Commit Graph

  • ea825a5aa6 Add Cuda support to mnist_hogwild (#508) Vitaly Fedyunin 2019-02-15 14:14:58 -05:00
  • fe8abc3c81 Remove deprecated & unused Variable() (#496) 박승원 2019-01-24 12:28:20 +09:00
  • 0d3fe14a1c Use argmax() instead of max()[1] in mnist/main.py (#494) 박승원 2019-01-16 23:50:47 +09:00
  • 29a38c6c44 Add cpp folder for C++ frontend examples (#492) Peter Goldsborough 2019-01-15 07:22:30 -08:00
  • 5d27fdb7a4 Update README.md (#493) pavel- 2019-01-15 17:17:04 +02:00
  • 635e64c1aa Add missing rank argument (#490) nsvanirudh 2019-01-14 02:22:58 -05:00
  • e0929a4253 remove more spurious perplexity references Soumith Chintala 2018-12-29 22:36:50 -05:00
  • 97e3e13d36 remove references to PTB perplexity numbers Soumith Chintala 2018-12-29 22:33:23 -05:00
  • c5985a81a8 Divide args.workers by ngpus_per_node (#485) daquexian 2018-12-27 13:27:16 +08:00
  • 44053c5f90 Add missing world size argument (#484) daquexian 2018-12-27 13:24:19 +08:00
  • d6b52110ba delete .swp file. (#474) Yüduo Wu 2018-12-15 16:59:37 -08:00
  • d8dd8caeec Mnist update (#469) surgan12 2018-12-13 11:36:32 +05:30
  • 35a9d849a5 fix a typo (#466) KAI ZHAO 2018-12-10 11:55:32 +08:00
  • 64f829ce49 Examples dcgan (#464) surgan12 2018-12-09 11:44:55 +05:30
  • 6d08877a1f python 2 fix Soumith Chintala 2018-12-06 22:52:53 -08:00
  • 15e27719d7 ImageNet, distributed bug fixes (#462) Teng Li 2018-12-04 21:46:59 -08:00
  • 91f230a13b Fix missing required argument for --evaluate. (#455) sulc 2018-11-27 17:42:45 +01:00
  • 12e434016f Fix name of dataset (#451) Ben Eyal 2018-11-27 07:26:31 +02:00
  • df70aa77bf fix broken args\n\nFor multiple long options python takes the first as attribute name by default. The latest commit swapped the order of options that causes the script an error. Added dest option to explicitly specify the valid attribute name (#454) Naofumi Tomita 2018-11-26 15:53:49 -05:00
  • 537f697187 Unified order of flag names in argparser: short flag names ('-f') go before long flag names ('--foo'). (#449) Timothy Gebhard 2018-11-21 17:52:20 +01:00
  • 0fadadceaa [Distributed] Multiprocessing Distributed Training for ImageNet Example (#441) Teng Li 2018-11-15 12:37:39 -08:00
  • 81f47e8ea4 Bug fix in the calculation of the validation loss (#427) trault14 2018-10-29 22:25:04 +01:00
  • 05ed879d19 modified the hogwild example to perform testing outside the different training processes (#426) Shagun Sodhani 2018-10-28 22:55:41 -04:00
  • 502e45deea replace Upsample layer with interpolate function (#424) Varun Agrawal 2018-10-16 23:57:04 -04:00
  • 323079f577 Use torchtext 0.3.1 (#423) Kento Nozawa 2018-10-15 00:42:34 +09:00
  • 0c1fcc8f9d Update all prec occurrences to acc (#420) Vipin Pillai 2018-10-11 14:32:08 -04:00
  • 95d5fddfb5 Fix bug in reparameterize() (#419) Durk Kingma 2018-10-04 21:58:04 -07:00
  • e128edac1d Add tokenization (#417) Joost Bastings 2018-09-25 04:37:51 +02:00
  • 8473762c88 precision -> accuracy Soumith Chintala 2018-09-24 20:46:24 -04:00
  • 753d086ba7 removed two deprecated function calls, added __name__ check to address multithreading bug in dataloader (#414) Jacob Austin 2018-09-20 19:24:13 -07:00
  • 1189d6a2bd Fix MNIST example for perf test (#413) perftests Will Feng 2018-09-17 13:55:13 -04:00
  • 6fd43cd04f Updated clipping to torch.nn.utils.clip_grad_norm_ (#403) Kartik Srivastava 2018-08-20 22:44:44 +08:00
  • 2fc0211d30 fix fast_neural_style --export_onnx (#397) Bowen Bao 2018-07-30 08:06:56 -07:00
  • 75e7c75469 Fix snli device to work with newest torchtext (#394) Tongzhou Wang 2018-07-26 17:05:43 -04:00
  • b5f361262f fix fast neural transfer on cuda (#392) Tongzhou Wang 2018-07-25 21:41:03 -04:00
  • 82a00b3c3a Normalized loss in actor-critic and REINFORCE examples. AC converges more than 2x faster now, REINFORCE about 10% faster. r.aristov 2018-07-14 00:36:23 +03:00
  • 98b1fce0d2 Add support for deteriminstic behavior in ImageNet benchmark. (#381) Jerry Ma 2018-07-07 07:26:05 -07:00
  • 29c2ed8ca6 Make the data preparation process cross-platform (#379) peterjc123 2018-07-04 21:26:33 +08:00
  • f83508117b open file explicitely using utf-8 (#366) siran 2018-06-04 11:02:40 -04:00
  • 19a7fa361c missing "," (#365) 0.4 Shuxin Zheng 2018-06-01 21:48:00 +08:00
  • f9820471d6 Convert to proper script with main(), fix #352 (#353) Shital Shah 2018-05-15 01:25:42 -07:00
  • 9dc6538693 fix 0.4 version compliance (#354) Sergei Chicherin 2018-05-14 17:14:02 +03:00
  • eee5ca3b68 Export Word Language Model to ONNX (#348) Lu Fang 2018-05-03 18:57:54 -07:00
  • 0604520baf Onnx export (#322) Sebastian Meßmer 2018-04-25 18:08:41 -07:00
  • 83f1b5c266 Fix call to Tensor.cuda() Sam Gross 2018-04-25 08:52:29 -07:00
  • c7df3c00f1 VAE: use F.sigmoid as it is parameter-free (#338) Xingdong Zuo 2018-04-25 13:31:46 +00:00
  • 56114a62cd Update main.py Soumith Chintala 2018-04-25 08:47:52 -04:00
  • d7e452ef69 Update main.py (#337) Alfredo Canziani 2018-04-24 18:29:41 -04:00
  • 645c7c386e codemod for 0.4 (#331) (#336) Soumith Chintala 2018-04-24 14:22:23 -04:00
  • 01c54ac947 codemod for 0.4 (#331) Tongzhou Wang 2018-04-24 10:45:24 -04:00
  • 287aa76db6 Fix clip_grad_norm use for word_language_model example (#332) Will Feng 2018-04-19 18:21:58 -04:00
  • de5ba9f2b6 Fix type check on hidden state (#324) Sam Gross 2018-04-03 11:48:04 -04:00
  • dcdabc22b3 Fix arg of LSUN: db_path -> root (#319) 0.3.1 0.3 Jason Park 2018-03-23 00:11:00 +09:00
  • c66593f169 fix accuracy bug Soumith Chintala 2018-03-05 09:35:36 -05:00
  • ce169fad9d Fix warnings in word_language_model due to Tensor/Variable merge (#309) Sam Gross 2018-02-23 12:38:40 -05:00
  • fa6a20d07f Fixes for Tensor and Variable merge (#307) Sam Gross 2018-02-22 15:06:08 -05:00
  • 8952fdcc2e Fixes for perf tests (#302) Will Feng 2018-02-09 12:05:51 -05:00
  • 4ef2d4d0c8 Fix actor_critic example (#301) Sam Gross 2018-02-08 16:38:10 -05:00
  • 963f7d1777 Fix VAE losses (sum over everything) (#296) Kai Arulkumaran 2018-02-02 09:21:47 -05:00
  • 8256aee444 Fix UserWarning in two examples (#293) Lawrence Neal 2018-01-29 00:40:22 -08:00
  • e11e0796fc Fix indentation to be self-consistent (#279) Justin 2018-01-06 19:06:50 +01:00
  • e23a9b4aea fix: Multinomial is now Categorical Brandon Lin 2017-12-07 20:25:28 -08:00
  • d5678bc8ac Replace WikiText-2 files with correct dataset (#264) Stephen Merity 2017-12-05 00:42:15 -08:00
  • ca90734250 Fix bugs and improve performance Sam Gross 2017-11-13 17:06:14 -08:00
  • 82cef4452f Fix action indexing Kaixhin 2017-11-07 09:16:14 +00:00
  • 9faf2c65f7 Update RL examples to use torch.distributions instead of reinforce Kaixhin 2017-11-04 12:57:51 +00:00
  • e0d33a69be Scale -> Resize + RandomSizedCrop -> RandomResizedCrop Alykhan Tejani 2017-10-01 15:53:48 +01:00
  • cf74c81ec0 Swap PTB for Wikitext-2 (which is open access) Adam Lerer 2017-11-24 14:47:24 -05:00
  • 62d5ca57af Add linear layer to time series prediction Thomas Viehmann 2017-11-09 09:01:06 +01:00
  • fad775979e Consistent newlines Nikolai Morin 2017-10-26 14:51:38 +02:00
  • 7532a61f95 Document the data arrangement in word_language_model Nikolai Morin 2017-10-26 14:44:09 +02:00
  • 23f8abfccf bug fix: vocab.load_vectors signature update Siva Reddy 2017-10-21 18:17:05 -07:00
  • 9a02f2aa4a Fix an argument instruction in mnist_hogwild Guangshuo Chen 2017-10-25 13:52:50 +02:00
  • 7d0d413425 Add link to script for preparing imagenet val data Achal Dave 2017-10-05 23:56:21 -04:00
  • aa7adf0848 Remove unused math import Kai Arulkumaran 2017-10-01 23:41:28 +01:00
  • 9fe431ed5b Fix VAE loss + improve reconstruction viz Kaixhin 2017-10-01 23:26:57 +01:00
  • 5f24730b50 Add support for CUDA Kaixhin 2017-10-01 13:31:19 +01:00
  • ab7cb38969 Balance VAE losses, add reconstruction + sampling Kaixhin 2017-10-01 13:24:59 +01:00
  • ddf9e3094a h rather than c should be fed into the next layer of LSTM (#222) David Tolpin 2017-09-26 01:01:37 +03:00
  • 3648cbc1a8 vae: Fix UserWarning (#220) Ryuichi Yamamoto 2017-09-22 22:36:21 +09:00
  • 30b9c0e479 Update README.md (#219) saikrishna_1996 2017-09-21 16:07:29 -04:00
  • a72359893f change lr to 0.8 to fix the issue for 0.2 boathit 2017-09-21 13:14:13 +08:00
  • 407bd3e69e minor spelling, evaluted->evaluated Brett Koonce 2017-09-15 10:37:38 -07:00
  • 930ae27d64 Change test DataLoader to use the test batch size Robin Dinse 2017-09-05 20:15:44 +02:00
  • 2dca104044 reinforcement_learning fix reward threshold pbartet 2017-08-14 11:47:51 +02:00
  • 86bc3e516d Add model_names back. Fixes #195 Bobby DeSimone 2017-08-07 17:16:17 -07:00
  • fb9ca4d975 mnist 0.2 fixes Soumith Chintala 2017-07-20 14:16:58 -07:00
  • 9053040f22 fix for 0.2 Soumith Chintala 2017-07-20 14:10:55 -07:00
  • 9012fae6e4 fix for 0.2 Soumith Chintala 2017-07-20 13:23:35 -07:00
  • 5c2b513a8e Change "gain" -> "calculate_gain" (#192) Kai Arulkumaran 2017-08-02 15:22:49 +01:00
  • b0a116e3fe Add an option to perform distributed ImageNet training (#185) Adam Paszke 2017-07-28 17:34:26 -04:00
  • 0722b2f4f7 Remove unused imports in SR example (#190) Kai Arulkumaran 2017-07-23 20:44:16 +01:00
  • d165984825 Use nn.init from core to init SR example (#189) Kai Arulkumaran 2017-07-23 19:28:14 +01:00
  • 6b17f79eee README: Correct case and add link to PyTorch (#188) Ken Fehling 2017-07-22 22:07:14 -04:00
  • a4e69725c0 Fix test data in time_sequence_prediction (#186) Zhengping Che 2017-07-21 11:43:00 -07:00
  • 55179e774d 0.2 fixes v0.2.0 Soumith Chintala 2017-07-20 14:22:19 -07:00
  • 270254c3ab mnist 0.2 fixes Soumith Chintala 2017-07-20 14:16:58 -07:00
  • 2e7e1cd754 fix for 0.2 Soumith Chintala 2017-07-20 14:10:55 -07:00
  • ae9c33feef fix for 0.2 Soumith Chintala 2017-07-20 13:23:35 -07:00
  • 10b22dc917 Fix test_epoch typo (#183) boscotsang 2017-07-20 23:02:55 +08:00