109 Commits

Author	SHA1	Message	Date
Matthew Honnibal	bd8e84998a	Add nO attribute to TextCategorizer model	2017-10-04 16:07:30 +02:00
Matthew Honnibal	f8a0614527	Improve textcat model slightly	2017-10-04 15:15:53 +02:00
Matthew Honnibal	39798b0172	Uncomment layernorm adjustment hack	2017-10-04 15:12:09 +02:00
Matthew Honnibal	774f5732bd	Fix dimensionality of textcat when no vectors available	2017-10-04 14:55:15 +02:00
Matthew Honnibal	af75b74208	Unset LayerNorm backwards compat hack	2017-10-03 20:47:10 -05:00
Matthew Honnibal	5cbefcba17	Set backwards compatibility flag	2017-10-03 20:29:58 +02:00
Matthew Honnibal	5454b20cd7	Update thinc imports for 6.9	2017-10-03 20:07:17 +02:00
Matthew Honnibal	e514d6aa0a	Import thinc modules more explicitly, to avoid cycles	2017-10-03 18:49:25 +02:00
Matthew Honnibal	f6330d69e6	Default embed size to 7000	2017-09-28 08:07:41 -05:00
Matthew Honnibal	1a37a2c0a0	Update training defaults	2017-09-27 11:48:07 -05:00
Matthew Honnibal	e34e70673f	Allow tagger models to be built with pre-defined tok2vec layer	2017-09-26 05:51:52 -05:00
Matthew Honnibal	63bd87508d	Don't use iterated convolutions	2017-09-23 04:39:17 -05:00
Matthew Honnibal	4348c479fc	Merge pre-trained vectors and noshare patches	2017-09-22 20:07:28 -05:00
Matthew Honnibal	4bd6a12b1f	Fix Tok2Vec	2017-09-23 02:58:54 +02:00
Matthew Honnibal	980fb6e854	Refactor Tok2Vec	2017-09-22 09:38:36 -05:00
Matthew Honnibal	d9124f1aa3	Add link_vectors_to_models function	2017-09-22 09:38:22 -05:00
Matthew Honnibal	a186596307	Add 'reapply' combinator, for iterated CNN	2017-09-22 09:37:03 -05:00
Matthew Honnibal	40a4873b70	Fix serialization of model options	2017-09-21 13:07:26 -05:00
Matthew Honnibal	20193371f5	Don't share CNN, to reduce complexities	2017-09-21 14:59:48 +02:00
Matthew Honnibal	f5144f04be	Add argument for CNN maxout pieces	2017-09-20 19:14:41 -05:00
Matthew Honnibal	78301b2d29	Avoid comparison to None in Tok2Vec	2017-09-20 00:19:34 +02:00
Matthew Honnibal	3fa76c17d1	Refactor Tok2Vec	2017-09-18 15:00:05 -05:00
Matthew Honnibal	7b3f391f80	Try dropping the Affine layer, conditionally	2017-09-18 11:35:59 -05:00
Matthew Honnibal	2148ae605b	Dont use iterated convolutions	2017-09-17 17:36:04 -05:00
Matthew Honnibal	8f42f8d305	Remove unused 'preprocess' argument in Tok2Vec'	2017-09-17 12:30:16 -05:00
Matthew Honnibal	8f913a74ca	Fix defaults and args to build_tagger_model	2017-09-17 05:46:36 -05:00
Matthew Honnibal	2a93404da6	Support optional pre-trained vectors in tensorizer model	2017-09-16 12:45:37 -05:00
Matthew Honnibal	24ff6b0ad9	Fix parsing and tok2vec models	2017-09-06 05:50:58 -05:00
Matthew Honnibal	16e25ce3b5	Merge branch 'develop' of https://github.com/explosion/spaCy into develop	2017-09-04 09:26:53 -05:00
Matthew Honnibal	9f512e657a	Fix drop_layer calculation	2017-09-04 09:26:38 -05:00
Matthew Honnibal	c0eaba8b28	Fix low-data textcat	2017-09-02 15:17:32 +02:00
Matthew Honnibal	a3b69bcb3d	Add low_data mode in textcat	2017-09-02 14:56:30 +02:00
Matthew Honnibal	a824cf8f9a	Adjust text classification model	2017-09-02 11:41:00 +02:00
Matthew Honnibal	ac040b99bb	Add support for pre-trained vectors in text classifier	2017-09-01 16:39:55 +02:00
Matthew Honnibal	6d4e8e14ca	Merge branch 'develop' of https://github.com/explosion/spaCy into develop	2017-08-25 12:37:16 -05:00
Matthew Honnibal	4ce5531389	Use layer norm instead of batch norm	2017-08-25 12:37:10 -05:00
Matthew Honnibal	1c5c256e58	Fix fine_tune when optimizer is None	2017-08-23 10:51:33 +02:00
Matthew Honnibal	9c580ad28a	Merge branch 'develop' of https://github.com/explosion/spaCy into develop	2017-08-22 17:02:04 -05:00
Matthew Honnibal	a4633fff6f	Restore use of batch norm in model	2017-08-22 17:01:58 -05:00
Matthew Honnibal	df2745eb08	Merge branch 'develop' of https://github.com/explosion/spaCy into develop	2017-08-22 19:00:43 +02:00
Matthew Honnibal	18b64e79ec	Fix fine tuning	2017-08-21 19:18:26 -05:00
Matthew Honnibal	a21d8f3f0b	Add predict paths to _ml models	2017-08-21 23:23:45 +02:00
Matthew Honnibal	80acbc5f1f	Fix fine-tune weight mixture	2017-08-21 14:15:29 -05:00
Matthew Honnibal	c10f63bf10	Initialize fine tuning to 0.5	2017-08-20 15:59:48 -05:00
Matthew Honnibal	8a59718fd6	Fix fine-tuning	2017-08-20 18:17:35 +02:00
Matthew Honnibal	bae59bf92f	Remove BiLSTM import	2017-08-18 22:46:59 +02:00
Matthew Honnibal	fe90dfc390	Restore changes from nn-beam-parser to spacy/_ml	2017-08-18 22:38:28 +02:00
Matthew Honnibal	ce321b0322	Restore changes from nn-beam-parser to spacy/_ml	2017-08-18 22:24:46 +02:00
Matthew Honnibal	931509d96a	Merge branch 'develop' of https://github.com/explosion/spaCy into develop	2017-08-18 21:57:15 +02:00
Matthew Honnibal	263366729e	Don't import BiLSTM	2017-08-18 21:56:31 +02:00