Richard Liaw
26c975ec66
sync
2020-06-17 21:51:22 -07:00
Richard Liaw
fdc9242bc1
small-changes
2020-06-17 19:42:53 -07:00
Richard Liaw
a5a3ed722c
move-ps
2020-06-16 20:33:21 -07:00
Richard Liaw
d1de4b1ea9
debug
2020-06-16 19:12:48 -07:00
Richard Liaw
d8705e1291
set
2020-06-16 13:19:15 -07:00
Richard Liaw
8cc49c5a03
some-allreduce
2020-06-16 12:17:01 -07:00
Richard Liaw
ef2af90f54
with-gpu
2020-06-15 21:59:27 -07:00
Richard Liaw
54951aa976
distributed
2020-06-15 01:30:25 -07:00
Richard Liaw
82a3a3b9a7
lint
2020-06-12 19:34:20 -07:00
Richard Liaw
e6536279d4
ray-init
2020-06-12 19:12:17 -07:00
Matthew Honnibal
a1c5b694be
Small fixes to train defaults
2020-06-12 02:22:13 +02:00
Sofie Van Landeghem
c0f4a1e43b
train is from-config by default ( #5575 )
...
* verbose and tag_map options
* adding init_tok2vec option and only changing the tok2vec that is specified
* adding omit_extra_lookups and verifying textcat config
* wip
* pretrain bugfix
* add replace and resume options
* train_textcat fix
* raw text functionality
* improve UX when KeyError or when input data can't be parsed
* avoid unnecessary access to goldparse in TextCat pipe
* save performance information in nlp.meta
* add noise_level to config
* move nn_parser's defaults to config file
* multitask in config - doesn't work yet
* scorer offering both F and AUC options, need to be specified in config
* add textcat verification code from old train script
* small fixes to config files
* clean up
* set default config for ner/parser to allow create_pipe to work as before
* two more test fixes
* small fixes
* cleanup
* fix NER pickling + additional unit test
* create_pipe as before
2020-06-12 02:02:07 +02:00
Ines Montani
d93cbeb14f
Add warning for loose version constraints ( #5536 )
...
* Add warning for loose version constraints
* Update wording [ci skip]
* Tweak error message
Co-authored-by: Matthew Honnibal <honnibal+gh@gmail.com>
2020-06-05 12:42:15 +02:00
Matthew Honnibal
8411d4f4e6
Merge pull request #5543 from svlandeg/feature/pretrain-config
...
pretrain from config
2020-06-04 19:07:12 +02:00
svlandeg
3ade455fd3
formatting
2020-06-04 16:09:55 +02:00
svlandeg
776d4f1190
cleanup
2020-06-04 16:07:30 +02:00
svlandeg
6b027d7689
remove duplicate model definition of tok2vec layer
2020-06-04 15:49:23 +02:00
svlandeg
1775f54a26
small little fixes
2020-06-03 22:17:02 +02:00
svlandeg
07886a3de3
rename init_tok2vec to resume
2020-06-03 22:00:25 +02:00
svlandeg
4ed6278663
small fixes to pretrain config, init_tok2vec TODO
2020-06-03 19:32:40 +02:00
svlandeg
ffe0451d09
pretrain from config
2020-06-03 14:45:00 +02:00
Ines Montani
a8875d4a4b
Fix typo
2020-06-03 14:42:39 +02:00
Ines Montani
4e0610d0d4
Update warning codes
2020-06-03 14:37:09 +02:00
Ines Montani
810fce3bb1
Merge branch 'develop' into master-tmp
2020-06-03 14:36:59 +02:00
Adriane Boyd
b0ee76264b
Remove debugging
2020-06-03 14:20:42 +02:00
Adriane Boyd
1d8168d1fd
Fix problems with lower and whitespace in variants
...
Port relevant changes from #5361 :
* Initialize lower flag explicitly
* Handle whitespace words from GoldParse correctly when creating raw
text with orth variants
2020-06-03 14:15:58 +02:00
Adriane Boyd
10d938f221
Update default cfg dir in train CLI
2020-06-03 14:15:50 +02:00
Adriane Boyd
f1f9c8b417
Port train CLI updates
...
Updates from #5362 and fix from #5387 :
* `train`:
* if training on GPU, only run evaluation/timing on CPU in the first
iteration
* if training is aborted, exit with a non-0 exit status
2020-06-03 14:03:43 +02:00
svlandeg
eac12cbb77
make dropout in embed layers configurable
2020-06-03 11:50:16 +02:00
svlandeg
e91485dfc4
add discard_oversize parameter, move optimizer to training subsection
2020-06-03 10:04:16 +02:00
svlandeg
03c58b488c
prevent infinite loop, custom warning
2020-06-03 10:00:21 +02:00
svlandeg
6504b7f161
Merge remote-tracking branch 'upstream/develop' into feature/pretrain-config
2020-06-03 08:30:16 +02:00
svlandeg
c5ac382f0a
fix name clash
2020-06-02 22:24:57 +02:00
svlandeg
2bf5111ecf
additional test with discard_oversize=False
2020-06-02 22:09:37 +02:00
svlandeg
aa6271b16c
extending algorithm to deal better with edge cases
2020-06-02 22:05:08 +02:00
svlandeg
f2e162fc60
it's only oversized if the tolerance level is also exceeded
2020-06-02 19:59:04 +02:00
svlandeg
ef834b4cd7
fix comments
2020-06-02 19:50:44 +02:00
svlandeg
6208d322d3
slightly more challenging unit test
2020-06-02 19:47:30 +02:00
svlandeg
6651fafd5c
using overflow buffer for examples within the tolerance margin
2020-06-02 19:43:39 +02:00
svlandeg
85b0597ed5
add test for minibatch util
2020-06-02 18:26:21 +02:00
svlandeg
5b350a6c99
bugfix of the bugfix
2020-06-02 17:49:33 +02:00
svlandeg
fdfd822936
rewrite minibatch_by_words function
2020-06-02 15:22:54 +02:00
svlandeg
ec52e7f886
add oversize examples before StopIteration returns
2020-06-02 13:21:55 +02:00
svlandeg
e0f9f448f1
remove Tensorizer
2020-06-01 23:38:48 +02:00
Ines Montani
b5ae2edcba
Merge pull request #5516 from explosion/feature/improve-model-version-deps
2020-05-31 12:54:01 +02:00
Ines Montani
dc186afdc5
Add warning
2020-05-30 15:34:54 +02:00
Ines Montani
b7aff6020c
Make functions more general purpose and update docstrings and tests
2020-05-30 15:18:53 +02:00
Ines Montani
a7e370bcbf
Don't override spaCy version
2020-05-30 15:03:18 +02:00
Ines Montani
e47e5a4b10
Use more sophisticated version parsing logic
2020-05-30 15:01:58 +02:00
Ines Montani
4fd087572a
WIP: improve model version deps
2020-05-28 12:51:37 +02:00