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
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
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
NAUSICAA\Julian
227f3b5dff
Merge branch 'master' of https://github.com/n-waves/ulmfit-multilingual into text_cols
2019-03-07 13:18:52 -03:00
Piotr Czapla
7b2ac9e94b
Add ability to use random-init=True
2019-03-01 17:41:07 +01:00
Piotr Czapla
11b2b14523
Fix -m ulmfit tar method
2019-03-01 17:39:58 +01:00
NAUSICAA\Julian
fd022c6826
Merge branch 'master' of https://github.com/n-waves/ulmfit-multilingual into text_cols
2019-02-27 23:30:59 -03:00
Marcin
f750586114
Remove redundant bptt param
2019-02-26 18:39:38 +01:00
Piotr Czapla
c6e0373170
Make models use bptt parameter
2019-02-26 18:06:23 +01:00
NAUSICAA\Julian
dd74a6d282
Remove previous multi field patch
2019-02-24 23:15:20 -03:00
NAUSICAA\Julian
06cd4d4d0b
Merge branch 'master' of https://github.com/n-waves/ulmfit-multilingual into text_cols
2019-02-24 22:58:46 -03:00
Piotr Czapla
260faa703c
Correct the label smoothing implementation
2019-02-22 16:55:35 +01:00
Piotr Czapla
015f04ec08
Training with noise & label smoothing
2019-02-22 12:02:18 +01:00
Piotr Czapla
99d6b22447
Correct noise generation training + convenience functions
2019-02-20 10:24:30 +01:00
NAUSICAA\Julian
13ae29a95d
Set mark_fields in True
2019-02-18 22:20:24 -03:00
Piotr Czapla
7dc7aac327
Merge all columns in classification task into first column
...
This should fix CLS issues.
2019-02-18 21:50:04 +01:00
Piotr Czapla
9fbcf56df3
Add different learning schedules, with default to the old schedule
...
use --lr-sched=1cycle for better results
2019-02-18 21:49:16 +01:00
NAUSICAA\Julian
458c06f779
Fix df name
2019-02-18 17:05:11 -03:00
NAUSICAA\Julian
dbb929e3ca
Adding more text cols to use all CLS data
2019-02-18 16:54:00 -03:00
Piotr Czapla
ce6cc607ae
Add saving itos.pkl so that the LM can be used to finetuning
2019-02-17 23:18:42 +01:00
Piotr Czapla
c3276da062
Fixing Imdb loading
2019-02-17 23:18:12 +01:00
Piotr Czapla
119417fb6e
Disable early stopping as it was causing OOMs
2019-02-17 23:04:25 +01:00
Piotr Czapla
490c792278
Upgrade to the recent the todays version of Fastai
2019-02-17 23:03:54 +01:00
Piotr Czapla
8733487d55
Make the validate vs train decision based on the existance of cls_last.pth istead of a model directory
2019-02-15 01:17:53 +01:00
Piotr Czapla
5dced1e488
Remove bidir
2019-02-15 01:16:37 +01:00
Piotr Czapla
0e6534ad7b
Expose num_lm_epochs in ulmfit eval
2019-02-15 01:11:39 +01:00
Piotr Czapla
0f084168c1
Merge branch 'master' of https://github.com/n-waves/ulmfit-multilingual
2019-02-14 22:35:29 +01:00
Piotr Czapla
cd47b3b5dc
Fix use_moses=True for mldoc so that it is identical to wiki with uses_moses=False
...
The issue was that Moses was executed after pre_rules when use_moses = True, But when data set was pre tokenized with Moses (use_moses=False) the pre_rules were executed after.
So our wikipedia had the following processing:
- raw text
- Moses
- pre_rules
- split(' ') # fastai BaseTokenizer
- post_rules
- sentence piece
While mldoc had the following tokenziation
- raw text
- pre_rules
- Moses
- post_rules
- sentence piece
After fix I've retrained the classfiers (without finetuning) and I haven't notice huge changes in the performance. 4 languages received slight improvment 4 got a slight decrease in performance.
2019-02-14 22:35:20 +01:00
Piotr Czapla
c28c0fde16
Make ulmfit eval more secure and give more flexibility in dataset_template
...
The dataset_template can use lang as additional token to construct globs patterns.
2019-02-14 22:28:25 +01:00
Marcin
fdac9f7ccd
Save only the best LM model
2019-02-14 14:01:33 +01:00
Marcin
b609951561
Fix path of pretrained model
2019-02-14 00:00:56 +01:00
Piotr Czapla
1340f4235c
Improve ulmfit eval to allow for zeroshot laser evaluation
2019-02-13 15:29:55 +01:00
Marcin
0e2211dfaf
Merge branch 'master' into use-configs
2019-02-13 14:10:59 +01:00
Marcin