Commit Graph

  • fac18dc776 * Upd intro page Matthew Honnibal 2015-02-01 18:12:46 +1100
  • af07eb508f Merge branch 'master' of ssh://github.com/honnibal/spaCy Matthew Honnibal 2015-02-01 18:11:44 +1100
  • dea1245311 * Require advanced version of cymem Matthew Honnibal 2015-02-01 17:04:59 +1100
  • ac20a53509 * Change version to 0.4.0 Matthew Honnibal 2015-02-01 16:54:13 +1100
  • 3d44dd32a1 * Upd project status note Matthew Honnibal 2015-02-01 16:47:33 +1100
  • 7e82c485f8 Update README.md honnibal 2015-02-01 16:40:46 +1100
  • 70dda04023 * Upd quickstart docs Matthew Honnibal 2015-02-01 16:40:15 +1100
  • 42e25dce79 * Fix quickstart docs Matthew Honnibal 2015-02-01 16:35:12 +1100
  • 46aaf624ce Update README.md honnibal 2015-02-01 16:29:16 +1100
  • 7d9d8c1489 * Fix test_tag_names again Matthew Honnibal 2015-02-01 16:25:03 +1100
  • d1a5091052 * Require six Matthew Honnibal 2015-02-01 16:24:50 +1100
  • c260180207 * Add six to requirements Matthew Honnibal 2015-02-01 16:24:33 +1100
  • 2ec023a994 * Add WordNet license file Matthew Honnibal 2015-02-01 16:11:53 +1100
  • 7a3f22bb51 Merge branch 'master' of ssh://github.com/honnibal/spaCy Matthew Honnibal 2015-02-01 15:57:07 +1100
  • d1c9e7ff70 * Make test_tag_names version-portable Matthew Honnibal 2015-02-01 15:40:20 +1100
  • 9f76b94582 Update README.md honnibal 2015-02-01 14:47:18 +1100
  • a6ead1f35e Update README.md honnibal 2015-02-01 14:42:35 +1100
  • 10f2e9cfdc Update README.md honnibal 2015-02-01 14:35:19 +1100
  • a20fdbd8ee * Upd download script Matthew Honnibal 2015-02-01 13:22:23 +1100
  • 6259947d30 Merge branch 'master' of ssh://github.com/honnibal/spaCy Matthew Honnibal 2015-02-01 13:14:48 +1100
  • 1f31774078 * Python3 fix for test_tag_names Matthew Honnibal 2015-02-01 13:14:20 +1100
  • 76d9394cb4 * Fix vocab.pyx for Python3 Matthew Honnibal 2015-02-01 13:14:04 +1100
  • cc216c2012 Update README.md honnibal 2015-01-31 23:26:12 +1100
  • 3d545d62c5 * Note version, but say Python3 is broken Matthew Honnibal 2015-01-31 23:18:37 +1100
  • 2525e3ca76 * Add known issue to updates page Matthew Honnibal 2015-01-31 23:08:16 +1100
  • fc32a213ed * Advertise new version Matthew Honnibal 2015-01-31 23:05:17 +1100
  • 754f4aed8e * Inc version Matthew Honnibal 2015-01-31 22:49:46 +1100
  • 63abdf154c * Hastily hack download file Matthew Honnibal 2015-01-31 22:48:32 +1100
  • 1af5ee1a69 * Add updates page Matthew Honnibal 2015-01-31 22:43:52 +1100
  • 7de00c5a79 * Try not holding a reference to Pool, since that seems to confuse the GC Matthew Honnibal 2015-01-31 22:10:22 +1100
  • ce3ae8b5d9 * Fix platform-specific lexicon bug. Matthew Honnibal 2015-01-31 16:38:58 +1100
  • a1ed574b7b * Fix default model path for English Matthew Honnibal 2015-01-31 16:38:27 +1100
  • 543fe5a280 * Comment training function back in Matthew Honnibal 2015-01-31 16:37:57 +1100
  • 018e0bfa24 * Bug fixes to parse navigation Matthew Honnibal 2015-01-31 16:37:13 +1100
  • e013555b25 * Add option to download script Matthew Honnibal 2015-01-31 13:51:56 +1100
  • a3955fd8d5 * Require plac Matthew Honnibal 2015-01-31 13:50:40 +1100
  • 951684c4ac * Upd travis.yml Matthew Honnibal 2015-01-31 13:50:30 +1100
  • 0fd4a71bed * Add provisional tests for sentence boundary detection Matthew Honnibal 2015-01-31 13:46:11 +1100
  • df749f257d * Upd train script Matthew Honnibal 2015-01-31 13:44:37 +1100
  • 08ca5c8970 * Add sent_end flag to TokenC struct Matthew Honnibal 2015-01-31 13:44:16 +1100
  • 024cfd485c * Pass tag_strings as a tuple, to support new Tokens API Matthew Honnibal 2015-01-31 13:43:37 +1100
  • 77d62d0179 * Large refactor of Token objects, making them much thinner. This is to support fast parse-tree navigation. Matthew Honnibal 2015-01-31 13:42:58 +1100
  • 88170e6295 * Supply dep_strings as a tuple, for the changed API on Tokens Matthew Honnibal 2015-01-31 13:42:09 +1100
  • 0981d68022 * Set a sent_end flag during parsing, for later use Matthew Honnibal 2015-01-31 13:41:46 +1100
  • 251dbf24d7 * Fix unintialised variable error Matthew Honnibal 2015-01-30 20:46:34 +1100
  • 83a4df5a1a * Fix download script Matthew Honnibal 2015-01-30 20:40:42 +1100
  • 6f9ebc2f34 * Fix download script Matthew Honnibal 2015-01-30 20:33:19 +1100
  • a2bed49ac7 * Upd travis.yml Matthew Honnibal 2015-01-30 20:27:35 +1100
  • 8b85d0bb8a * Only download small data if no data dir exists Matthew Honnibal 2015-01-30 20:27:14 +1100
  • e03b1fea22 * Don't download full data during test Matthew Honnibal 2015-01-30 20:12:33 +1100
  • 2da694f65e * Don't load parser in test_pre_punct Matthew Honnibal 2015-01-30 20:11:47 +1100
  • e88ceda0ab * Set PYTHONPATH in travis.yml Matthew Honnibal 2015-01-30 19:48:34 +1100
  • 6c081dd1fc * Handle failure when numpy headers are already installed correctly Matthew Honnibal 2015-01-30 19:48:19 +1100
  • 1a7a1c2771 * Fix Issue #16: tokens recurse when printing Matthew Honnibal 2015-01-30 19:47:50 +1100
  • cb95ef6934 * Fix download script Matthew Honnibal 2015-01-30 19:28:43 +1100
  • e578bd37bd * Fix download script Matthew Honnibal 2015-01-30 18:59:31 +1100
  • df52014d12 * Fix download script Matthew Honnibal 2015-01-30 18:36:24 +1100
  • f0bbffca8d * Fix the way numpy headers are installed during compilation from source Matthew Honnibal 2015-01-30 18:14:45 +1100
  • 1ef0e04aa0 * Change travis.yml to build from source, instead of from PyPi. PyPi checking will be done in a different branch. Matthew Honnibal 2015-01-30 18:06:02 +1100
  • 0f95712189 * Improve accuracy reporting during training Matthew Honnibal 2015-01-30 18:05:06 +1100
  • b68f563c2f * Fix Issue #14: Improve parsing API Matthew Honnibal 2015-01-30 18:04:41 +1100
  • 998b607f65 * Upd download script, having it download all data if there's no data/ directory, allowing easier compilation from source Matthew Honnibal 2015-01-30 18:04:01 +1100
  • 0b53fd7daa * Add test for parse tree navigation Matthew Honnibal 2015-01-30 18:02:58 +1100
  • ef2493a3bd * Upd gitignore Matthew Honnibal 2015-01-30 16:49:44 +1100
  • d5d1578e44 * Add manifest file Matthew Honnibal 2015-01-30 16:49:02 +1100
  • 0a1ec40f76 * Add draft work on features Matthew Honnibal 2015-01-30 16:46:52 +1100
  • 7d432b7e39 * Add tests for vector-space model Matthew Honnibal 2015-01-30 16:45:45 +1100
  • 48b98e3fb5 * Add test for tag names Matthew Honnibal 2015-01-30 16:45:11 +1100
  • 613a195f92 * Add test for indices Matthew Honnibal 2015-01-30 16:44:29 +1100
  • 03cc2ee08e * Add test for numpy array transport Matthew Honnibal 2015-01-30 16:43:55 +1100
  • d20eeac156 * Start work on lexrank tutorial Matthew Honnibal 2015-01-30 16:42:43 +1100
  • b3f9b199cf Merge branch 'punctparse' Matthew Honnibal 2015-01-30 16:38:56 +1100
  • ca7577d8a9 * Allow parsers and taggers to be trained on text without gold pre-processing. Matthew Honnibal 2015-01-30 16:36:24 +1100
  • 67d6e53a69 * Ensure parser and tagger function correctly when training from missing values, indicated by -1 Matthew Honnibal 2015-01-30 14:08:56 +1100
  • 4ff180db74 * Fix off-by-one error in commit 0a7fceb Matthew Honnibal 2015-01-30 12:49:33 +1100
  • d0e08a5b57 * Upd index tests Matthew Honnibal 2015-01-30 12:35:13 +1100
  • 0a7fcebdf7 * Fix Issue #12: Incorrect token.idx calculations for some punctuation, in the presence of token cache Matthew Honnibal 2015-01-30 12:33:38 +1100
  • b38093237e * More debug prints Matthew Honnibal 2015-01-30 11:15:54 +1100
  • 35a18250cc * Upd tests, avoiding unnecessary processing to make testing faster Matthew Honnibal 2015-01-30 10:41:55 +1100
  • 5458f220f8 * Fix quickstart instructions Matthew Honnibal 2015-01-30 10:31:25 +1100
  • 11ed65b93c * Work on alignment, for evaluation with non-gold preprocessing Matthew Honnibal 2015-01-30 10:31:03 +1100
  • ebf7d2fab1 * Use non-joint sbd, for more simplicity and fewer classes Matthew Honnibal 2015-01-29 06:22:03 +1100
  • d05c5bf141 * Remove comment Matthew Honnibal 2015-01-29 05:19:27 +1100
  • b4348ce1c3 * Messily use unsegmented sentences to train the parser Matthew Honnibal 2015-01-29 04:21:13 +1100
  • 320b045daa * Oracle now consistent over gold standard derivation Matthew Honnibal 2015-01-29 03:41:58 +1100
  • f590382134 * Work on sbd Matthew Honnibal 2015-01-29 03:18:29 +1100
  • 8178963183 Merge c670b4a659 into fe5f34c37c Osman Başkaya 2015-01-28 11:34:18 +0000
  • 9e78d673d5 * Fix quickstart installation docs Matthew Honnibal 2015-01-28 14:28:34 +1100
  • fe5f34c37c Merge branch 'master' of ssh://github.com/honnibal/spaCy Matthew Honnibal 2015-01-28 14:01:00 +1100
  • 781dd712dc * Fix numpy commit problem Matthew Honnibal 2015-01-28 14:00:20 +1100
  • b08c0ce54e * Fix numpy install problem Matthew Honnibal 2015-01-28 13:58:33 +1100
  • 9171284d62 * Fix compile-from-source instructions Matthew Honnibal 2015-01-28 12:27:44 +1100
  • 1884a7a0be * Attach comment with paper Matthew Honnibal 2015-01-28 03:18:43 +1100
  • a2d6b195db * Add messy Break transitions, carefully following the scheme of Dd Zhang et al (2013) Matthew Honnibal 2015-01-28 03:09:45 +1100
  • f9ee5d9934 * Build a python list of word strings, for debugging Matthew Honnibal 2015-01-28 01:06:13 +1100
  • d819101571 * Improve error message on oracle failure Matthew Honnibal 2015-01-28 00:58:03 +1100
  • 1d39562946 * Ws Matthew Honnibal 2015-01-28 00:57:16 +1100
  • 97d91ada3a Update README.md honnibal 2015-01-27 21:55:36 +1100
  • 51ca3a285d Update README.md honnibal 2015-01-27 21:55:23 +1100
  • e6c3d3471f * Tweak documentation for Tokens, and hide constructor as __cinit__ Matthew Honnibal 2015-01-27 18:57:52 +1100