Thesis/pytorch-dnc
1
0
Fork 0
Code Issues Pull Requests Packages Projects Releases Wiki Activity

Commit Graph

  • 2026a8939d progressively make things differentiable ixaxaar 2017-12-04 01:48:03 +0530
  • 75264f2d80 Use inner product as similarity metric ixaxaar 2017-12-03 20:23:33 +0530
  • 54759c820c implement usage ixaxaar 2017-12-03 17:09:59 +0530
  • 7fa5d3614f reset indexes instead of creating new ones ixaxaar 2017-12-02 12:02:55 +0530
  • 5105ba7e90 test script on various distros docker ixaxaar 2017-12-02 01:30:57 +0530
  • af79ac10ef distribution independent script to some extent ixaxaar 2017-12-02 01:05:41 +0530
  • 9896bac743 add script to install faiss ixaxaar 2017-12-02 00:01:11 +0530
  • 3ec9216f8d Increase reserved memory fraction for indexes ixaxaar 2017-12-01 23:49:12 +0530
  • 85c61682c8 travis log file limit ixaxaar 2017-12-01 20:53:06 +0530
  • 01e63eb084 travis log file limit ixaxaar 2017-12-01 20:43:44 +0530
  • dbeb0ba378 Remove SDNC tests ixaxaar 2017-12-01 19:48:39 +0530
  • 7743d04e6c 😭 ixaxaar 2017-12-01 14:01:52 +0530
  • 54ac865401 work goddammit ixaxaar 2017-12-01 13:52:00 +0530
  • 6456f7d0aa and again ixaxaar 2017-12-01 13:38:25 +0530
  • b5f27a3daf try again ixaxaar 2017-12-01 13:23:57 +0530
  • dc05308b38 try by copying libgomp ixaxaar 2017-12-01 13:11:16 +0530
  • 0df4eb8d4c install swig ixaxaar 2017-12-01 11:41:53 +0530
  • 0e2cccbcf7 try installing cuda again ixaxaar 2017-12-01 11:30:51 +0530
  • 672ffe130e no gpu on travis ixaxaar 2017-12-01 11:24:09 +0530
  • 6e5ffa1770 install cuda travis ixaxaar 2017-12-01 01:00:53 +0530
  • 11cc2107d7
         Merge pull request #17 from ixaxaar/scale_interface Russi Chatterjee 2017-12-01 00:47:24 +0530
  • e895425cd4 Install cuda ixaxaar 2017-12-01 00:37:51 +0530
  • f137ea3689 Default encoding to utf8 ixaxaar 2017-11-30 23:24:24 +0530
  • 6f82bc454f Link to libopenblas.so.3 ixaxaar 2017-11-30 22:51:47 +0530
  • 1b55e2d0ec travis add openblas ixaxaar 2017-11-30 22:20:14 +0530
  • 2afeffcf40 trigger travis ixaxaar 2017-11-30 22:12:54 +0530
  • 049c293a83 finally no nans :relief: ixaxaar 2017-11-30 20:07:52 +0530
  • d7f916627a Get SDNC to work with FAISS, much faster ixaxaar 2017-11-30 14:54:51 +0530
  • 856efaab23 Add FAISS based indexing ixaxaar 2017-11-30 00:44:26 +0530
  • 379d719c76 Use FAISS instead of FLANN ixaxaar 2017-11-29 15:41:50 +0530
  • 7b1b53d4c5 Add SDNC to copy task- slow as hell 😭 ixaxaar 2017-11-27 23:59:13 +0530
  • 57b082d5f0 Write ixaxaar 2017-11-27 23:44:21 +0530
  • 33249133d9 Add sdnc nn ixaxaar 2017-11-27 16:46:53 +0530
  • 14f0f67b2d supports only one read key ixaxaar 2017-11-27 16:45:49 +0530
  • 00b561e4da Sparse reads barebones ixaxaar 2017-11-27 15:58:14 +0530
  • 41a96c29d3 Test for SDNC ixaxaar 2017-11-27 13:51:40 +0530
  • 8aee625101 FLANN read ixaxaar 2017-11-27 13:51:17 +0530
  • c4ec88a58e save work ixaxaar 2017-11-27 12:02:41 +0530
  • 30487965a6 Commit to keep track of changes ixaxaar 2017-11-24 16:41:19 +0530
  • f0bd0151f3 Tests for no memory pass ixaxaar 2017-11-21 17:28:49 +0530
  • 1a465dee0c Tests for no memory pass ixaxaar 2017-11-21 17:28:49 +0530
  • e55736751f Add num_hidden_layers to nn layers ixaxaar 2017-11-21 17:28:27 +0530
  • e3b4513730 Add num_hidden_layers to nn layers ixaxaar 2017-11-21 17:28:27 +0530
  • e74dc042d9
         Merge pull request #14 from MaxwellRebo/patch-1 Russi Chatterjee 2017-11-17 18:52:45 +0530
  • 32c690992a change unsupported character MaxwellRebo 2017-11-16 15:00:23 -0500
  • e9dace7a9d
         Update README.md Maxwell Rebo 2017-11-16 14:53:18 -0500
  • 909abd2bbc Add more learning rules to the copy task ixaxaar 2017-11-16 01:46:56 +0530
  • cc0308ba68 Add more learning rules to the copy task ixaxaar 2017-11-16 01:46:56 +0530
  • 213d5d5623 Update readme ixaxaar 2017-11-14 21:16:49 +0530
  • ea401eb798 Update readme ixaxaar 2017-11-14 21:16:49 +0530
  • 4055eb0d3e Use independent linear units like the official implementation ixaxaar 2017-11-14 21:15:11 +0530
  • 0d0f4c9ff6 Use independent linear units like the official implementation ixaxaar 2017-11-14 21:15:11 +0530
  • 029a963a62 Scale weights according to paper ixaxaar 2017-11-14 21:14:26 +0530
  • 72e56463ce Scale weights according to paper ixaxaar 2017-11-14 21:14:26 +0530
  • 06d18a0fc3 Bump version 0.0.6 ixaxaar 2017-11-13 00:49:58 +0530
  • 60fcbc569f Bump version ixaxaar 2017-11-13 00:49:58 +0530
  • 680488a3b4
         Merge pull request #11 from ixaxaar/debug Russi Chatterjee 2017-11-13 00:46:26 +0530
  • 7f4a8e892d Merge pull request #11 from ixaxaar/debug Russi Chatterjee 2017-11-13 00:46:26 +0530
  • e61841769a Update readme ixaxaar 2017-11-13 00:34:59 +0530
  • a28a0760d5 Update readme ixaxaar 2017-11-13 00:34:59 +0530
  • 2cf9ca59c8 Update dashboard image ixaxaar 2017-11-12 16:10:48 +0530
  • fb274aa3fb Update dashboard image ixaxaar 2017-11-12 16:10:48 +0530
  • f9cc0529b1 Update copy task defaults and readme ixaxaar 2017-11-12 15:56:33 +0530
  • 1e487f5dde Update copy task defaults and readme ixaxaar 2017-11-12 15:56:33 +0530
  • 256b323cdc Add optim option to copy_task ixaxaar 2017-11-12 14:05:53 +0530
  • 1e7c15d842 Add optim option to copy_task ixaxaar 2017-11-12 14:05:53 +0530
  • 70b3388051 Cleanup unused code ixaxaar 2017-11-12 13:58:52 +0530
  • 768481ea01 Cleanup unused code ixaxaar 2017-11-12 13:58:52 +0530
  • 6e58890709 Rewrite allocation vector part ixaxaar 2017-11-12 13:36:17 +0530
  • 89c379db75 Rewrite allocation vector part ixaxaar 2017-11-12 13:36:17 +0530
  • 8dc0a7cd41 Fine tune and delete redundant code ixaxaar 2017-11-11 22:21:41 +0530
  • 37dce1f118 Fine tune and delete redundant code ixaxaar 2017-11-11 22:21:41 +0530
  • d1443e1ef9 Put dropout in RNNs, clip controller outputs, cleanup ixaxaar 2017-11-11 22:21:04 +0530
  • f219984b78 Put dropout in RNNs, clip controller outputs, cleanup ixaxaar 2017-11-11 22:21:04 +0530
  • 5072093e7d Remove useless artifacts ixaxaar 2017-11-11 15:04:32 +0530
  • d865f65d4a Remove useless artifacts ixaxaar 2017-11-11 15:04:32 +0530
  • 5e6e16c39e Concat read vectors, dont add them according to paper ixaxaar 2017-11-11 00:35:21 +0530
  • 73b550ea53 Concat read vectors, dont add them according to paper ixaxaar 2017-11-11 00:35:21 +0530
  • 80f2067099 udpate readme ixaxaar 2017-11-10 21:32:14 +0530
  • a88b18a0b1 udpate readme ixaxaar 2017-11-10 21:32:14 +0530
  • aa9592d811 RNNs with CUDNN implementations, make whether to forward pass thorugh memory a controllable, update readme ixaxaar 2017-11-10 21:29:48 +0530
  • 51caa2e2ce RNNs with CUDNN implementations, make whether to forward pass thorugh memory a controllable, update readme ixaxaar 2017-11-10 21:29:48 +0530
  • ec3e35b83a Merge read vectors correctly, option to not forward pass through memory, correctly apply linear layers ixaxaar 2017-11-10 14:30:37 +0530
  • 67d9722231 Merge read vectors correctly, option to not forward pass through memory, correctly apply linear layers ixaxaar 2017-11-10 14:30:37 +0530
  • c4ce2426eb Stack DNCs properly ixaxaar 2017-11-09 00:33:24 +0530
  • f7dc1b5aab Stack DNCs properly ixaxaar 2017-11-09 00:33:24 +0530
  • eb96e64de0 Change dashboard image ixaxaar 2017-11-02 20:41:41 +0530
  • 5158e85f5d Change dashboard image ixaxaar 2017-11-02 20:41:41 +0530
  • 0bab77b7ca Debug complete memory ixaxaar 2017-11-02 20:27:55 +0530
  • 4079060908 Debug complete memory ixaxaar 2017-11-02 20:27:55 +0530
  • ab68a85774 Add new debug images ixaxaar 2017-11-01 15:19:57 +0530
  • d307acc6e4 Add new debug images ixaxaar 2017-11-01 15:19:57 +0530
  • f2149ca3bc Add new debug images ixaxaar 2017-11-01 15:19:39 +0530
  • fa95c713b8 Add new debug images ixaxaar 2017-11-01 15:19:39 +0530
  • e86f213f66 not reliable advise ixaxaar 2017-11-01 15:08:17 +0530
  • 9e1b0a2983 not reliable advise ixaxaar 2017-11-01 15:08:17 +0530
  • 4dde9c49f8 debug time ixaxaar 2017-11-01 15:04:30 +0530
  • af1a77ca7f debug time ixaxaar 2017-11-01 15:04:30 +0530
  • c8bfa6cbe6 Format attribution links ixaxaar 2017-11-01 13:08:32 +0530
  • 4700ecbac3 Format attribution links ixaxaar 2017-11-01 13:08:32 +0530