This repository has been archived by the owner on Jan 15, 2024. It is now read-only.
Update for Block API #1261
Merged
Update for Block API #1261
Conversation
This file contains bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
- Remove params and prefix arguments for MXNet 2 and update parameter sharing implementation - Remove Block.name_scope() for MXNet 2 - Remove self.params.get() and self.params.get_constant()
sxjscience
reviewed
Jul 10, 2020
|
@zheyuye @hymzoque Would you also help review? |
|
zheyuye
reviewed
Jul 11, 2020
| self.mlm_decoder[-1].share_parameters(word_embed_params) | ||
| self.backbone_model.token_type_embed.share_parameters(token_type_embed_params) | ||
| self.backbone_model.token_pos_embed.share_parameters(token_pos_embed_params) | ||
| self.backbone_model.embed_layer_norm.share_parameters(embed_layer_norm_params) |
There was a problem hiding this comment.
From the above code, we can share weights in two ways: 1. weight = weight 2.share_parameters, is that correct? Do we need consistency on this part?
There was a problem hiding this comment.
Yes, both is supported. The 1. weight = weight way works well if you only need to replace a single parameter and you know it's location. The second way is more general and doesn't require users to specify the position of the weight (but just pass a dictionary containing the weights with proper name).
zheyuye
reviewed
Jul 11, 2020
There was a problem hiding this comment.
Thanks to @leezu for revising, this is generally good but may require some extra efforts on the conversion toolkits which are highly dependent on the prefix as https://github.com/leezu/gluon-nlp/blob/a79101da3a40d5212e419fa1f46a40e9ad3e7eb3/scripts/conversion_toolkits/convert_tf_hub_model.py#L134-L166
| params=embed_layer.collect_params( | ||
| '(.*_embed|.*_inter_proj)')) | ||
| div_val=div_val) | ||
| layer_with_shared_proj_embed.share_parameters(embed_layer.collect_params('(.*_embed|.*_inter_proj)')) |
There was a problem hiding this comment.
It seems that we cant locate to these parameters since prefix are removed
https://github.com/leezu/gluon-nlp/blob/a79101da3a40d5212e419fa1f46a40e9ad3e7eb3/src/gluonnlp/layers.py#L916-L917
There was a problem hiding this comment.
It's because the regex needs to be updated. It contains _ character which is no longer used and should be replaced by \..
|
@dmlc/gluon-nlp-committers let's halt other merges in the numpy branch to yield for this change. |
sxjscience
reviewed
Jul 12, 2020
tests/test_models.py
Outdated
| net.hybridize() | ||
| num_params, num_fixed_params = count_parameters(net.collect_params()) | ||
| assert num_params > 0 | ||
| @pytest.mark.parametrize('name', list_backbone_names()) |
There was a problem hiding this comment.
I moved to use sequential test intentionally because running multiple tests in parallel may cause some memory issues.
There was a problem hiding this comment.
In that case we may mark it as serial. Having a single large tests makes it very hard to reproduce failures of specific models, because the test will always test all models. It's not a good development experience.
There was a problem hiding this comment.
May be we should remove the forward test of xlmr, which is too large.
sxjscience
reviewed
Jul 12, 2020
| @@ -120,6 +120,7 @@ def test_adaptive_embedding(vocab_size, cutoffs, embed_size, units, div_val): | |||
| [1000, None, 1.0]]) | |||
| @pytest.mark.parametrize('embed_size', [128]) | |||
| @pytest.mark.parametrize('in_units', [16]) | |||
| # TODO This test even passes without sharing the parameters. It needs to be improved. | |||
There was a problem hiding this comment.
If that's the case, we should revise the test. (May be in a later PR).
There was a problem hiding this comment.
Yes, we should revise the test in a later PR. I just noticed that the test passed prior even when I disabled the parameter sharing or even when the parameter sharing is handled wrongly (such as using invalid regex in collect_params)
There was a problem hiding this comment.
Need to investigate the test after this gets merged. I'm also having issues about parameter sharing in the layout PR so need to wait for this one.
@sxjscience is this performance as expected? |
Yes, we need to check "best_exact" and "best_f1". |
|
@szha @sxjscience Yes, the |
8 tasks
@zheyuye generally the scripts can be updated by replacing the It' can be done in a separate PR if we decide to keep the conversion scripts in the release (which may require adding tests). |
|
@leezu This sounds great. Let's leave conversion scripts alone for now and re-revise these with some useful test cases once this PR merged. |
|
|
|
@leezu It's expected. I propose to merge this in. |
Codecov Report
@@ Coverage Diff @@
## numpy #1261 +/- ##
==========================================
+ Coverage 82.52% 82.53% +0.01%
==========================================
Files 38 38
Lines 5500 5446 -54
==========================================
- Hits 4539 4495 -44
+ Misses 961 951 -10
|
sxjscience
approved these changes
Jul 16, 2020
zheyuye
added a commit
to zheyuye/gluon-nlp
that referenced
this pull request
Jul 17, 2020
commit 35a586676036f627bffd0d3c753c6cd0a70d63cf
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 17 10:10:14 2020 +0800
Squashed commit of the following:
commit 673344d
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 15 22:43:07 2020 +0800
CharTokenizer
commit 8dabfd6
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 15 15:47:24 2020 +0800
lowercase
commit f5c94a6
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jul 14 17:45:28 2020 +0800
test
commit dc55fc9
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jul 14 05:45:01 2020 +0800
tiny update on run_squad
commit 4defc7a
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 23:18:08 2020 +0800
update testings
commit 2719e81
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 23:08:32 2020 +0800
re-upload xlmr
commit cd0509d
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 22:30:47 2020 +0800
fix get_pretrained
commit 8ed8a72
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 22:28:13 2020 +0800
re-upload roberta
commit 5811d40
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 18:27:23 2020 +0800
update
commit 44a09a3
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sat Jul 11 15:06:33 2020 +0800
fix
commit 4074a26
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 10 16:08:49 2020 +0800
inference without horovod
commit 31cb953
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 18:41:55 2020 +0800
update
commit 838be2a
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 15:14:39 2020 +0800
horovod for squad
commit 1d374a2
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 12:09:19 2020 +0800
fix
commit e4fba39
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 10:35:08 2020 +0800
remove multiply_grads
commit 007f07e
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jul 7 11:26:38 2020 +0800
multiply_grads
commit b8c85bb
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 6 12:28:56 2020 +0800
fix ModelForQABasic
commit 0e13a58
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sat Jul 4 18:42:12 2020 +0800
clip_grad_global_norm with zeros max_grad_norm
commit bd270f2
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 3 20:21:31 2020 +0800
fix roberta
commit 4fc564c
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 3 19:36:08 2020 +0800
update hyper-parameters of adamw
commit 59cffbf
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 3 16:25:46 2020 +0800
try
commit a84f782
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 2 20:39:03 2020 +0800
fix mobilebert
commit 4bc3a96
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 2 11:14:39 2020 +0800
layer-wise decay
commit 07186d5
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 2 02:14:43 2020 +0800
revise
commit a5a6475
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 1 19:50:20 2020 +0800
topk
commit 34ee884
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 1 19:25:09 2020 +0800
index_update
commit 74178e2
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 1 00:48:32 2020 +0800
rename
commit fa011aa
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 23:40:28 2020 +0800
update
commit 402d625
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 21:40:30 2020 +0800
multiprocessing for wiki
commit ddbde75
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 20:41:35 2020 +0800
fix bookcorpus
commit 6cc5ccd
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 16:39:12 2020 +0800
fix wiki
commit 9773efd
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 15:52:13 2020 +0800
fix openwebtext
commit 1fb8eb8
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 19:51:25 2020 +0800
upload gluon_electra_small_owt
commit ca83fac
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 18:09:48 2020 +0800
revise train_transformer
commit 1450f5c
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 18:07:04 2020 +0800
revise
commit b460bbe
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 17:24:00 2020 +0800
repeat for pretraining
commit 8ee381b
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 17:06:43 2020 +0800
repeat
commit aea936f
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 16:39:22 2020 +0800
fix mobilebert
commit eead164
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 18:44:28 2020 +0800
fix
commit 8645115
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 17:27:43 2020 +0800
update
commit 2b7f7a3
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 17:18:00 2020 +0800
fix roberta
commit 86702fe
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 16:27:43 2020 +0800
use_segmentation
commit 6d03d7a
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 15:52:40 2020 +0800
fix
commit 5c0ca43
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 15:49:48 2020 +0800
fix token_ids
commit ff7aae8
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 13:56:07 2020 +0800
fix xlmr
commit 2070b86
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 13:54:26 2020 +0800
fix roberta
commit 70a1887
Author: Leonard Lausen <lausen@amazon.com>
Date: Fri Jul 17 00:07:08 2020 +0000
Update for Block API (dmlc#1261)
- Remove params and prefix arguments for MXNet 2 and update
parameter sharing implementation
- Remove Block.name_scope() for MXNet 2
- Remove self.params.get() and self.params.get_constant()
commit ea9152b
Author: Xingjian Shi <xshiab@connect.ust.hk>
Date: Thu Jul 16 15:42:04 2020 -0700
Fixes to make the CI more stable (dmlc#1265)
* Some fixes to make the CI more stable
* add retries
* Update tokenizers.py
commit a646c34
Author: ht <wawawa@akane.waseda.jp>
Date: Sun Jul 12 02:49:53 2020 +0800
[FEATURE] update backtranslation and add multinomial sampler (dmlc#1259)
* back translation bash
* split "lang-pair" para in clean_tok_para_corpus
* added clean_tok_mono_corpus
* fix
* add num_process para
* fix
* fix
* add yml
* rm yml
* update cfg name
* update evaluate
* added max_update / save_interval_update params
* fix
* fix
* multi gpu inference
* fix
* update
* update multi gpu inference
* fix
* fix
* split evaluate and parallel infer
* fix
* test
* fix
* update
* add comments
* fix
* remove todo comment
* revert remove todo comment
* raw lines remove duplicated '\n'
* update multinomaial sampler
* fix
* fix
* fix
* fix
* sampling
* update script
* fix
* add test_case with k > 1 in topk sampling
* fix multinomial sampler
* update docs
* comments situation eos_id = None
* fix
Co-authored-by: Hu <huta@a483e74650ff.ant.amazon.com>
commit 83e1f13
Author: Leonard Lausen <lausen@amazon.com>
Date: Thu Jul 9 20:57:55 2020 -0700
Use Amazon S3 Transfer Acceleration (dmlc#1260)
commit cd48efd
Author: Leonard Lausen <lausen@amazon.com>
Date: Tue Jul 7 17:39:42 2020 -0700
Update codecov action to handle different OS and Python versions (dmlc#1254)
codecov/codecov-action#80 (comment)
commit 689eba9
Author: Sheng Zha <szha@users.noreply.github.com>
Date: Tue Jul 7 09:55:34 2020 -0700
[CI] AWS batch job tool for GluonNLP (Part I) (dmlc#1251)
* AWS batch job tool for GluonNLP
* limit range
Co-authored-by: Xingjian Shi <xshiab@connect.ust.hk>
commit e06ff01
Author: Leonard Lausen <lausen@amazon.com>
Date: Tue Jul 7 08:36:24 2020 -0700
Pin mxnet version range on CI (dmlc#1257)
3 tasks
sxjscience
pushed a commit
that referenced
this pull request
Aug 1, 2020
* fix roberta
* fix xlmr
* fix token_ids
* fix
* use_segmentation
* fix roberta
* update
* fix
* fix mobilebert
* repeat
* repeat for pretraining
* revise
* revise train_transformer
* upload gluon_electra_small_owt
* fix openwebtext
* fix wiki
* fix bookcorpus
* multiprocessing for wiki
* update
* rename
* index_update
* topk
* revise
* layer-wise decay
* fix mobilebert
* try
* update hyper-parameters of adamw
* fix roberta
* clip_grad_global_norm with zeros max_grad_norm
* fix ModelForQABasic
* multiply_grads
* remove multiply_grads
* fix
* horovod for squad
* update
* inference without horovod
* fix
* update
* re-upload roberta
* fix get_pretrained
* re-upload xlmr
* update testings
* tiny update on run_squad
* test
* lowercase
* CharTokenizer
* Squashed commit of the following:
commit 35a586676036f627bffd0d3c753c6cd0a70d63cf
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 17 10:10:14 2020 +0800
Squashed commit of the following:
commit 673344d
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 15 22:43:07 2020 +0800
CharTokenizer
commit 8dabfd6
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 15 15:47:24 2020 +0800
lowercase
commit f5c94a6
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jul 14 17:45:28 2020 +0800
test
commit dc55fc9
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jul 14 05:45:01 2020 +0800
tiny update on run_squad
commit 4defc7a
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 23:18:08 2020 +0800
update testings
commit 2719e81
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 23:08:32 2020 +0800
re-upload xlmr
commit cd0509d
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 22:30:47 2020 +0800
fix get_pretrained
commit 8ed8a72
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 22:28:13 2020 +0800
re-upload roberta
commit 5811d40
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 13 18:27:23 2020 +0800
update
commit 44a09a3
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sat Jul 11 15:06:33 2020 +0800
fix
commit 4074a26
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 10 16:08:49 2020 +0800
inference without horovod
commit 31cb953
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 18:41:55 2020 +0800
update
commit 838be2a
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 15:14:39 2020 +0800
horovod for squad
commit 1d374a2
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 12:09:19 2020 +0800
fix
commit e4fba39
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 9 10:35:08 2020 +0800
remove multiply_grads
commit 007f07e
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jul 7 11:26:38 2020 +0800
multiply_grads
commit b8c85bb
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jul 6 12:28:56 2020 +0800
fix ModelForQABasic
commit 0e13a58
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sat Jul 4 18:42:12 2020 +0800
clip_grad_global_norm with zeros max_grad_norm
commit bd270f2
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 3 20:21:31 2020 +0800
fix roberta
commit 4fc564c
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 3 19:36:08 2020 +0800
update hyper-parameters of adamw
commit 59cffbf
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Fri Jul 3 16:25:46 2020 +0800
try
commit a84f782
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 2 20:39:03 2020 +0800
fix mobilebert
commit 4bc3a96
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 2 11:14:39 2020 +0800
layer-wise decay
commit 07186d5
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Thu Jul 2 02:14:43 2020 +0800
revise
commit a5a6475
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 1 19:50:20 2020 +0800
topk
commit 34ee884
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 1 19:25:09 2020 +0800
index_update
commit 74178e2
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Wed Jul 1 00:48:32 2020 +0800
rename
commit fa011aa
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 23:40:28 2020 +0800
update
commit 402d625
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 21:40:30 2020 +0800
multiprocessing for wiki
commit ddbde75
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 20:41:35 2020 +0800
fix bookcorpus
commit 6cc5ccd
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 16:39:12 2020 +0800
fix wiki
commit 9773efd
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Tue Jun 30 15:52:13 2020 +0800
fix openwebtext
commit 1fb8eb8
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 19:51:25 2020 +0800
upload gluon_electra_small_owt
commit ca83fac
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 18:09:48 2020 +0800
revise train_transformer
commit 1450f5c
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 18:07:04 2020 +0800
revise
commit b460bbe
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 17:24:00 2020 +0800
repeat for pretraining
commit 8ee381b
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 17:06:43 2020 +0800
repeat
commit aea936f
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Mon Jun 29 16:39:22 2020 +0800
fix mobilebert
commit eead164
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 18:44:28 2020 +0800
fix
commit 8645115
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 17:27:43 2020 +0800
update
commit 2b7f7a3
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 17:18:00 2020 +0800
fix roberta
commit 86702fe
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 16:27:43 2020 +0800
use_segmentation
commit 6d03d7a
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 15:52:40 2020 +0800
fix
commit 5c0ca43
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 15:49:48 2020 +0800
fix token_ids
commit ff7aae8
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 13:56:07 2020 +0800
fix xlmr
commit 2070b86
Author: ZheyuYe <zheyu.ye1995@gmail.com>
Date: Sun Jun 28 13:54:26 2020 +0800
fix roberta
commit 70a1887
Author: Leonard Lausen <lausen@amazon.com>
Date: Fri Jul 17 00:07:08 2020 +0000
Update for Block API (#1261)
- Remove params and prefix arguments for MXNet 2 and update
parameter sharing implementation
- Remove Block.name_scope() for MXNet 2
- Remove self.params.get() and self.params.get_constant()
commit ea9152b
Author: Xingjian Shi <xshiab@connect.ust.hk>
Date: Thu Jul 16 15:42:04 2020 -0700
Fixes to make the CI more stable (#1265)
* Some fixes to make the CI more stable
* add retries
* Update tokenizers.py
commit a646c34
Author: ht <wawawa@akane.waseda.jp>
Date: Sun Jul 12 02:49:53 2020 +0800
[FEATURE] update backtranslation and add multinomial sampler (#1259)
* back translation bash
* split "lang-pair" para in clean_tok_para_corpus
* added clean_tok_mono_corpus
* fix
* add num_process para
* fix
* fix
* add yml
* rm yml
* update cfg name
* update evaluate
* added max_update / save_interval_update params
* fix
* fix
* multi gpu inference
* fix
* update
* update multi gpu inference
* fix
* fix
* split evaluate and parallel infer
* fix
* test
* fix
* update
* add comments
* fix
* remove todo comment
* revert remove todo comment
* raw lines remove duplicated '\n'
* update multinomaial sampler
* fix
* fix
* fix
* fix
* sampling
* update script
* fix
* add test_case with k > 1 in topk sampling
* fix multinomial sampler
* update docs
* comments situation eos_id = None
* fix
Co-authored-by: Hu <huta@a483e74650ff.ant.amazon.com>
commit 83e1f13
Author: Leonard Lausen <lausen@amazon.com>
Date: Thu Jul 9 20:57:55 2020 -0700
Use Amazon S3 Transfer Acceleration (#1260)
commit cd48efd
Author: Leonard Lausen <lausen@amazon.com>
Date: Tue Jul 7 17:39:42 2020 -0700
Update codecov action to handle different OS and Python versions (#1254)
codecov/codecov-action#80 (comment)
commit 689eba9
Author: Sheng Zha <szha@users.noreply.github.com>
Date: Tue Jul 7 09:55:34 2020 -0700
[CI] AWS batch job tool for GluonNLP (Part I) (#1251)
* AWS batch job tool for GluonNLP
* limit range
Co-authored-by: Xingjian Shi <xshiab@connect.ust.hk>
commit e06ff01
Author: Leonard Lausen <lausen@amazon.com>
Date: Tue Jul 7 08:36:24 2020 -0700
Pin mxnet version range on CI (#1257)
* frozen_params
* remove conversion to a sperate pr
* fix
* fix
* update
* test
* revise
* update performance numbers
* update apply_layerwisw_decay
* use shuffle
* fix mobilebert
* fix vocab_file
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
parameter sharing implementation
CI will pass once apache/mxnet#18619 is merged
Please review the API changes. Scripts are not updated by this PR, but at least some will be updated and included here following verification of fine-tuning performance and NMT training. Note that it's not required to re-generate the parameter files.
Thanks to @acphile for his hard work on the Gluon API refactor on the MXNet side (apache/mxnet@cb54a4a)