-01-08 11:21:22 -0800
  • 1481b24b1b
         Update README.md Yaguang 2018-12-24 00:47:17 -0800
  • ac5321fcf7
         Update README.md Yaguang 2018-11-03 11:54:22 -0700
  • a8aa1732a9
         Update README.md Yaguang 2018-10-12 13:30:53 -0700
  • d59d44e4f0 Code refactor. Yaguang 2018-10-01 17:45:46 -0700
  • bbc06b6c0c Add log level support and load_dataset method. Yaguang 2018-10-01 17:45:28 -0700
  • 88d9fc86d1 Update instructions for graph generation. Yaguang 2018-10-01 10:54:56 -0700
  • e6645df191 Update README. Yaguang 2018-10-01 09:59:47 -0700
  • 2e4b8c868f
         Merge pull request #9 from liyaguang/v2 Yaguang 2018-10-01 09:56:22 -0700
  • 9520e6cf85 Update pretrained model. Yaguang 2018-10-01 09:47:23 -0700
  • e0212cc178 Update README and requirements. Yaguang 2018-09-30 22:15:27 -0700
  • 7e05414d99 Combine Val and Test model. Yaguang 2018-09-30 22:14:45 -0700
  • 5dc36fed7c Refactor DCRNN Model. Yaguang 2018-09-30 21:53:40 -0700
  • 2f2d748b45 Add AMSGrad from https://github.com/taki0112/AMSGrad-Tensorflow for stablized Adam Training. Yaguang 2018-09-30 21:52:20 -0700
  • e8d6c08db0 Merge log_helper into utils and change logging mechanism. Yaguang 2018-09-26 11:33:21 -0700
  • 9ec161543b Code refactoring, including data loading, logging, configuration, removing redundant code. Yaguang 2018-09-26 11:19:00 -0700
  • 80e156c580
         Update README.md Yaguang 2018-07-22 18:55:35 +0800
  • 67bd763b18
         Update README.md Yaguang 2018-07-22 18:53:50 +0800
  • 17927239a7 Update the pretrained model and results. Yaguang 2018-06-07 11:18:55 +0800
  • e08002d72d Update the seq2seq training logic. Yaguang 2018-06-07 11:16:33 +0800
  • 562514cc30 Add a more data-efficient training data generation method. Yaguang 2018-06-07 10:41:47 +0800
  • e93435c598 Replace json with yaml for configuration. Yaguang 2018-04-18 11:51:35 -0700
  • 3e94a0ff0e Adds support for python 3.5 and python 3.6. Yaguang 2018-04-18 11:50:39 -0700
  • fcdf62d6de Update reference. Yaguang 2018-03-30 13:42:12 -0700
  • 90e39f5a4a Update README to include the lat/lon file. Yaguang 2018-03-29 17:58:52 -0700
  • 82922c8308 Adds the file containing sensor lat/lon. Yaguang 2018-03-29 17:48:51 -0700