339 Commits
Author SHA1 Message Date
Piotr Czapla a7ac4f5170 missing file 2019-06-10 20:14:57 +02:00
Piotr Czapla 7d59c4ecf2 Sentence piece has new option -fix that changes coverage to 99.95% 2019-06-10 20:14:11 +02:00
Piotr Czapla ce0a29f385 Make it possible to set ftseed in poleval19_init 2019-06-06 17:45:13 +02:00
Piotr Czapla 854391a130 Merge branch 'reproduce-poleval' of https://github.com/n-waves/ulmfit-multilingual into reproduce-poleval 2019-05-19 00:17:27 +02:00
Piotr Czapla 38e131f802 Make the ulmfit compatible with old fastai (ulmfit_multilingual) 2019-05-19 00:17:03 +02:00
Marcin 7326df2c98 Train sentencepiece on unsup 2019-05-19 00:12:34 +02:00
Piotr Czapla 6f4db9de1a Fix ensemble output (by removing index column) 2019-05-18 22:50:55 +02:00
Piotr Czapla 06c32ae2dc Add ability to create folders when saving ensemble output 2019-05-18 21:59:12 +02:00
Piotr Czapla f38f8c0670 fix issue in the databunch name generation 2019-05-18 21:48:32 +02:00
Piotr Czapla 8927e45fb2 Add ensemble commnad 2019-05-18 21:48:08 +02:00
Marcin 15c2d0663a Save predictions for a proper dataset 2019-05-18 14:42:35 +02:00
Piotr Czapla 7a33ea5d1f Add no-test + ability to start training from the classifcation data set without wiki 2019-05-18 13:10:29 +02:00
Piotr Czapla 75b934bb83 Merge branch 'reproduce-poleval' of https://github.com/n-waves/ulmfit-multilingual into reproduce-poleval 2019-05-15 18:00:45 +02:00
Piotr Czapla 7c9129670e Poleval 19 ablation experiments 2019-05-15 18:00:40 +02:00
Marcin 30f5a7901c Save predictions 2019-05-14 15:34:40 +02:00
Piotr Czapla 7955db37ad Merge branch 'reproduce-poleval' of https://github.com/n-waves/ulmfit-multilingual into reproduce-poleval 2019-05-14 13:27:03 +02:00
Piotr Czapla ad3c11edb8 Add ability to turnoff weighted CrossEntropy 2019-05-14 13:26:57 +02:00
Piotr Czapla 6dc4a5f102 Add Kappa and Mathew score calcualtion + ls command to main 2019-05-14 11:53:16 +02:00
Marcin 746c6ca185 Don't reset lmseed 2019-05-14 06:57:02 +02:00
Marcin fe31778d70 Add option to remove duplicates 2019-05-14 02:30:25 +02:00
Marcin 84405b6e6c Allow bwd in model name 2019-05-13 20:11:11 +02:00
Piotr Czapla b065e36f1b Fix double name exception 2019-05-13 20:02:48 +02:00
Piotr Czapla 9101a0313c Let poleval19_full accept name 2019-05-13 19:53:45 +02:00
Piotr Czapla 29b600b8b5 Add ability to check different model save 2019-05-13 17:47:57 +02:00
Piotr Czapla b55715e4f6 Fix ftseed used in pretrain_lm so ulmfit lm works again 2019-05-13 17:47:28 +02:00
Piotr Czapla 6ef75e4dd7 Fix issue in poleval_eval + add valid metrics 2019-05-13 09:17:59 +02:00
Piotr Czapla 1c3044f5df fix issue with cls_best add lmseed to poleval 2019-05-13 03:44:35 +02:00
Piotr Czapla 53eadfc232 Add poleval19_full and fix name inference from paths 2019-05-13 03:25:12 +02:00
Piotr Czapla d11deac01a Add poleval commands + saving evaluation metrics to cache 2019-05-13 02:51:01 +02:00
Piotr Czapla c42bd8d9ed Update to eval to support seeds values 2019-05-13 00:07:46 +02:00
Piotr Czapla 724233585e Make it compatible with fastai 51 2019-05-12 22:19:02 +02:00
Marcin 50b73510de Add workaround for metrics init 2019-05-12 21:43:54 +02:00
Marcin 6aff0a9e22 Add more seeds 2019-05-12 21:29:32 +02:00
Marcin 3125ce7d8b Fix seeds 2019-05-11 16:59:59 +02:00
Piotr Czapla 8f2b7edee4 Merge pull request #45 from cahya-wirawan/tokens_min
Added an option to set the minimal limit of tokens per article
2019-05-07 00:09:31 +02:00
Marcin 27a720b105 Make decoder bias optional
These changes were made during poleval'19 competition to try to
reproduce results from fastai v0.7 in v1.0.
2019-05-01 20:57:20 +02:00
Cahya Wirawan 616bec18dc Added an option to set the minimal limit of tokens per article 2019-04-30 11:48:14 +02:00
Piotr Czapla f5306b4064 Add a few examples of commands avaliable in ulmfit-multilingual 2019-04-20 14:34:26 +02:00
Piotr Czapla f9032ad846 Update to README.md
to show the newest changes to the command line.
2019-04-20 14:28:45 +02:00
Piotr Czapla a312572004 Merge branch 'master' of https://github.com/n-waves/ulmfit-multilingual 2019-04-19 10:34:18 +02:00
Piotr Czapla 9f494b3ec6 Remove old bilm code that wasn't working 2019-04-19 10:32:35 +02:00
Julian Eisenschlos ba186113dc Merge pull request #43 from n-waves/fix-tests
Fix tests
2019-04-15 15:14:40 -03:00
NAUSICAA\Julian 2b7ec26f8b Lower LM accuracy test threshold 2019-04-15 15:00:46 -03:00
NAUSICAA\Julian e4d0529a3d Removing more bi-directional test 2019-04-15 14:41:08 -03:00
Piotr Czapla 94be5e6e0f Remove unused tests
Skip the test that were developed to support BiLM training that due to the changes in fastai stopped compiling
2019-04-14 19:19:12 +02:00
Piotr Czapla 8facc21cfa Add F1 metrics to evaluate 2019-04-14 19:13:43 +02:00
NAUSICAA\Julian 65397d7ae4 Bi-dir not supported yet 2019-04-13 16:38:58 -03:00
NAUSICAA\Julian 7e0a8517b6 Remove config params from language_model_learner funcion 2019-04-13 16:08:01 -03:00
NAUSICAA\Julian edf3b8d76e Adding missing param arch 2019-04-13 12:32:19 -03:00
Piotr Czapla a948d7df91 Merge branch 'master' of https://github.com/n-waves/ulmfit-multilingual 2019-03-26 21:29:33 +01:00