url stringlengths 62 66 | repository_url stringclasses 1
value | labels_url stringlengths 76 80 | comments_url stringlengths 71 75 | events_url stringlengths 69 73 | html_url stringlengths 50 56 | id int64 377M 2.15B | node_id stringlengths 18 32 | number int64 1 29.2k | title stringlengths 1 487 | user dict | labels list | state stringclasses 2
values | locked bool 2
classes | assignee dict | assignees list | comments list | created_at int64 1.54k 1.71k | updated_at int64 1.54k 1.71k | closed_at int64 1.54k 1.71k ⌀ | author_association stringclasses 4
values | active_lock_reason stringclasses 2
values | body stringlengths 0 234k ⌀ | reactions dict | timeline_url stringlengths 71 75 | state_reason stringclasses 3
values | draft bool 2
classes | pull_request dict |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
https://api.github.com/repos/huggingface/transformers/issues/1310 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1310/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1310/comments | https://api.github.com/repos/huggingface/transformers/issues/1310/events | https://github.com/huggingface/transformers/issues/1310 | 496,730,008 | MDU6SXNzdWU0OTY3MzAwMDg= | 1,310 | Redundant sep_token_extra option for RoBERTa Fine-tuning | {
"login": "todpole3",
"id": 4227871,
"node_id": "MDQ6VXNlcjQyMjc4NzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4227871?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/todpole3",
"html_url": "https://github.com/todpole3",
"followers_url": "https://api.github.com/users/todpo... | [] | closed | false | null | [] | [
"Myle Ott from Facebook commented on the Twitter thread (https://twitter.com/myleott/status/1175750596630056961) confirming that there is an extra separator being used, so there should be details I did not understand well.\r\n\r\nI will revisit this issue when I understand it better.",
"The `sep_token_extra` para... | 1,569 | 1,569 | 1,569 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....): RoBERTa
Language I am using the model on (English, Chinese....): English
## Context
I was reading the code on RoBERTa fine-tuning and noticed the [`sep_token_extra` option](https://github.com/huggingface/pytorch-transformers/sea... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1310/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1310/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1309 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1309/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1309/comments | https://api.github.com/repos/huggingface/transformers/issues/1309/events | https://github.com/huggingface/transformers/issues/1309 | 496,697,242 | MDU6SXNzdWU0OTY2OTcyNDI= | 1,309 | Best loss | {
"login": "jasonmusespresso",
"id": 24786001,
"node_id": "MDQ6VXNlcjI0Nzg2MDAx",
"avatar_url": "https://avatars.githubusercontent.com/u/24786001?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jasonmusespresso",
"html_url": "https://github.com/jasonmusespresso",
"followers_url": "https://... | [] | closed | false | null | [] | [] | 1,569 | 1,569 | 1,569 | NONE | null | I am building a classifier by adopting codes from `run_glue.py`. There is plenty of optimization in training and tuning hyperparameters. Could anyone explain the difference between loss, tr_loss and logging_loss in these parts?
https://github.com/huggingface/pytorch-transformers/blob/a2d4950f5c909f7bb4ea7c06afa6cdecde... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1309/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1309/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1308 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1308/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1308/comments | https://api.github.com/repos/huggingface/transformers/issues/1308/events | https://github.com/huggingface/transformers/issues/1308 | 496,660,604 | MDU6SXNzdWU0OTY2NjA2MDQ= | 1,308 | Planned support for new Grover 1.5B models? | {
"login": "GenTxt",
"id": 22547261,
"node_id": "MDQ6VXNlcjIyNTQ3MjYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22547261?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/GenTxt",
"html_url": "https://github.com/GenTxt",
"followers_url": "https://api.github.com/users/GenTxt/fo... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"No short-term plan to implement this ourselves, but we'd welcome a PR (especially one involving the original authors for validation).",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
... | 1,569 | 1,575 | 1,575 | NONE | null | Thanks for the great repo.
Just wondering if there's any planned support for the new Grover 1.5B models?
https://github.com/rowanz/grover (original 1.5B now available via download_model.py)
https://github.com/vanyacohen/opengpt2-1.5B-gpu-inference (slightly different variation)
Cheers
| {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1308/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1308/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1307 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1307/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1307/comments | https://api.github.com/repos/huggingface/transformers/issues/1307/events | https://github.com/huggingface/transformers/issues/1307 | 496,563,723 | MDU6SXNzdWU0OTY1NjM3MjM= | 1,307 | mask_tokens sometimes masks special tokens | {
"login": "kaushaltrivedi",
"id": 3465437,
"node_id": "MDQ6VXNlcjM0NjU0Mzc=",
"avatar_url": "https://avatars.githubusercontent.com/u/3465437?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kaushaltrivedi",
"html_url": "https://github.com/kaushaltrivedi",
"followers_url": "https://api.gith... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi, thank you for the bug report. Indeed, this does seem problematic. I'm looking into it.",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,569 | 1,575 | 1,575 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....): RoBERTa
Language I am using the model on (English, Chinese....):
The problem arise when using:
* [ ] the official example scripts: (give details) run_lm_finetuning
* [ ] my own modified scripts: (give details)
The tasks I am w... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1307/reactions",
"total_count": 3,
"+1": 3,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1307/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1306 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1306/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1306/comments | https://api.github.com/repos/huggingface/transformers/issues/1306/events | https://github.com/huggingface/transformers/issues/1306 | 496,440,436 | MDU6SXNzdWU0OTY0NDA0MzY= | 1,306 | Which model is best to used for language model rescoring for ASR | {
"login": "LearnedVector",
"id": 8495552,
"node_id": "MDQ6VXNlcjg0OTU1NTI=",
"avatar_url": "https://avatars.githubusercontent.com/u/8495552?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LearnedVector",
"html_url": "https://github.com/LearnedVector",
"followers_url": "https://api.github.... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Same as https://github.com/google-research/bert/issues/35",
"And https://github.com/huggingface/transformers/issues/37",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n",
"GPT2 can ... | 1,568 | 1,586 | 1,576 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
Hello all, I am wanting to use this library to rescore an output from an automatic speech recognition model. Still learning a lot about language model so out of curiosity for anyone who's tried, which model has given you the best perfor... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1306/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1306/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1305 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1305/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1305/comments | https://api.github.com/repos/huggingface/transformers/issues/1305/events | https://github.com/huggingface/transformers/issues/1305 | 496,407,463 | MDU6SXNzdWU0OTY0MDc0NjM= | 1,305 | Dataset format and Best Practices For Language Model Fine-tuning | {
"login": "HanGuo97",
"id": 18187806,
"node_id": "MDQ6VXNlcjE4MTg3ODA2",
"avatar_url": "https://avatars.githubusercontent.com/u/18187806?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/HanGuo97",
"html_url": "https://github.com/HanGuo97",
"followers_url": "https://api.github.com/users/Han... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"I am facing the same issue as there is no proper format available for defining the train and test dataset. \r\nAs usual, I use .csv file in a format of columns with (UID, Text, and Labels). But according to the Wiki.txt its more of arbitrary format.\r\n\r\nAny help would be appreciated.",
"I'm having the same is... | 1,568 | 1,575 | 1,575 | NONE | null | ## ❓ Questions & Help
Hi, thanks for making this code base available!
I have two questions, one on the input format of for fine-tuning the language model on custom dataset, and one on (unreasonably-)long data preprocessing time. Thanks in advance for any help!
- I'm trying to fine-tune the BERT Model on an ext... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1305/reactions",
"total_count": 8,
"+1": 8,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1305/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1304 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1304/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1304/comments | https://api.github.com/repos/huggingface/transformers/issues/1304/events | https://github.com/huggingface/transformers/issues/1304 | 496,349,252 | MDU6SXNzdWU0OTYzNDkyNTI= | 1,304 | max_len_single_sentence should be max_len - 2 for RoBERTa | {
"login": "kaushaltrivedi",
"id": 3465437,
"node_id": "MDQ6VXNlcjM0NjU0Mzc=",
"avatar_url": "https://avatars.githubusercontent.com/u/3465437?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kaushaltrivedi",
"html_url": "https://github.com/kaushaltrivedi",
"followers_url": "https://api.gith... | [] | closed | false | {
"login": "julien-c",
"id": 326577,
"node_id": "MDQ6VXNlcjMyNjU3Nw==",
"avatar_url": "https://avatars.githubusercontent.com/u/326577?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/julien-c",
"html_url": "https://github.com/julien-c",
"followers_url": "https://api.github.com/users/julien-... | [
{
"login": "julien-c",
"id": 326577,
"node_id": "MDQ6VXNlcjMyNjU3Nw==",
"avatar_url": "https://avatars.githubusercontent.com/u/326577?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/julien-c",
"html_url": "https://github.com/julien-c",
"followers_url": "https://api.gith... | [
"I think you may be right and we've been meaning to fix this. cf recent discussion @LysandreJik @VictorSanh ",
"thanks. Adding LM fine-tuning to fast-bert. Have added a workaround for now :)",
"Also see https://github.com/pytorch/fairseq/issues/1187"
] | 1,568 | 1,569 | 1,569 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....): RoBERTa
Language I am using the model on (English, Chinese....): English
The problem arise when using:
* [ ] the official example scripts: (give details) run_lm_finetuning
* [ ] my own modified scripts: (give details)
The task... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1304/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1304/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1303 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1303/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1303/comments | https://api.github.com/repos/huggingface/transformers/issues/1303/events | https://github.com/huggingface/transformers/issues/1303 | 496,343,629 | MDU6SXNzdWU0OTYzNDM2Mjk= | 1,303 | Getting an unexpected EOF when trying to download 'bert-large-uncased-whole-word-masking-finetuned-squad' model. | {
"login": "agrawalarpan",
"id": 4469890,
"node_id": "MDQ6VXNlcjQ0Njk4OTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/4469890?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/agrawalarpan",
"html_url": "https://github.com/agrawalarpan",
"followers_url": "https://api.github.com... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n",
"In my environment, **it works as expected**!\r\n\r\n_Environment_:\r\n\r\n- **Python**: 3.6.9\r\n- **O.S.** : Linux-4.15.0-70-generic-x... | 1,568 | 1,580 | 1,580 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....): Bert
Language I am using the model on (English, Chinese....): English
The problem arise when using:
* [ ] the official example scripts: (give details): BertForQuestionAnswering
* [ ] my own modified scripts: (give details)
The... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1303/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1303/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1302 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1302/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1302/comments | https://api.github.com/repos/huggingface/transformers/issues/1302/events | https://github.com/huggingface/transformers/issues/1302 | 496,339,584 | MDU6SXNzdWU0OTYzMzk1ODQ= | 1,302 | Rectified Adam + LARS | {
"login": "i404788",
"id": 50617709,
"node_id": "MDQ6VXNlcjUwNjE3NzA5",
"avatar_url": "https://avatars.githubusercontent.com/u/50617709?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/i404788",
"html_url": "https://github.com/i404788",
"followers_url": "https://api.github.com/users/i40478... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"From what I can tell, Radam makes automatic warmup and LARS is good but requires more calculations per batch. Before implementing it here it's worth do testing to tell if it's a good idea.",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further... | 1,568 | 1,580 | 1,580 | NONE | null | ## 🚀 Feature
There has been a lot of buzz around the new Radam and Ralamb (Radam + LARS) optimizers, and I was wondering if it could also be implemented in pytorch-transformers.
## Motivation
It seems to have consistent performance improvements. It also seems to handle different learning rates a lot better. And... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1302/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1302/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1301 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1301/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1301/comments | https://api.github.com/repos/huggingface/transformers/issues/1301/events | https://github.com/huggingface/transformers/pull/1301 | 496,303,595 | MDExOlB1bGxSZXF1ZXN0MzE5NjgyMjc3 | 1,301 | RBERT implementation | {
"login": "RichJackson",
"id": 7306627,
"node_id": "MDQ6VXNlcjczMDY2Mjc=",
"avatar_url": "https://avatars.githubusercontent.com/u/7306627?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RichJackson",
"html_url": "https://github.com/RichJackson",
"followers_url": "https://api.github.com/us... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/1301?src=pr&el=h1) Report\n> Merging [#1301](https://codecov.io/gh/huggingface/transformers/pull/1301?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/2dc8cb87341223e86220516951bb4ad84f880b4a?src=pr&el=desc) will **i... | 1,568 | 1,592 | 1,571 | NONE | null | As per #1250, this PR describes an additional classification head for BERT for relationship classification tasks. This work is originally documented in [this paper](https://arxiv.org/pdf/1905.08284.pdf). In addition, the new head can be used with RoBERTa, producing a new SOTA as far as I know....
I have included a n... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1301/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1301/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1301",
"html_url": "https://github.com/huggingface/transformers/pull/1301",
"diff_url": "https://github.com/huggingface/transformers/pull/1301.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1301.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1300 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1300/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1300/comments | https://api.github.com/repos/huggingface/transformers/issues/1300/events | https://github.com/huggingface/transformers/issues/1300 | 496,203,995 | MDU6SXNzdWU0OTYyMDM5OTU= | 1,300 | ❓ Why the criterion of XLNet LMHeadModel use ignore_index = -1 ? | {
"login": "astariul",
"id": 43774355,
"node_id": "MDQ6VXNlcjQzNzc0MzU1",
"avatar_url": "https://avatars.githubusercontent.com/u/43774355?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/astariul",
"html_url": "https://github.com/astariul",
"followers_url": "https://api.github.com/users/ast... | [] | closed | false | null | [] | [
"you should set padding and all the labels to be ignored to -1. In Bert/XLNet training, we usually only use 15% of the tokens as labels."
] | 1,568 | 1,570 | 1,570 | CONTRIBUTOR | null | In the XLNetLMHeadModel, the criterion used to compute the loss uses `ignore_index=-1` :
https://github.com/huggingface/pytorch-transformers/blob/9f995b99d4c4067662c3bd4f1274315c0839deeb/pytorch_transformers/modeling_xlnet.py#L927-L931
**Why ?**
Isn't it supposed to ignore the padding index ID, i.e. 5 ? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1300/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1300/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1299 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1299/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1299/comments | https://api.github.com/repos/huggingface/transformers/issues/1299/events | https://github.com/huggingface/transformers/issues/1299 | 496,115,778 | MDU6SXNzdWU0OTYxMTU3Nzg= | 1,299 | What is the best CPU inference acceleration solution for BERT now? | {
"login": "guotong1988",
"id": 4702353,
"node_id": "MDQ6VXNlcjQ3MDIzNTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/4702353?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/guotong1988",
"html_url": "https://github.com/guotong1988",
"followers_url": "https://api.github.com/us... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Give us a little more details about your `(latency, compute)` constraints.",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,574 | 1,574 | CONTRIBUTOR | null | Thank you very much.
Thank you very much.
Thank you very much. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1299/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1299/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1298 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1298/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1298/comments | https://api.github.com/repos/huggingface/transformers/issues/1298/events | https://github.com/huggingface/transformers/pull/1298 | 496,105,829 | MDExOlB1bGxSZXF1ZXN0MzE5NTI2NTc1 | 1,298 | fix annotation | {
"login": "Maxpa1n",
"id": 34930289,
"node_id": "MDQ6VXNlcjM0OTMwMjg5",
"avatar_url": "https://avatars.githubusercontent.com/u/34930289?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Maxpa1n",
"html_url": "https://github.com/Maxpa1n",
"followers_url": "https://api.github.com/users/Maxpa1... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1298?src=pr&el=h1) Report\n> Merging [#1298](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1298?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/9f995b99d4c4067662c3bd4f1274315c0839deeb?... | 1,568 | 1,568 | 1,568 | CONTRIBUTOR | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1298/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1298/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1298",
"html_url": "https://github.com/huggingface/transformers/pull/1298",
"diff_url": "https://github.com/huggingface/transformers/pull/1298.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1298.patch",
"merged_at": 1... | |
https://api.github.com/repos/huggingface/transformers/issues/1297 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1297/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1297/comments | https://api.github.com/repos/huggingface/transformers/issues/1297/events | https://github.com/huggingface/transformers/pull/1297 | 495,869,185 | MDExOlB1bGxSZXF1ZXN0MzE5MzMzNDA1 | 1,297 | add support for file I/O | {
"login": "rajarsheem",
"id": 6441313,
"node_id": "MDQ6VXNlcjY0NDEzMTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/6441313?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rajarsheem",
"html_url": "https://github.com/rajarsheem",
"followers_url": "https://api.github.com/users... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1297?src=pr&el=h1) Report\n> Merging [#1297](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1297?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/0d1dad6d5323cf627cb8d7ddd428856ab8475f6b?... | 1,568 | 1,570 | 1,570 | NONE | null | sometimes we need to process multiple prompts from a file and generate multiple sequences. also, writing results to a file would be less verbose and fast. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1297/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1297/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1297",
"html_url": "https://github.com/huggingface/transformers/pull/1297",
"diff_url": "https://github.com/huggingface/transformers/pull/1297.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1297.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1296 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1296/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1296/comments | https://api.github.com/repos/huggingface/transformers/issues/1296/events | https://github.com/huggingface/transformers/pull/1296 | 495,847,974 | MDExOlB1bGxSZXF1ZXN0MzE5MzE2NDMx | 1,296 | Added ValueError for duplicates in list of added tokens | {
"login": "danai-antoniou",
"id": 32068609,
"node_id": "MDQ6VXNlcjMyMDY4NjA5",
"avatar_url": "https://avatars.githubusercontent.com/u/32068609?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/danai-antoniou",
"html_url": "https://github.com/danai-antoniou",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/1296?src=pr&el=h1) Report\n> Merging [#1296](https://codecov.io/gh/huggingface/transformers/pull/1296?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/391db836ab7ed2ca61c51a7cf1b135b6ab92be58?src=pr&el=desc) will **n... | 1,568 | 1,570 | 1,570 | CONTRIBUTOR | null | Very small addition to raise an error if the list of tokens passed to `add_tokens` contains duplicates. This otherwise raises cryptic errors down the line. Happy to update it to `Warning` if someone believes there's any reason for duplicates to be allowed here. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1296/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1296/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1296",
"html_url": "https://github.com/huggingface/transformers/pull/1296",
"diff_url": "https://github.com/huggingface/transformers/pull/1296.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1296.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1295 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1295/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1295/comments | https://api.github.com/repos/huggingface/transformers/issues/1295/events | https://github.com/huggingface/transformers/issues/1295 | 495,778,253 | MDU6SXNzdWU0OTU3NzgyNTM= | 1,295 | Where are BERT's pretrained Embeddings loaded? | {
"login": "BramVanroy",
"id": 2779410,
"node_id": "MDQ6VXNlcjI3Nzk0MTA=",
"avatar_url": "https://avatars.githubusercontent.com/u/2779410?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BramVanroy",
"html_url": "https://github.com/BramVanroy",
"followers_url": "https://api.github.com/users... | [] | closed | false | null | [] | [
"- \"token type embeddings\" are the BERT paper's segment embeddings \r\n- embeddings are inside the pretrained weights",
"Ah that makes sense. So there are no \"separate\" word2vec-style pretrained embedding models for the different types of embeddings which one could load with `nn.Embedding().from_pretrained`. ... | 1,568 | 1,584 | 1,568 | COLLABORATOR | null | I am trying to better understand the difference between the different types of embeddings that BERT uses (from the BERT paper: token, segment, position). For this purpose, I was hoping to put some print statement in the `pytorch_transformers` source code to see how the IDs change into vector representations for each ty... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1295/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1295/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1294 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1294/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1294/comments | https://api.github.com/repos/huggingface/transformers/issues/1294/events | https://github.com/huggingface/transformers/pull/1294 | 495,654,422 | MDExOlB1bGxSZXF1ZXN0MzE5MTU4NzI4 | 1,294 | Delete n_special reference in docstring | {
"login": "sshleifer",
"id": 6045025,
"node_id": "MDQ6VXNlcjYwNDUwMjU=",
"avatar_url": "https://avatars.githubusercontent.com/u/6045025?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sshleifer",
"html_url": "https://github.com/sshleifer",
"followers_url": "https://api.github.com/users/ss... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1294?src=pr&el=h1) Report\n> Merging [#1294](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1294?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/0d1dad6d5323cf627cb8d7ddd428856ab8475f6b?... | 1,568 | 1,569 | 1,569 | CONTRIBUTOR | null | I don't think the `n_special` param is used, even in `**kwargs`. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1294/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1294/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1294",
"html_url": "https://github.com/huggingface/transformers/pull/1294",
"diff_url": "https://github.com/huggingface/transformers/pull/1294.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1294.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1293 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1293/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1293/comments | https://api.github.com/repos/huggingface/transformers/issues/1293/events | https://github.com/huggingface/transformers/issues/1293 | 495,606,621 | MDU6SXNzdWU0OTU2MDY2MjE= | 1,293 | cannot import name 'XLNetForMultipleChoice' but python can import | {
"login": "yyHaker",
"id": 18585628,
"node_id": "MDQ6VXNlcjE4NTg1NjI4",
"avatar_url": "https://avatars.githubusercontent.com/u/18585628?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yyHaker",
"html_url": "https://github.com/yyHaker",
"followers_url": "https://api.github.com/users/yyHake... | [] | closed | false | null | [] | [
"I found maybe the current code is not consistent with the pip package pytorch_transformers, so when use the pip package it does't work, but when just run the code without the pip package, it can work, but you need change some path to make the code work correctly!",
"Hi, I believe this was fixed with @VictorSanh'... | 1,568 | 1,569 | 1,569 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert):
Language I am using the model on (English):
when I use the following command to run run_multiple_choice.py, like:
'''
python examples/run_multiple_choice.py --model_type bert --task_name race --model_name_or_path bert_large --do_train --do_... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1293/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1293/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1292 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1292/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1292/comments | https://api.github.com/repos/huggingface/transformers/issues/1292/events | https://github.com/huggingface/transformers/issues/1292 | 495,589,979 | MDU6SXNzdWU0OTU1ODk5Nzk= | 1,292 | Fine Tuning GPT2 on wikitext-103-raw | {
"login": "snaik2016",
"id": 18183245,
"node_id": "MDQ6VXNlcjE4MTgzMjQ1",
"avatar_url": "https://avatars.githubusercontent.com/u/18183245?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/snaik2016",
"html_url": "https://github.com/snaik2016",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"@snaik2016 I ran into the same issue and had to parallelize my code to make it faster. Also getting rid of the while loop and list splicing in the TextDataset class with a for loop made it much quicker.",
"Please check #1830 . I made some tuning on a training part. But I guess it'll still take many days for 1.8M... | 1,568 | 1,579 | 1,579 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
Running pytorch-transformers\examples\run_lm_finetuning.py. This is stuck at load_and_cache_examples step. I just see message like
WARNING - pytorch_transformers.tokenization_utils - This tokenizer does not make use of special tok... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1292/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1292/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1291 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1291/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1291/comments | https://api.github.com/repos/huggingface/transformers/issues/1291/events | https://github.com/huggingface/transformers/issues/1291 | 495,523,074 | MDU6SXNzdWU0OTU1MjMwNzQ= | 1,291 | traced_model | {
"login": "HongyanJiao",
"id": 44488820,
"node_id": "MDQ6VXNlcjQ0NDg4ODIw",
"avatar_url": "https://avatars.githubusercontent.com/u/44488820?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/HongyanJiao",
"html_url": "https://github.com/HongyanJiao",
"followers_url": "https://api.github.com/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Which model did you use？",
"> Which model did you use？\r\n\r\nxlnet",
"Hi! Could you show the inputs you use to trace your model?",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
... | 1,568 | 1,575 | 1,575 | NONE | null | when I ran: traced_model = torch.jit.trace(model, (input_ids,))
I got:
/home/jhy/py3.6/lib/python3.6/site-packages/torch/tensor.py:389: RuntimeWarning: Iterating over a tensor might cause the trace to be incorrect. Passing a tensor of different shape won't change the number of iterations executed (and might lead to e... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1291/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1291/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1290 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1290/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1290/comments | https://api.github.com/repos/huggingface/transformers/issues/1290/events | https://github.com/huggingface/transformers/issues/1290 | 495,477,472 | MDU6SXNzdWU0OTU0Nzc0NzI= | 1,290 | MemoryError on run_lm_finetuning.py | {
"login": "echan00",
"id": 6287299,
"node_id": "MDQ6VXNlcjYyODcyOTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/6287299?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/echan00",
"html_url": "https://github.com/echan00",
"followers_url": "https://api.github.com/users/echan00/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n",
"I also have the same problem...."
] | 1,568 | 1,576 | 1,574 | CONTRIBUTOR | null | Previous versions of finetune_on_pregenerated.py had a `--reduce_memory` parameter to keep memory requirements from going overboard, it seems like it is no longer available in the new run_lm_finetuning.py file? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1290/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1290/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1289 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1289/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1289/comments | https://api.github.com/repos/huggingface/transformers/issues/1289/events | https://github.com/huggingface/transformers/pull/1289 | 495,445,981 | MDExOlB1bGxSZXF1ZXN0MzE4OTkyNzI2 | 1,289 | Adding Adapters | {
"login": "zphang",
"id": 1668462,
"node_id": "MDQ6VXNlcjE2Njg0NjI=",
"avatar_url": "https://avatars.githubusercontent.com/u/1668462?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zphang",
"html_url": "https://github.com/zphang",
"followers_url": "https://api.github.com/users/zphang/foll... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1289?src=pr&el=h1) Report\n> Merging [#1289](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1289?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/0d1dad6d5323cf627cb8d7ddd428856ab8475f6b?... | 1,568 | 1,576 | 1,576 | CONTRIBUTOR | null | From: https://arxiv.org/pdf/1902.00751.pdf
Open to feedback!
* Implementing adapters requires a couple more hyperparameters that need to go into the BertConfig. Do let me know if there is an alternative to modifying the core Config object (maybe a subclass would work better?)
* If `use_adapter` is False, the ada... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1289/reactions",
"total_count": 3,
"+1": 3,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1289/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1289",
"html_url": "https://github.com/huggingface/transformers/pull/1289",
"diff_url": "https://github.com/huggingface/transformers/pull/1289.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1289.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1288 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1288/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1288/comments | https://api.github.com/repos/huggingface/transformers/issues/1288/events | https://github.com/huggingface/transformers/pull/1288 | 495,444,848 | MDExOlB1bGxSZXF1ZXN0MzE4OTkxNzgy | 1,288 | Typo with LM Fine tuning script | {
"login": "echan00",
"id": 6287299,
"node_id": "MDQ6VXNlcjYyODcyOTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/6287299?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/echan00",
"html_url": "https://github.com/echan00",
"followers_url": "https://api.github.com/users/echan00/... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/1288?src=pr&el=h1) Report\n> Merging [#1288](https://codecov.io/gh/huggingface/transformers/pull/1288?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/2dc8cb87341223e86220516951bb4ad84f880b4a?src=pr&el=desc) will **d... | 1,568 | 1,569 | 1,569 | CONTRIBUTOR | null | Typo with LM Fine tuning script | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1288/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1288/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1288",
"html_url": "https://github.com/huggingface/transformers/pull/1288",
"diff_url": "https://github.com/huggingface/transformers/pull/1288.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1288.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1287 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1287/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1287/comments | https://api.github.com/repos/huggingface/transformers/issues/1287/events | https://github.com/huggingface/transformers/issues/1287 | 495,435,291 | MDU6SXNzdWU0OTU0MzUyOTE= | 1,287 | TransfoXLLMHeadModel compatibility with pytorch 1.1.0 | {
"login": "sarahnlewis",
"id": 9419264,
"node_id": "MDQ6VXNlcjk0MTkyNjQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/9419264?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/sarahnlewis",
"html_url": "https://github.com/sarahnlewis",
"followers_url": "https://api.github.com/us... | [] | closed | false | null | [] | [
"Should be fixed on master and the new release (2.0)"
] | 1,568 | 1,569 | 1,569 | NONE | null | TransfoXLLMHeadModel._forward uses torch.Tensor.bool, which is not present in pytorch 1.1.0 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1287/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1287/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1286 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1286/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1286/comments | https://api.github.com/repos/huggingface/transformers/issues/1286/events | https://github.com/huggingface/transformers/issues/1286 | 495,337,957 | MDU6SXNzdWU0OTUzMzc5NTc= | 1,286 | Evaluation result.txt path suggestion | {
"login": "brian41005",
"id": 13401708,
"node_id": "MDQ6VXNlcjEzNDAxNzA4",
"avatar_url": "https://avatars.githubusercontent.com/u/13401708?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/brian41005",
"html_url": "https://github.com/brian41005",
"followers_url": "https://api.github.com/use... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Yes, why not, do you want to submit a PR for that?",
"> Yes, why not, do you want to submit a PR for that?\r\n\r\nThanks~\r\nBy the way, is there any code formatting requirement or a contribution docs for developers?\r\n",
"This issue has been automatically marked as stale because it has not had recent activit... | 1,568 | 1,582 | 1,575 | CONTRIBUTOR | null | ## 🚀 Feature
At pytorch-transformers/examples/**run_lm_finetuning**.py and **run_glue**.py
There is a line ```output_eval_file = os.path.join(eval_output_dir, "eval_results.txt")```
When setting evaluate_during_training **True**, `output_eval_file` will keep being overwritten.
I think `output_eval_file` can ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1286/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1286/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1285 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1285/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1285/comments | https://api.github.com/repos/huggingface/transformers/issues/1285/events | https://github.com/huggingface/transformers/issues/1285 | 495,307,780 | MDU6SXNzdWU0OTUzMDc3ODA= | 1,285 | GPT2 Tokenizer Decoding Adding Space | {
"login": "harkous",
"id": 5602332,
"node_id": "MDQ6VXNlcjU2MDIzMzI=",
"avatar_url": "https://avatars.githubusercontent.com/u/5602332?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/harkous",
"html_url": "https://github.com/harkous",
"followers_url": "https://api.github.com/users/harkous/... | [] | closed | false | null | [] | [
"Also getting this effect when using the reproduction code on my system.",
"It's not a bug. This is an artefact produced by BPE as explained here https://github.com/huggingface/pytorch-transformers/blob/d483cd8e469126bed081c59473bdf64ce74c8b36/pytorch_transformers/tokenization_gpt2.py#L106\r\n\r\nI think the solu... | 1,568 | 1,569 | 1,569 | CONTRIBUTOR | null | ## 🐛 Bug
The GPT-2 tokenizer's decoder now adds a space at the beginning of the string upon decoding.
(Potentially causing #1254)
Model I am using (Bert, XLNet....): GPT2
Language I am using the model on (English, Chinese....): English
The problem arise when using:
* [ ] the official example script... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1285/reactions",
"total_count": 3,
"+1": 3,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1285/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1284 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1284/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1284/comments | https://api.github.com/repos/huggingface/transformers/issues/1284/events | https://github.com/huggingface/transformers/pull/1284 | 495,230,014 | MDExOlB1bGxSZXF1ZXN0MzE4ODE1NzEy | 1,284 | Fix fp16 masking in PoolerEndLogits | {
"login": "slayton58",
"id": 4992598,
"node_id": "MDQ6VXNlcjQ5OTI1OTg=",
"avatar_url": "https://avatars.githubusercontent.com/u/4992598?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/slayton58",
"html_url": "https://github.com/slayton58",
"followers_url": "https://api.github.com/users/sl... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/1284?src=pr&el=h1) Report\n> Merging [#1284](https://codecov.io/gh/huggingface/transformers/pull/1284?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/2dc8cb87341223e86220516951bb4ad84f880b4a?src=pr&el=desc) will **d... | 1,568 | 1,569 | 1,569 | CONTRIBUTOR | null | Necessary to run xlnet squad fine-tuning with `--fp16 --fp16_opt_level="O2"`, otherwise loss is immediately `NaN` and fine-tuning cannot proceed.
Similar to #1249 | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1284/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1284/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1284",
"html_url": "https://github.com/huggingface/transformers/pull/1284",
"diff_url": "https://github.com/huggingface/transformers/pull/1284.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1284.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1283 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1283/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1283/comments | https://api.github.com/repos/huggingface/transformers/issues/1283/events | https://github.com/huggingface/transformers/issues/1283 | 495,089,778 | MDU6SXNzdWU0OTUwODk3Nzg= | 1,283 | Is training from scratch possible now? | {
"login": "Stamenov",
"id": 1288381,
"node_id": "MDQ6VXNlcjEyODgzODE=",
"avatar_url": "https://avatars.githubusercontent.com/u/1288381?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/Stamenov",
"html_url": "https://github.com/Stamenov",
"followers_url": "https://api.github.com/users/Stame... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"You can just instanciate the models without the `.from_pretraining()` like so:\r\n```python\r\nconfig = BertConfig(**optionally your favorite parameters**)\r\nmodel = BertForPretraining(config)\r\n```\r\n\r\nI added a flag to `run_lm_finetuning.py` that gets checked in the `main()`. Maybe this snipped helps (note,... | 1,568 | 1,581 | 1,574 | NONE | null | Do the models support training from scratch, together with original (paper) parameters? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1283/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1283/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1282 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1282/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1282/comments | https://api.github.com/repos/huggingface/transformers/issues/1282/events | https://github.com/huggingface/transformers/issues/1282 | 495,082,718 | MDU6SXNzdWU0OTUwODI3MTg= | 1,282 | start_position=0 in utils_squad.py when span is impossible | {
"login": "YeDeming",
"id": 12911231,
"node_id": "MDQ6VXNlcjEyOTExMjMx",
"avatar_url": "https://avatars.githubusercontent.com/u/12911231?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YeDeming",
"html_url": "https://github.com/YeDeming",
"followers_url": "https://api.github.com/users/YeD... | [] | closed | false | null | [] | [] | 1,568 | 1,568 | 1,568 | NONE | null | Hi,
https://github.com/huggingface/pytorch-transformers/blob/e768f2322abd2a2f60a3a6d64a6a94c2d957fe89/examples/utils_squad.py#L340-L351
when answer is out of span, start_position should be cls_index rather than 0 as L350
And in
https://github.com/huggingface/pytorch-transformers/blob/e768f2322abd2a2f60a3... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1282/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1282/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1280 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1280/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1280/comments | https://api.github.com/repos/huggingface/transformers/issues/1280/events | https://github.com/huggingface/transformers/issues/1280 | 494,984,636 | MDU6SXNzdWU0OTQ5ODQ2MzY= | 1,280 | FineTuning using single sentence document | {
"login": "tuhinjubcse",
"id": 3104771,
"node_id": "MDQ6VXNlcjMxMDQ3NzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/3104771?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tuhinjubcse",
"html_url": "https://github.com/tuhinjubcse",
"followers_url": "https://api.github.com/us... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi Tuhin, you can use `examples/run_lm_finetuning.py` now. The scripts in `examples/lm_finetuning` are deprecated (removed on master now).",
"Thomas i checked the wiki text 2 format and its confusing to me . Do we have to seperate documents by new lines ? My input file is a set of single sentence documents one p... | 1,568 | 1,577 | 1,577 | NONE | null | Hello
roBERTa is not using the next sentence prediction objective. I want to fine-tune the pre-trained model on an unlabelled corpus of domain-specific text (ULMFIT style intermediate pretraining).
The bottleneck is my examples are single short sentences, instead of a document with multiple sentences
The INPUT ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1280/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1280/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1279 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1279/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1279/comments | https://api.github.com/repos/huggingface/transformers/issues/1279/events | https://github.com/huggingface/transformers/issues/1279 | 494,962,869 | MDU6SXNzdWU0OTQ5NjI4Njk= | 1,279 | connection limit of pregenerate_training_data.py | {
"login": "ntubertchen",
"id": 7036778,
"node_id": "MDQ6VXNlcjcwMzY3Nzg=",
"avatar_url": "https://avatars.githubusercontent.com/u/7036778?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ntubertchen",
"html_url": "https://github.com/ntubertchen",
"followers_url": "https://api.github.com/us... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,574 | 1,574 | NONE | null | ## ❓ Questions & Help
Inside pregenerate_training_data.py, one can use multiprocessing to process each epoch in parallel.
However, since the communication between process is limited by pickle limit size. We can only transfer arguments less than 1GB and argument: docs are very likely to exceed this limit.
I sav... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1279/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1279/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1278 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1278/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1278/comments | https://api.github.com/repos/huggingface/transformers/issues/1278/events | https://github.com/huggingface/transformers/issues/1278 | 494,941,298 | MDU6SXNzdWU0OTQ5NDEyOTg= | 1,278 | 'Default process group is not initialized' Error | {
"login": "jasonmusespresso",
"id": 24786001,
"node_id": "MDQ6VXNlcjI0Nzg2MDAx",
"avatar_url": "https://avatars.githubusercontent.com/u/24786001?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/jasonmusespresso",
"html_url": "https://github.com/jasonmusespresso",
"followers_url": "https://... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"You can use `example/run_lm_finetuning` now, the scripts in the `example/lm_finetuning/` folder are deprecated (removed on master).",
"What kind input format is good for `example/run_lm_finetuning.py`?",
"Hi, there's an example using WikiText-2 in the [documentation](https://huggingface.co/pytorch-transformers... | 1,568 | 1,574 | 1,574 | NONE | null | I get this error when finetuning bert using code in `lm_finetuning/` folder, when I try to run it several GPUs.
```
Traceback (most recent call last):
File "finetune_on_pregenerated.py", line 330, in <module>
main()
File "finetune_on_pregenerated.py", line 323, in main
if n_gpu > 1 and torch.distrib... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1278/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1278/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1277 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1277/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1277/comments | https://api.github.com/repos/huggingface/transformers/issues/1277/events | https://github.com/huggingface/transformers/issues/1277 | 494,889,562 | MDU6SXNzdWU0OTQ4ODk1NjI= | 1,277 | No language embedding weights in pre-trained xlm models. | {
"login": "BalazsHoranyi",
"id": 5817677,
"node_id": "MDQ6VXNlcjU4MTc2Nzc=",
"avatar_url": "https://avatars.githubusercontent.com/u/5817677?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/BalazsHoranyi",
"html_url": "https://github.com/BalazsHoranyi",
"followers_url": "https://api.github.... | [] | closed | false | null | [] | [
"Indeed, I've checked and the 100 and 17 language models don't use language indices.\r\nJust supply `langs=None`.\r\nYou can see that in the official notebook from Facebook: https://github.com/facebookresearch/XLM/blob/master/generate-embeddings.ipynb",
"I see. I missed this issue here which explains it pretty we... | 1,568 | 1,569 | 1,569 | NONE | null | I'm trying to train a one-shot classification model using the given XLM pre-trained weights.
However, I noticed that for both `xlm-mlm-17-1280` and `xlm-mlm-100-1280`, That I kept receiving the warning `weights of XLMForSequenceClassification not initialized from pre-trained model: ['lang_embeddings.weight']`. I the... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1277/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1277/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1276 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1276/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1276/comments | https://api.github.com/repos/huggingface/transformers/issues/1276/events | https://github.com/huggingface/transformers/issues/1276 | 494,848,104 | MDU6SXNzdWU0OTQ4NDgxMDQ= | 1,276 | Write with Transformer: Please, add an autosave to browser cache! | {
"login": "varkarrus",
"id": 38511981,
"node_id": "MDQ6VXNlcjM4NTExOTgx",
"avatar_url": "https://avatars.githubusercontent.com/u/38511981?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/varkarrus",
"html_url": "https://github.com/varkarrus",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
},
{
"id": 1565794707,
"node_id": "MDU6TGFiZWwxNTY1Nzk0NzA3",
... | closed | false | null | [] | [
"Hi @varkarrus, thank you for your feature request. There is a \"save & publish\" button on the top right-hand side, which saves your document on a specific URL. Does this fit your needs?",
"Does it begin autosaving after you do that? If so, then probably!",
"Nope, it does not currently auto-save 🙁\r\n\r\nOne ... | 1,568 | 1,574 | 1,574 | NONE | null | ## 🚀 Feature
In Write with Transformer, the writing should be periodically saved to the browser cache, so that if the user accidentally refreshes the page, their work that they may have spent hours on won't be lost.
## Motivation
I just lost several hours worth of writing because I accidentally refreshed the ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1276/reactions",
"total_count": 3,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1276/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1275 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1275/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1275/comments | https://api.github.com/repos/huggingface/transformers/issues/1275/events | https://github.com/huggingface/transformers/pull/1275 | 494,650,464 | MDExOlB1bGxSZXF1ZXN0MzE4MzYxMzU5 | 1,275 | Implement fine-tuning BERT on CoNLL-2003 named entity recognition task | {
"login": "stecklin",
"id": 6171989,
"node_id": "MDQ6VXNlcjYxNzE5ODk=",
"avatar_url": "https://avatars.githubusercontent.com/u/6171989?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/stecklin",
"html_url": "https://github.com/stecklin",
"followers_url": "https://api.github.com/users/steck... | [] | closed | false | null | [] | [
"Thanks for adding this :+1: \r\n\r\nI've one suggestion for some improvement (rfc): can we make the `get_labels()` function a bit more configurable? E.g. reading the labels from a file `labels.txt` would be great, so I could use other datasets (e.g. GermEval, which has more labels) 🤔 What do you think 🤗",
"Th... | 1,568 | 1,593 | 1,571 | NONE | null | I added a script for fine-tuning BERT on the CoNLL-2003 named entity recognition task, as an example for token classification. This was requested in #1216.
I followed the structure of the run_glue example, and implemented the data processing in a way suitable for all transformer models (although currently token classi... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1275/reactions",
"total_count": 10,
"+1": 4,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 6,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1275/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1275",
"html_url": "https://github.com/huggingface/transformers/pull/1275",
"diff_url": "https://github.com/huggingface/transformers/pull/1275.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1275.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1274 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1274/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1274/comments | https://api.github.com/repos/huggingface/transformers/issues/1274/events | https://github.com/huggingface/transformers/pull/1274 | 494,650,320 | MDExOlB1bGxSZXF1ZXN0MzE4MzYxMjQy | 1,274 | Fixes #1263, add tokenization_with_offsets, gets tokens with offsets in the original text | {
"login": "michaelrglass",
"id": 35044941,
"node_id": "MDQ6VXNlcjM1MDQ0OTQx",
"avatar_url": "https://avatars.githubusercontent.com/u/35044941?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/michaelrglass",
"html_url": "https://github.com/michaelrglass",
"followers_url": "https://api.githu... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"I think this is an interesting addition and I like the way the PR is structured in general.\r\n\r\nBefore I dive in, could you lay down the status of the PR in terms of supported models, python version (we will still keep python 2 support for now), know issues and TO-DOs?",
"The fully supported models are BERT, ... | 1,568 | 1,583 | 1,583 | CONTRIBUTOR | null | This is similar to the utils_squad approach to getting offsets for the tokens but can also be used in other places where the tokens should have a correspondence to the original text to fix #1263.
| {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1274/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1274/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1274",
"html_url": "https://github.com/huggingface/transformers/pull/1274",
"diff_url": "https://github.com/huggingface/transformers/pull/1274.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1274.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1273 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1273/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1273/comments | https://api.github.com/repos/huggingface/transformers/issues/1273/events | https://github.com/huggingface/transformers/issues/1273 | 494,409,710 | MDU6SXNzdWU0OTQ0MDk3MTA= | 1,273 | ModuleNotFoundError: No module named 'pytorch_transformers.modeling' using convert_pytorch_checkpoint_to_tf.py | {
"login": "yangyaofei",
"id": 7934098,
"node_id": "MDQ6VXNlcjc5MzQwOTg=",
"avatar_url": "https://avatars.githubusercontent.com/u/7934098?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yangyaofei",
"html_url": "https://github.com/yangyaofei",
"followers_url": "https://api.github.com/users... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,574 | 1,574 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....):
BERT
Language I am using the model on (English, Chinese....):
Chinese
The problem arise when using:
* [✔️ ] the official example scripts: (give details)
run convert_pytorch_checkpoint_to_tf.py to generate the tf check point
## T... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1273/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1273/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1272 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1272/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1272/comments | https://api.github.com/repos/huggingface/transformers/issues/1272/events | https://github.com/huggingface/transformers/issues/1272 | 494,279,084 | MDU6SXNzdWU0OTQyNzkwODQ= | 1,272 | How long does it take? (BERT Model Finetuning using Masked ML objective) | {
"login": "echan00",
"id": 6287299,
"node_id": "MDQ6VXNlcjYyODcyOTk=",
"avatar_url": "https://avatars.githubusercontent.com/u/6287299?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/echan00",
"html_url": "https://github.com/echan00",
"followers_url": "https://api.github.com/users/echan00/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,574 | 1,574 | CONTRIBUTOR | null | I am about to finetune a multilingual BERT model using English and Chinese text from the legal domain.
My corpus is around 27GB, how long should I expect to train 3 epochs (default parameters) using a Google TPU? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1272/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1272/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1271 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1271/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1271/comments | https://api.github.com/repos/huggingface/transformers/issues/1271/events | https://github.com/huggingface/transformers/issues/1271 | 494,194,615 | MDU6SXNzdWU0OTQxOTQ2MTU= | 1,271 | get NaN loss when I run the example code run_squad.py | {
"login": "kugwzk",
"id": 15382517,
"node_id": "MDQ6VXNlcjE1MzgyNTE3",
"avatar_url": "https://avatars.githubusercontent.com/u/15382517?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/kugwzk",
"html_url": "https://github.com/kugwzk",
"followers_url": "https://api.github.com/users/kugwzk/fo... | [] | closed | false | null | [] | [
"Maybe the learning rate is too high?",
"I change a GPU node and this situation doesn't appear. I will change the learning rate and see results. Thanks a lot."
] | 1,568 | 1,568 | 1,568 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
I use the example run_squad.py code, and use the Readme's hyper-parameters, but I got nan loss when I trained a few batches. And I use the `autograd.detect_anomaly()` want to catch that.
The more information is below:
> File "/user... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1271/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1271/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1270 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1270/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1270/comments | https://api.github.com/repos/huggingface/transformers/issues/1270/events | https://github.com/huggingface/transformers/issues/1270 | 493,993,318 | MDU6SXNzdWU0OTM5OTMzMTg= | 1,270 | BERT returns different embedding for same sentence | {
"login": "rshah1990",
"id": 37735152,
"node_id": "MDQ6VXNlcjM3NzM1MTUy",
"avatar_url": "https://avatars.githubusercontent.com/u/37735152?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rshah1990",
"html_url": "https://github.com/rshah1990",
"followers_url": "https://api.github.com/users/... | [] | closed | false | null | [] | [
"Are you initializing from a pretrained model? If no, than this is normal behaviour: your weights are randomly initialized. If yes, make sure your model is in evaluation mode (```model.eval()```), this disables dropout and other random modules.",
"Thank you for quick response @srslynow . how to initialize weights... | 1,568 | 1,568 | 1,568 | NONE | null | I am using pre-trained BERT for creating features, for same sentence it produces different result in two different runs. Do we have to set some random state to produce consistent result? I am using pytorch-transformers for reading pre-trained model.
| {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1270/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1270/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1269 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1269/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1269/comments | https://api.github.com/repos/huggingface/transformers/issues/1269/events | https://github.com/huggingface/transformers/issues/1269 | 493,816,341 | MDU6SXNzdWU0OTM4MTYzNDE= | 1,269 | could you add an option to transfer variables from float32 to float16 in GPT2 model to reduce model size and accelerate the inference speed | {
"login": "carter54",
"id": 26741594,
"node_id": "MDQ6VXNlcjI2NzQxNTk0",
"avatar_url": "https://avatars.githubusercontent.com/u/26741594?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/carter54",
"html_url": "https://github.com/carter54",
"followers_url": "https://api.github.com/users/car... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"You should give a look at NVIDIA's apex library and PyTorch `model.half()` method.",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n",
"@thomwolf Do you know if the gpt2 model needs t... | 1,568 | 1,583 | 1,574 | NONE | null | ## 🚀 Feature
could you add an option to transfer variables from float32 to float16 in GPT2 model
## Motivation
reduce model size and accelerate the inference speed
## Additional context
<!-- Add any other context or screenshots about the feature request here. --> | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1269/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1269/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1268 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1268/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1268/comments | https://api.github.com/repos/huggingface/transformers/issues/1268/events | https://github.com/huggingface/transformers/issues/1268 | 493,765,687 | MDU6SXNzdWU0OTM3NjU2ODc= | 1,268 | How to use pytorch-transformers for transfer learning? | {
"login": "vasilynikita",
"id": 53129810,
"node_id": "MDQ6VXNlcjUzMTI5ODEw",
"avatar_url": "https://avatars.githubusercontent.com/u/53129810?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/vasilynikita",
"html_url": "https://github.com/vasilynikita",
"followers_url": "https://api.github.c... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hey, so not exactly a direct answer to your question, but bert outright doesn't do amazing on sentence similarity. This repo here should help with your question and their paper does a great job at explaining how their method works https://github.com/UKPLab/sentence-transformers. I think you will find better result... | 1,568 | 1,574 | 1,574 | NONE | null | ## ❓ Questions & Help
```python
pretrained_weights='bert-base-uncased'
tokenizer = tokenizer_class.from_pretrained(pretrained_weights)
model = model_class.from_pretrained(pretrained_weights)
features = []
input_ids = torch.tensor([self.tokenizer.encode(phrase, add_special_tokens=True)])
with torch.no_grad():... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1268/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1268/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1267 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1267/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1267/comments | https://api.github.com/repos/huggingface/transformers/issues/1267/events | https://github.com/huggingface/transformers/issues/1267 | 493,742,689 | MDU6SXNzdWU0OTM3NDI2ODk= | 1,267 | Accuracy not increasing with BERT Large model | {
"login": "ghost",
"id": 10137,
"node_id": "MDQ6VXNlcjEwMTM3",
"avatar_url": "https://avatars.githubusercontent.com/u/10137?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ghost",
"html_url": "https://github.com/ghost",
"followers_url": "https://api.github.com/users/ghost/followers",
"f... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi, I experienced this also in several experiments. BERT large is extremely sensitive to the random seed. Try some other seeds and you will likely get a performance at least on oar with the base model.\r\n\r\nI haven't it studied further why the large model is so sensitive to the random seed, but it appears that t... | 1,568 | 1,576 | 1,576 | NONE | null | I experimented with `BERT_base_cased` and `BERT_large_cased` model for multi class text classification. With BERT_base_cased, I got satisfactory results. When I tried with BERT_large_cased model, the accuracy is same for all the epochs
```
Epoch: 01 | Epoch Time: 0m 57s
*******train_loss,train_acc,valid_loss,vali... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1267/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1267/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1266 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1266/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1266/comments | https://api.github.com/repos/huggingface/transformers/issues/1266/events | https://github.com/huggingface/transformers/issues/1266 | 493,741,184 | MDU6SXNzdWU0OTM3NDExODQ= | 1,266 | Fine-tune distilbert-base-uncased under run_glue | {
"login": "YosiMass",
"id": 6850963,
"node_id": "MDQ6VXNlcjY4NTA5NjM=",
"avatar_url": "https://avatars.githubusercontent.com/u/6850963?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YosiMass",
"html_url": "https://github.com/YosiMass",
"followers_url": "https://api.github.com/users/YosiM... | [] | closed | false | null | [] | [
"Hello @YosiMass,\r\nThe simplest/more direct way to do transfer learning is indeed the 3rd solution.\r\nIf you use `run_glue.py`, the modification you made is correct. You also have to be careful since DistilBERT doesn't take `token_type_embeddings` as input --> [here](https://github.com/huggingface/pytorch-transf... | 1,568 | 1,569 | 1,569 | NONE | null | ## ❓ Questions & Help
I am a bit confused on how to fine-tune distilbert. I see three options
1. Fine-tune bert for the task and then use distillation.distiller
2. Fine-tune bert for the task and then use distillation.train
3. Fine-tune distilbert-base-uncased directly for the task using run_glue.py
I trie... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1266/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1266/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1265 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1265/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1265/comments | https://api.github.com/repos/huggingface/transformers/issues/1265/events | https://github.com/huggingface/transformers/issues/1265 | 493,704,912 | MDU6SXNzdWU0OTM3MDQ5MTI= | 1,265 | different results shown each time when I run the example code for BertForMultipleChoice | {
"login": "yuchuang1979",
"id": 38869115,
"node_id": "MDQ6VXNlcjM4ODY5MTE1",
"avatar_url": "https://avatars.githubusercontent.com/u/38869115?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/yuchuang1979",
"html_url": "https://github.com/yuchuang1979",
"followers_url": "https://api.github.c... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Yes!\r\n\r\nYou have to fine-tune BertForMultipleChoice to be able to use it.\r\n\r\n",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n",
"Does Bert add extra hidden layers that are r... | 1,568 | 1,692 | 1,574 | NONE | null | When I ran the following example provided for BertForMultipleChoice in the documentation, I've got different results each time when I run it. Does it mean that BertForMultipleChoice is only provided to fine-tune the BERT model with RocStories/SWAG like datasets, and no pretrained models (after the fine-tuning) are prov... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1265/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1265/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1264 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1264/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1264/comments | https://api.github.com/repos/huggingface/transformers/issues/1264/events | https://github.com/huggingface/transformers/issues/1264 | 493,637,149 | MDU6SXNzdWU0OTM2MzcxNDk= | 1,264 | Error running openai-gpt on ROCstories | {
"login": "roholazandie",
"id": 7584674,
"node_id": "MDQ6VXNlcjc1ODQ2NzQ=",
"avatar_url": "https://avatars.githubusercontent.com/u/7584674?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/roholazandie",
"html_url": "https://github.com/roholazandie",
"followers_url": "https://api.github.com... | [] | closed | false | null | [] | [
"Ok should be fixed now on master with e768f23"
] | 1,568 | 1,568 | 1,568 | NONE | null | ## 🐛 Bug
<!-- Important information -->
The model I am using: OpenAIGPT
The language I am using the model on: English
The problem arises when using:
* [ ] the official example scripts: When I try to run examples/single_model_script/run_openai_gpt.py I get this error:
```
Traceback (most recent call la... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1264/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1264/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1263 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1263/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1263/comments | https://api.github.com/repos/huggingface/transformers/issues/1263/events | https://github.com/huggingface/transformers/issues/1263 | 493,552,222 | MDU6SXNzdWU0OTM1NTIyMjI= | 1,263 | Offsets in original text from tokenizers | {
"login": "michaelrglass",
"id": 35044941,
"node_id": "MDQ6VXNlcjM1MDQ0OTQx",
"avatar_url": "https://avatars.githubusercontent.com/u/35044941?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/michaelrglass",
"html_url": "https://github.com/michaelrglass",
"followers_url": "https://api.githu... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"I am glad to see someone is working on this and really appreciate your work.\r\nCurrently I'm using LCS the original xlnet also uses to align the token and raw input to extract the answer highlighting.\r\nThis is really painful as it's slow and may fail at some corner case.\r\nCan't wait to see your pulled feature... | 1,568 | 1,574 | 1,574 | CONTRIBUTOR | null | ## 🚀 Feature
A new method for tokenizers: tokenize_with_offsets. In addition to returning the tokens, it returns the spans in the original text that the tokens correspond to.
After
tokens, offsets = tokenizer.tokenize_with_offsets(text)
then
tokens[i] maps to text[offsets[i, 0]:offsets[i, 1]]
## Motivation
... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1263/reactions",
"total_count": 3,
"+1": 3,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1263/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1262 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1262/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1262/comments | https://api.github.com/repos/huggingface/transformers/issues/1262/events | https://github.com/huggingface/transformers/issues/1262 | 493,385,577 | MDU6SXNzdWU0OTMzODU1Nzc= | 1,262 | run_generation.py 'encode' error for gpt2 and xlnet | {
"login": "GenTxt",
"id": 22547261,
"node_id": "MDQ6VXNlcjIyNTQ3MjYx",
"avatar_url": "https://avatars.githubusercontent.com/u/22547261?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/GenTxt",
"html_url": "https://github.com/GenTxt",
"followers_url": "https://api.github.com/users/GenTxt/fo... | [] | closed | false | null | [] | [
"Closing own non-issue. Needed to download missing files listed in the four applicable tokenizer scripts. Working 100%. On to fine-tuning."
] | 1,568 | 1,568 | 1,568 | NONE | null | Hello:
Been using nshepperd's tf repo and various excellent forks for fine-tuning and inference without issue.
Wanted to check out py-torch transformers and compare. First test is simple conditional sampling from the pytorch models:
python3 run_generation.py --model_type=xlnet --length=20 --model_name_or_path=... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1262/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1262/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1261 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1261/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1261/comments | https://api.github.com/repos/huggingface/transformers/issues/1261/events | https://github.com/huggingface/transformers/issues/1261 | 493,332,338 | MDU6SXNzdWU0OTMzMzIzMzg= | 1,261 | SequenceSummary / quenstion regarding summary types | {
"login": "cherepanovic",
"id": 10064548,
"node_id": "MDQ6VXNlcjEwMDY0NTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/10064548?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cherepanovic",
"html_url": "https://github.com/cherepanovic",
"followers_url": "https://api.github.c... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"That's how it's done in the respective original implementations.",
"it is very interesting. All other tokens will not be considered. \r\n\r\nDo you know whether other architectures has been tried out? Initially, when I had taken the transformer-model I created another output architecture. It is very interesting,... | 1,568 | 1,574 | 1,574 | NONE | null | in the class SequenceSummary(nn.Module) which is a part of {BERT, XLNet}ForSequenceClassification:
https://github.com/huggingface/pytorch-transformers/blob/32e1332acf6fd1ad372b81c296d43be441d3b0b1/pytorch_transformers/modeling_utils.py#L643-L644
in case of XLNet we can see that the last token will be taken, for B... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1261/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1261/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1260 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1260/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1260/comments | https://api.github.com/repos/huggingface/transformers/issues/1260/events | https://github.com/huggingface/transformers/issues/1260 | 493,332,321 | MDU6SXNzdWU0OTMzMzIzMjE= | 1,260 | XLNet tokenizer returns empty list instead of string for some indexes | {
"login": "andrey999333",
"id": 29929303,
"node_id": "MDQ6VXNlcjI5OTI5MzAz",
"avatar_url": "https://avatars.githubusercontent.com/u/29929303?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/andrey999333",
"html_url": "https://github.com/andrey999333",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | [
"Hi,\r\n`tokenizer.decode()` expect a sequence of ids as indicated in the doc/docstring: https://huggingface.co/pytorch-transformers/main_classes/tokenizer.html#pytorch_transformers.PreTrainedTokenizer.decode",
"well, as far as i see, the tokenizer can accept both - sequence or single index. For example:\r\n`toke... | 1,568 | 1,569 | 1,569 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....):
XLNet
Language I am using the model on (English, Chinese....):
English
The problem arise when using:
* [ ] my own modified scripts: (give details)
The tasks I am working on is:
* [ ] my own task or dataset: (give details... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1260/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1260/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1259 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1259/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1259/comments | https://api.github.com/repos/huggingface/transformers/issues/1259/events | https://github.com/huggingface/transformers/issues/1259 | 493,059,405 | MDU6SXNzdWU0OTMwNTk0MDU= | 1,259 | Cannot install the library | {
"login": "YimingSun60",
"id": 33106916,
"node_id": "MDQ6VXNlcjMzMTA2OTE2",
"avatar_url": "https://avatars.githubusercontent.com/u/33106916?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/YimingSun60",
"html_url": "https://github.com/YimingSun60",
"followers_url": "https://api.github.com/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,574 | 1,574 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
Traceback (most recent call last):
File "/usr/lib/python2.7/dist-packages/pip/basecommand.py", line 215, in main
status = self.run(options, args)
File "/usr/lib/python2.7/dist-packages/pip/commands/install.py", line 353, in... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1259/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1259/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1258 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1258/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1258/comments | https://api.github.com/repos/huggingface/transformers/issues/1258/events | https://github.com/huggingface/transformers/pull/1258 | 493,055,972 | MDExOlB1bGxSZXF1ZXN0MzE3MTEwODAx | 1,258 | fix padding_idx of RoBERTa model | {
"login": "ikuyamada",
"id": 426342,
"node_id": "MDQ6VXNlcjQyNjM0Mg==",
"avatar_url": "https://avatars.githubusercontent.com/u/426342?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ikuyamada",
"html_url": "https://github.com/ikuyamada",
"followers_url": "https://api.github.com/users/ikuy... | [] | closed | false | null | [] | [
"LGTM but let's have @julien-c or @LysandreJik confirm",
"lgtm too",
"It would be appreciated if you review this PR! @LysandreJik ",
"@ikuyamada Out of curiosity, in which cases did you need to specify this `padding_idx`? It shouldn't have any impact on the inference so are you training a model from scratch? ... | 1,568 | 1,569 | 1,569 | CONTRIBUTOR | null | The padding index of the pretrained RoBERTa model is 1, and 0 is assigned to `<s>` token. The padding index of the current RoBERTa model is set to be 0, therefore `<s>` is treated as padding.
This PR aims to fix this problem. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1258/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1258/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1258",
"html_url": "https://github.com/huggingface/transformers/pull/1258",
"diff_url": "https://github.com/huggingface/transformers/pull/1258.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1258.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1257 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1257/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1257/comments | https://api.github.com/repos/huggingface/transformers/issues/1257/events | https://github.com/huggingface/transformers/issues/1257 | 492,903,465 | MDU6SXNzdWU0OTI5MDM0NjU= | 1,257 | Training time increased from 45 min per epoch to 6 hours per epoch in colab | {
"login": "bvy007",
"id": 6167208,
"node_id": "MDQ6VXNlcjYxNjcyMDg=",
"avatar_url": "https://avatars.githubusercontent.com/u/6167208?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/bvy007",
"html_url": "https://github.com/bvy007",
"followers_url": "https://api.github.com/users/bvy007/foll... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi, \r\n\r\nFound a solution to my problem to some extent. I had cloned the latest apex repo and testing out with PyTorch-pretrained-Bert which is causing the problem to take more time for execution. I took the older apex repo and tested the code and working as earlier (45 min per epoch)\r\n\r\nBut when I was usin... | 1,568 | 1,574 | 1,574 | NONE | null | ## 📚 Migration
<!-- Important information -->
Model I am using (Bert): PyTorch-pretrained-Bert and BERT in pytorch-transformers
Language I am using the model on : English
The problem arise when using:
* the official example scripts: run_squad.py in pytorch-transformers
The tasks I am working on is:
*... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1257/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1257/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1256 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1256/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1256/comments | https://api.github.com/repos/huggingface/transformers/issues/1256/events | https://github.com/huggingface/transformers/issues/1256 | 492,850,584 | MDU6SXNzdWU0OTI4NTA1ODQ= | 1,256 | Could you please implement a Adafactor optimizer? :) | {
"login": "christophschuhmann",
"id": 22318853,
"node_id": "MDQ6VXNlcjIyMzE4ODUz",
"avatar_url": "https://avatars.githubusercontent.com/u/22318853?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/christophschuhmann",
"html_url": "https://github.com/christophschuhmann",
"followers_url": "ht... | [] | closed | false | null | [] | [
"What didn't work for you with the fairseq implementation?\r\n\r\nIt seems pretty self-contained: https://github.com/pytorch/fairseq/blob/master/fairseq/optim/adafactor.py#L65-L213",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity... | 1,568 | 1,598 | 1,598 | NONE | null | ## 🚀 Feature
Could you please implement a Adafactor optimizer? :)
( https://arxiv.org/abs/1804.04235 )
## Motivation
In contrast to Adam it requires much less GPU memory.
I tried to use the FairSeq implementation for the pytorch-transformers, but I'm no expert and I couldn't get it done.
Could you ple... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1256/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1256/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1255 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1255/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1255/comments | https://api.github.com/repos/huggingface/transformers/issues/1255/events | https://github.com/huggingface/transformers/issues/1255 | 492,803,150 | MDU6SXNzdWU0OTI4MDMxNTA= | 1,255 | examples/lm_finetuning/simple_lm_finetuning.py crashes with cublas runtime error | {
"login": "fredriko",
"id": 5771366,
"node_id": "MDQ6VXNlcjU3NzEzNjY=",
"avatar_url": "https://avatars.githubusercontent.com/u/5771366?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/fredriko",
"html_url": "https://github.com/fredriko",
"followers_url": "https://api.github.com/users/fredr... | [] | closed | false | null | [] | [
"Additional info: simple_lm_finetuning.py works with pytorch-transformers version 1.1.0, but not with version 1.2.0.",
"Maybe related to the change in order of parameters to BertModel's forward method ?\r\nSee #1246",
"Hi, the lm finetuning examples are now replaced by `examples/run_lm_finetuning`"
] | 1,568 | 1,568 | 1,568 | NONE | null | ## Possible bug: simple_lm_finetuning.py crashes with cublas runtime error🐛 Bug
<!-- Important information -->
### TL;DR
I'm trying to finetune the existing English **bert-base-uncased** model according to the examples in `examples/lm_finetuning/README.md` on IMDB data, but fail. The `simple_lm_finetuning.py` scr... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1255/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1255/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1254 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1254/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1254/comments | https://api.github.com/repos/huggingface/transformers/issues/1254/events | https://github.com/huggingface/transformers/issues/1254 | 492,801,100 | MDU6SXNzdWU0OTI4MDExMDA= | 1,254 | Write With Transformer adding spaces? | {
"login": "zacharymacleod",
"id": 6412653,
"node_id": "MDQ6VXNlcjY0MTI2NTM=",
"avatar_url": "https://avatars.githubusercontent.com/u/6412653?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zacharymacleod",
"html_url": "https://github.com/zacharymacleod",
"followers_url": "https://api.gith... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
},
{
"id": 1565794707,
"node_id": "MDU6TGFiZWwxNTY1Nzk0NzA3",
... | closed | false | null | [] | [
"Commenting to say I have also noticed this. Also I would assume the small amount of tokens being generated per autocomplete lately is because of compute concerns, not time concerns. It is a bit limiting.",
"Also having this problem!",
"This issue has been automatically marked as stale because it has not had re... | 1,568 | 1,574 | 1,574 | NONE | null | This issue didn't happen before, but now whenever you use the autocomplete it always adds a space to the beginning, even when a space is not needed. I.E: when adding a comma / period to the end of a sentence, when starting a new line, or (most egregiously) when finishing a word that was only partially written.
As an... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1254/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1254/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1253 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1253/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1253/comments | https://api.github.com/repos/huggingface/transformers/issues/1253/events | https://github.com/huggingface/transformers/issues/1253 | 492,558,944 | MDU6SXNzdWU0OTI1NTg5NDQ= | 1,253 | Running XLNet on Squad | {
"login": "LeonCrashCode",
"id": 5652525,
"node_id": "MDQ6VXNlcjU2NTI1MjU=",
"avatar_url": "https://avatars.githubusercontent.com/u/5652525?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LeonCrashCode",
"html_url": "https://github.com/LeonCrashCode",
"followers_url": "https://api.github.... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | {
"login": "VictorSanh",
"id": 16107619,
"node_id": "MDQ6VXNlcjE2MTA3NjE5",
"avatar_url": "https://avatars.githubusercontent.com/u/16107619?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VictorSanh",
"html_url": "https://github.com/VictorSanh",
"followers_url": "https://api.github.com/use... | [
{
"login": "VictorSanh",
"id": 16107619,
"node_id": "MDQ6VXNlcjE2MTA3NjE5",
"avatar_url": "https://avatars.githubusercontent.com/u/16107619?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/VictorSanh",
"html_url": "https://github.com/VictorSanh",
"followers_url": "https:... | [
"Seems like the run_squad script is in bad shape now. It just doesn't work.",
"same question.. also running this script with XLNet on Squad, is ~10 F1 points below BERT-Large-WWM. The difference in preprocessing as pointed out above could be one of the reasons.",
"This issue has been automatically marked as sta... | 1,568 | 1,574 | 1,574 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
This is the padding problem. In GLUE codes in the examples, the padding for XLNet is on the left of the input, but in Squad codes, the padding is on right. I was wondering which one is correct.
Also, the inputs of `convert_example... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1253/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1253/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1252 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1252/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1252/comments | https://api.github.com/repos/huggingface/transformers/issues/1252/events | https://github.com/huggingface/transformers/pull/1252 | 492,340,210 | MDExOlB1bGxSZXF1ZXN0MzE2NTMyMTk3 | 1,252 | Max encoding length + corresponding tests | {
"login": "LysandreJik",
"id": 30755778,
"node_id": "MDQ6VXNlcjMwNzU1Nzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/30755778?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LysandreJik",
"html_url": "https://github.com/LysandreJik",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1252?src=pr&el=h1) Report\n> Merging [#1252](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1252?src=pr&el=desc) into [glue-example](https://codecov.io/gh/huggingface/pytorch-transformers/commit/5583711822f79d8b3b7e7ba2560748cc0cf... | 1,568 | 1,576 | 1,568 | MEMBER | null | The encoding function eases the encoding of sequences across tokenizers. The addition of the `head_mask` return further removes the pressure on the user to manually check the added special tokens.
There is currently no easy method to truncate the encoded sequences while keeping the special tokens intact. This PR a... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1252/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1252/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1252",
"html_url": "https://github.com/huggingface/transformers/pull/1252",
"diff_url": "https://github.com/huggingface/transformers/pull/1252.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1252.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1251 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1251/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1251/comments | https://api.github.com/repos/huggingface/transformers/issues/1251/events | https://github.com/huggingface/transformers/issues/1251 | 492,278,744 | MDU6SXNzdWU0OTIyNzg3NDQ= | 1,251 | Why you need DistilBertModel class? | {
"login": "tomohideshibata",
"id": 16042472,
"node_id": "MDQ6VXNlcjE2MDQyNDcy",
"avatar_url": "https://avatars.githubusercontent.com/u/16042472?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tomohideshibata",
"html_url": "https://github.com/tomohideshibata",
"followers_url": "https://api... | [] | closed | false | null | [] | [
"This is a smaller model compared to the original, and is thus better suitable for usage on embedded devices / devices without large gpu's. Their blog posts explains this: https://medium.com/huggingface/distilbert-8cf3380435b5.",
"Hello @tomohideshibata,\r\nTrue, we could use the same code base for BERT and Disti... | 1,568 | 1,568 | 1,568 | CONTRIBUTOR | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
You have `DistilBertModel`, `DistilBertForSequenceClassification`, etc. in `modeling_distilbert.py`. Why you need these classes? How about using `BertModel`, `BertForSequenceClassification`, etc.?
I found the weight names are diffe... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1251/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1251/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1250 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1250/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1250/comments | https://api.github.com/repos/huggingface/transformers/issues/1250/events | https://github.com/huggingface/transformers/issues/1250 | 492,082,448 | MDU6SXNzdWU0OTIwODI0NDg= | 1,250 | R-BERT implementation | {
"login": "RichJackson",
"id": 7306627,
"node_id": "MDQ6VXNlcjczMDY2Mjc=",
"avatar_url": "https://avatars.githubusercontent.com/u/7306627?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/RichJackson",
"html_url": "https://github.com/RichJackson",
"followers_url": "https://api.github.com/us... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This is mostly a new head for Bert, right?\r\nIf so, yes I think it could be a nice addition.\r\nIs the tokenizer different as well?",
"That's correct. The new head is [here](https://github.com/azdatascience/pytorch-transformers/blob/rbert/pytorch_transformers/modeling_bert.py#L832) (Think that's in the right pl... | 1,568 | 1,574 | 1,574 | NONE | null | ## 🚀 Feature
An implementation of the R-BERT architecture for relationship classification
## Motivation
Hi @Huggingface. A recent paper describes an architecture for relationship classification called [R-BERT](https://arxiv.org/pdf/1905.08284.pdf), which claims SOTA performance on the Semeval 2010 Task 8 chal... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1250/reactions",
"total_count": 4,
"+1": 3,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 1,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1250/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1249 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1249/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1249/comments | https://api.github.com/repos/huggingface/transformers/issues/1249/events | https://github.com/huggingface/transformers/pull/1249 | 492,073,765 | MDExOlB1bGxSZXF1ZXN0MzE2MzE3MjE0 | 1,249 | fixed: hard coding for max and min number will out of range in fp16, which will cause nan. | {
"login": "ziliwang",
"id": 13744942,
"node_id": "MDQ6VXNlcjEzNzQ0OTQy",
"avatar_url": "https://avatars.githubusercontent.com/u/13744942?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ziliwang",
"html_url": "https://github.com/ziliwang",
"followers_url": "https://api.github.com/users/zil... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1249?src=pr&el=h1) Report\n> Merging [#1249](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1249?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/364920e216c16d73c782a61a4cf6652e541fbe18?... | 1,568 | 1,568 | 1,568 | CONTRIBUTOR | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1249/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1249/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1249",
"html_url": "https://github.com/huggingface/transformers/pull/1249",
"diff_url": "https://github.com/huggingface/transformers/pull/1249.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1249.patch",
"merged_at": 1... | |
https://api.github.com/repos/huggingface/transformers/issues/1248 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1248/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1248/comments | https://api.github.com/repos/huggingface/transformers/issues/1248/events | https://github.com/huggingface/transformers/issues/1248 | 492,060,366 | MDU6SXNzdWU0OTIwNjAzNjY= | 1,248 | model_type for gpt | {
"login": "tuhinjubcse",
"id": 3104771,
"node_id": "MDQ6VXNlcjMxMDQ3NzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/3104771?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tuhinjubcse",
"html_url": "https://github.com/tuhinjubcse",
"followers_url": "https://api.github.com/us... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi Tuhin, adding GPT/GPT2 is a tiny bit more complicated since these models need to have special tokens added to their vocabulary prior to fine-tuning (just a few more lines of code though).\r\n\r\nDo you want to try to make a PR with these other models?",
"Okay I will add bert as the model type i was just not s... | 1,568 | 1,574 | 1,574 | NONE | null | ## ❓ Questions & Help
To run openai-gpt model in run_glue.py what should we provide as model_type? The model classes contains
MODEL_CLASSES = {
'bert'
'xlnet'
'xlm'
'roberta'
} | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1248/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1248/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1247 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1247/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1247/comments | https://api.github.com/repos/huggingface/transformers/issues/1247/events | https://github.com/huggingface/transformers/issues/1247 | 491,969,538 | MDU6SXNzdWU0OTE5Njk1Mzg= | 1,247 | KnowBert | {
"login": "rishibommasani",
"id": 47439426,
"node_id": "MDQ6VXNlcjQ3NDM5NDI2",
"avatar_url": "https://avatars.githubusercontent.com/u/47439426?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rishibommasani",
"html_url": "https://github.com/rishibommasani",
"followers_url": "https://api.gi... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi,\r\nWe only add models when there are pretrained weights released.\r\nThis doesn't seem to be the case for KnowBert, or maybe I missed them?",
"Yes, looking into if they are releasing pretrained weights, I incorrectly assumed they were. ",
"Wow! Really looking forward to this.\r\nI really feel that models t... | 1,568 | 1,691 | 1,579 | NONE | null | As has amazingly and remarkably become a standard response to a new model announcement, will this new transformer model be implemented:
https://arxiv.org/pdf/1909.04164.pdf - KnowBert at EMNLP19. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1247/reactions",
"total_count": 1,
"+1": 1,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1247/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1246 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1246/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1246/comments | https://api.github.com/repos/huggingface/transformers/issues/1246/events | https://github.com/huggingface/transformers/issues/1246 | 491,836,654 | MDU6SXNzdWU0OTE4MzY2NTQ= | 1,246 | breaking change | {
"login": "akosenkov",
"id": 1645649,
"node_id": "MDQ6VXNlcjE2NDU2NDk=",
"avatar_url": "https://avatars.githubusercontent.com/u/1645649?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/akosenkov",
"html_url": "https://github.com/akosenkov",
"followers_url": "https://api.github.com/users/ak... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Indeed. What do you mean by \"examples\"? The docstrings examples?",
"Most of examples folder.\r\nIn particular run_swag.py and lm finetuning scripts.",
"just came here to say the same. ",
"Indeed, I've fixed and cleaned up the examples in 8334993 (the lm finetuning examples are now replaced by `run_lm_finet... | 1,568 | 1,574 | 1,574 | NONE | null | Great job, just in case it went unnoticed:
from revision 995e38b7af1aa325b994246e1bfcc7bf7c9b6b4f
to revision 2c177a87eb5faab8a0abee907ff75898b4886689
examples are broken due to changed orders of parameters in
pytorch_transformers/modeling_bert.py
```
< def forward(self, input_ids, token_type_ids=None, a... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1246/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1246/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1245 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1245/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1245/comments | https://api.github.com/repos/huggingface/transformers/issues/1245/events | https://github.com/huggingface/transformers/issues/1245 | 491,826,672 | MDU6SXNzdWU0OTE4MjY2NzI= | 1,245 | Different performance between pip install vs. download zip code | {
"login": "wyin-Salesforce",
"id": 53835505,
"node_id": "MDQ6VXNlcjUzODM1NTA1",
"avatar_url": "https://avatars.githubusercontent.com/u/53835505?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/wyin-Salesforce",
"html_url": "https://github.com/wyin-Salesforce",
"followers_url": "https://api... | [] | closed | false | null | [] | [
"sorry I found the problem: the pip source code and the zip code downloaded are different, especially for the \"BertForSequenceClassification\" class"
] | 1,568 | 1,568 | 1,568 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
Hi guys,
I meet a weird problem. Basically I am using BERT for sentence pair classification, such as MNLI, RTE tasks.
I installed the pytorch_transformers by "pip install pytorch-transformers" and
> from pytorch_transformers.mo... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1245/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1245/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1244 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1244/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1244/comments | https://api.github.com/repos/huggingface/transformers/issues/1244/events | https://github.com/huggingface/transformers/issues/1244 | 491,815,631 | MDU6SXNzdWU0OTE4MTU2MzE= | 1,244 | unconditional generation with run_generation.py | {
"login": "ehsan-soe",
"id": 12740904,
"node_id": "MDQ6VXNlcjEyNzQwOTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/12740904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ehsan-soe",
"html_url": "https://github.com/ehsan-soe",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi, by unconditional generation do you mean generating a sequence from no context? If so, if using GPT-2, you can set your initial context to be: `<|endoftext|>`. This will generate sequences with no other initial context.\r\n\r\nYou could do so like this:\r\n\r\n```\r\nfrom pytorch_transformers import GPT2Tokeniz... | 1,568 | 1,586 | 1,574 | NONE | null | ## ❓ Questions & Help
Is it possible to have unconditional generation with ``` run_generation.py``` ? I realized the previous script ```run_gpt2.py``` had this option.
Can we use the same ```start_token``` trick? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1244/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1244/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1243 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1243/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1243/comments | https://api.github.com/repos/huggingface/transformers/issues/1243/events | https://github.com/huggingface/transformers/issues/1243 | 491,807,387 | MDU6SXNzdWU0OTE4MDczODc= | 1,243 | Can pytorch-transformers be used to get XLM sentence embeddings for multiple languages? | {
"login": "pbutenee",
"id": 7203172,
"node_id": "MDQ6VXNlcjcyMDMxNzI=",
"avatar_url": "https://avatars.githubusercontent.com/u/7203172?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pbutenee",
"html_url": "https://github.com/pbutenee",
"followers_url": "https://api.github.com/users/pbute... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,573 | 1,573 | NONE | null | ## ❓ Questions & Help
I tried to create a class to get the XLM sentence embeddings in multiple languages.
```
class XLMSentenceEmbeddings(pt.XLMPreTrainedModel):
def __init__(self, config):
super(XLMSentenceEmbeddings, self).__init__(config)
self.transformer = pt.XLMModel(config)
def fo... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1243/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1243/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1242 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1242/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1242/comments | https://api.github.com/repos/huggingface/transformers/issues/1242/events | https://github.com/huggingface/transformers/issues/1242 | 491,783,752 | MDU6SXNzdWU0OTE3ODM3NTI= | 1,242 | Special tokens / XLNet | {
"login": "cherepanovic",
"id": 10064548,
"node_id": "MDQ6VXNlcjEwMDY0NTQ4",
"avatar_url": "https://avatars.githubusercontent.com/u/10064548?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/cherepanovic",
"html_url": "https://github.com/cherepanovic",
"followers_url": "https://api.github.c... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi, in the case of sequence classification, XLNet does indeed use special tokens. For sentence pairs, it looks like this:\r\n\r\n```\r\nA [SEP] B [SEP][CLS]\r\n```\r\n\r\nYou can either create those yourself or use the flag `add_special_tokens` from the `encode` function as follows:\r\n\r\n```\r\ntokenizer.encode(... | 1,568 | 1,578 | 1,574 | NONE | null | it is necessary to add [CLS], [SEP] tokens in case of XLNet transformers?
Thanks!
*I used only the tokenizer.encode() function even if the sample had several sentences and I didn't set any special tokens. I think it was not the right way, isn't? It was done for a classification task. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1242/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1242/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1241 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1241/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1241/comments | https://api.github.com/repos/huggingface/transformers/issues/1241/events | https://github.com/huggingface/transformers/pull/1241 | 491,770,366 | MDExOlB1bGxSZXF1ZXN0MzE2MDczMzk0 | 1,241 | Fixing typo in gpt2 for doc site's class link | {
"login": "mattolson93",
"id": 32203230,
"node_id": "MDQ6VXNlcjMyMjAzMjMw",
"avatar_url": "https://avatars.githubusercontent.com/u/32203230?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/mattolson93",
"html_url": "https://github.com/mattolson93",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"👍 "
] | 1,568 | 1,568 | 1,568 | CONTRIBUTOR | null | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1241/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1241/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1241",
"html_url": "https://github.com/huggingface/transformers/pull/1241",
"diff_url": "https://github.com/huggingface/transformers/pull/1241.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1241.patch",
"merged_at": 1... | |
https://api.github.com/repos/huggingface/transformers/issues/1240 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1240/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1240/comments | https://api.github.com/repos/huggingface/transformers/issues/1240/events | https://github.com/huggingface/transformers/issues/1240 | 491,755,814 | MDU6SXNzdWU0OTE3NTU4MTQ= | 1,240 | ModuleNotFoundError in distillation/scripts/binarized_data.py | {
"login": "MatejUlcar",
"id": 26550612,
"node_id": "MDQ6VXNlcjI2NTUwNjEy",
"avatar_url": "https://avatars.githubusercontent.com/u/26550612?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/MatejUlcar",
"html_url": "https://github.com/MatejUlcar",
"followers_url": "https://api.github.com/use... | [] | closed | false | null | [] | [
"Hello @MatejUlcar \r\nThanks for pointing that out. I fixed it once and for all by having a local logger (and not importing the global one) in commit 32e1332acf6fd1ad372b81c296d43be441d3b0b1. "
] | 1,568 | 1,568 | 1,568 | NONE | null | Hello,
importing logger (still? I've seen previous issues, but this is the first time I'm running the code myself) throws the ModuleNotFoundError.
```
Traceback (most recent call last):
File "examples/distillation/scripts/binarized_data.py", line 25, in <module>
from examples.distillation.utils import l... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1240/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1240/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1239 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1239/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1239/comments | https://api.github.com/repos/huggingface/transformers/issues/1239/events | https://github.com/huggingface/transformers/issues/1239 | 491,632,680 | MDU6SXNzdWU0OTE2MzI2ODA= | 1,239 | how to finetuning with roberta-large | {
"login": "zhaoguangxiang",
"id": 17742385,
"node_id": "MDQ6VXNlcjE3NzQyMzg1",
"avatar_url": "https://avatars.githubusercontent.com/u/17742385?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/zhaoguangxiang",
"html_url": "https://github.com/zhaoguangxiang",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | [
"What do you mean by \"latest\" version of PyTorch tranformers.\r\n\r\nAre you using a release or installing from source from master?",
"> What do you mean by \"latest\" version of PyTorch tranformers.\r\n> Are you using a release or installing from source from master?\r\nThanks for your reply.\r\npytorch transfo... | 1,568 | 1,568 | 1,568 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Roberta-large):
Language I am using the model on (English, ):
The problem arise when using:
* [ ] my own modified scripts: (give details)
export CUDA_VISIBLE_DEVICES=2
export GLUE_DIR=/home/zhaoguangxiang/bert/glue_data
DATA=MNLI
NUM_CLASSES=3
... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1239/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1239/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1238 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1238/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1238/comments | https://api.github.com/repos/huggingface/transformers/issues/1238/events | https://github.com/huggingface/transformers/pull/1238 | 491,617,708 | MDExOlB1bGxSZXF1ZXN0MzE1OTQ5ODAx | 1,238 | BLUE | {
"login": "anhnt170489",
"id": 24732444,
"node_id": "MDQ6VXNlcjI0NzMyNDQ0",
"avatar_url": "https://avatars.githubusercontent.com/u/24732444?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/anhnt170489",
"html_url": "https://github.com/anhnt170489",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"@LysandreJik @julien-c one of you want to give a look?",
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1238?src=pr&el=h1) Report\n> Merging [#1238](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1238?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-t... | 1,568 | 1,570 | 1,570 | NONE | null | In this PR:
- I add BertForMultiLabelClassification, RobertaForTokenClassification, RobertaForMultiLabelClassification.
- I add examples for Finetuning the BERT, RoBERTa models for tasks on BLUE (https://github.com/ncbi-nlp/BLUE_Benchmark). BLUE (Biomedical Language Understanding Evaluation) is similar to GLUE, but... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1238/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1238/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1238",
"html_url": "https://github.com/huggingface/transformers/pull/1238",
"diff_url": "https://github.com/huggingface/transformers/pull/1238.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1238.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1237 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1237/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1237/comments | https://api.github.com/repos/huggingface/transformers/issues/1237/events | https://github.com/huggingface/transformers/issues/1237 | 491,581,367 | MDU6SXNzdWU0OTE1ODEzNjc= | 1,237 | Issue in fine-tuning distilbert on Squad 1.0 | {
"login": "pragnakalpdev",
"id": 48992007,
"node_id": "MDQ6VXNlcjQ4OTkyMDA3",
"avatar_url": "https://avatars.githubusercontent.com/u/48992007?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pragnakalpdev",
"html_url": "https://github.com/pragnakalpdev",
"followers_url": "https://api.githu... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hello @pragnakalpdev,\r\nDid you change the `run_squad.py` file to include distilbert ([here](https://github.com/huggingface/pytorch-transformers/blob/master/examples/run_squad.py#L58) for instance)?\r\nCan you check the warnings ``Weights from XXX not used in YYY``?",
"Hello @VictorSanh,\r\nThank You for your r... | 1,568 | 1,615 | 1,577 | NONE | null | When I tried to finetune the distillbert (using run_squad.py in examples folder), the model reaches the F1 score of 17.43 on dev set but you have mentioned that the F1 score is 86.2. Can you help me with what I am doing wrong at the time of fine-tuning?
Below is the command that I am using
python ./examples/run_squ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1237/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1237/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1236 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1236/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1236/comments | https://api.github.com/repos/huggingface/transformers/issues/1236/events | https://github.com/huggingface/transformers/issues/1236 | 491,478,744 | MDU6SXNzdWU0OTE0Nzg3NDQ= | 1,236 | Roberta for squad | {
"login": "search4mahesh",
"id": 4182331,
"node_id": "MDQ6VXNlcjQxODIzMzE=",
"avatar_url": "https://avatars.githubusercontent.com/u/4182331?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/search4mahesh",
"html_url": "https://github.com/search4mahesh",
"followers_url": "https://api.github.... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,573 | 1,573 | NONE | null | Hi,
Please add Roberta for squad.
Thanks
Mahesh | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1236/reactions",
"total_count": 5,
"+1": 5,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1236/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1235 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1235/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1235/comments | https://api.github.com/repos/huggingface/transformers/issues/1235/events | https://github.com/huggingface/transformers/issues/1235 | 491,441,661 | MDU6SXNzdWU0OTE0NDE2NjE= | 1,235 | Quick questions about details | {
"login": "syahrulhamdani",
"id": 23093968,
"node_id": "MDQ6VXNlcjIzMDkzOTY4",
"avatar_url": "https://avatars.githubusercontent.com/u/23093968?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/syahrulhamdani",
"html_url": "https://github.com/syahrulhamdani",
"followers_url": "https://api.gi... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,573 | 1,573 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
Can someone explain the difference between `mem_len`, `mlen`, and `ext_len` in `TransfoXLModel`? While the documentation has stated below
![Screen Shot 2019-09-10 at 11 18 40](https://user-images.githubusercontent.com/23093968/645... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1235/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1235/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1234 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1234/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1234/comments | https://api.github.com/repos/huggingface/transformers/issues/1234/events | https://github.com/huggingface/transformers/issues/1234 | 491,409,856 | MDU6SXNzdWU0OTE0MDk4NTY= | 1,234 | ❓ How to finetune `token_type_ids` of RoBERTa ? | {
"login": "astariul",
"id": 43774355,
"node_id": "MDQ6VXNlcjQzNzc0MzU1",
"avatar_url": "https://avatars.githubusercontent.com/u/43774355?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/astariul",
"html_url": "https://github.com/astariul",
"followers_url": "https://api.github.com/users/ast... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"What I have done is :\r\n\r\n```python\r\nmodel = RobertaModel.from_pretrained('roberta-base')\r\nmodel.config.type_vocab_size = 2\r\nsingle_emb = model.embeddings.token_type_embeddings\r\nmodel.embeddings.token_type_embeddings = torch.nn.Embedding(2, single_emb.embedding_dim)\r\nmodel.embeddings.token_type_embedd... | 1,568 | 1,671 | 1,578 | CONTRIBUTOR | null | ## ❓ Questions & Help
RoBERTa model does not use `token_type_ids`.
However it is mentioned in the documentation :
> you will have to train it during finetuning
Indeed, I would like to train it during finetuning. I tried to load the model with :
`model = RobertaModel.from_pretrained('roberta-base', type_v... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1234/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1234/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1233 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1233/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1233/comments | https://api.github.com/repos/huggingface/transformers/issues/1233/events | https://github.com/huggingface/transformers/pull/1233 | 491,381,533 | MDExOlB1bGxSZXF1ZXN0MzE1NzYxMTMy | 1,233 | Fix to prevent crashing on assert len(tokens_b)>=1 | {
"login": "searchivarius",
"id": 825650,
"node_id": "MDQ6VXNlcjgyNTY1MA==",
"avatar_url": "https://avatars.githubusercontent.com/u/825650?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/searchivarius",
"html_url": "https://github.com/searchivarius",
"followers_url": "https://api.github.co... | [] | closed | false | null | [] | [
"Ok, merging.\r\nNote that we are in the process of deprecating these finetuning scripts and replacing them with the common `run_lm_finetuning.py` which handles several models.",
"Thank you @thomwolf sorry didn't realize this was deprecated."
] | 1,568 | 1,568 | 1,568 | CONTRIBUTOR | null | Thank you for the awesome library! One little issue that I have sometimes with somewhat "noisy" text is that Bert tokenizer fails to process some weird stuff. In such a case, one unexpectedly gets no tokens and the converter fails on assert with a message like this one:
```
Epoch: 0%| | 0/1 [00:00<?, ?it/s... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1233/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1233/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1233",
"html_url": "https://github.com/huggingface/transformers/pull/1233",
"diff_url": "https://github.com/huggingface/transformers/pull/1233.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1233.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1232 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1232/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1232/comments | https://api.github.com/repos/huggingface/transformers/issues/1232/events | https://github.com/huggingface/transformers/issues/1232 | 491,330,031 | MDU6SXNzdWU0OTEzMzAwMzE= | 1,232 | Can't reproduce XNLI zero-shot results from MBERT in Chinese | {
"login": "edchengg",
"id": 20430102,
"node_id": "MDQ6VXNlcjIwNDMwMTAy",
"avatar_url": "https://avatars.githubusercontent.com/u/20430102?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/edchengg",
"html_url": "https://github.com/edchengg",
"followers_url": "https://api.github.com/users/edc... | [] | closed | false | null | [] | [
"Hey @edchengg, I'm running into the same problem. Were you able to figure this out? thanks\r\n",
"Anyone else who stumbles here.\r\n**Fix**: Just use bert-based-multilingual-cased as shown here https://huggingface.co/transformers/v2.3.0/examples.html.\r\nWhen I used Google's mBERT and made it pyTorch compatible ... | 1,568 | 1,596 | 1,568 | NONE | null | ## ❓ Questions & Help
Hi guys,
I am trying to reproduce the XNLI zero-shot transfer results from MBERT.
With the same code and same checkpoint but different language for the test set, I am not able to reproduce the results for Chinese, Arabic, and Urdu. Does anyone encounter the same problem? Thanks!
Model |E... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1232/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1232/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1231 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1231/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1231/comments | https://api.github.com/repos/huggingface/transformers/issues/1231/events | https://github.com/huggingface/transformers/issues/1231 | 491,265,028 | MDU6SXNzdWU0OTEyNjUwMjg= | 1,231 | Unable to load DistilBertModel after training | {
"login": "dalefwillis",
"id": 2707681,
"node_id": "MDQ6VXNlcjI3MDc2ODE=",
"avatar_url": "https://avatars.githubusercontent.com/u/2707681?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/dalefwillis",
"html_url": "https://github.com/dalefwillis",
"followers_url": "https://api.github.com/us... | [] | closed | false | null | [] | [
"Hi @dalefwillis,\r\nYou simply have to rename your last checkpoint (I guess in your case it's _\"model_epoch_2.pth\"_) to _\"pytorch_model.bin\"_ --> `mv model_epoch_2.pth pytorch_model.bin`.\r\nI updated the training code so that the very last _\"model_epoch_*.pth\"_ checkpoint is also saved as _\"pytorch_model.b... | 1,568 | 1,568 | 1,568 | NONE | null | ## ❓ Questions & Help
I'm following the example to train a DistilBert model from scratch from: examples/distillation/README.md
I perform the training step:
```
python examples/distillation/train.py --dump_path ser_dir/sm_training_1 --data_file data/sm_bin_text.bert-base-uncased.pickle --token_counts data/s... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1231/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1231/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1230 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1230/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1230/comments | https://api.github.com/repos/huggingface/transformers/issues/1230/events | https://github.com/huggingface/transformers/issues/1230 | 491,187,452 | MDU6SXNzdWU0OTExODc0NTI= | 1,230 | How to deal with oov tokens with pretrained models | {
"login": "skurzhanskyi",
"id": 17638837,
"node_id": "MDQ6VXNlcjE3NjM4ODM3",
"avatar_url": "https://avatars.githubusercontent.com/u/17638837?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/skurzhanskyi",
"html_url": "https://github.com/skurzhanskyi",
"followers_url": "https://api.github.c... | [] | closed | false | null | [] | [
"Hello! We have a method called `add_tokens` in our tokenizers that does just that. [Here's the relevant information](https://huggingface.co/pytorch-transformers/main_classes/tokenizer.html#pytorch_transformers.PreTrainedTokenizer.add_tokens) in the documentation.",
"Thanks a lot for your answer. That's exactly w... | 1,568 | 1,568 | 1,568 | NONE | null | ## ❓ Questions & Help
Can you, please, give an advice on how to handle out of vocabulary word? Just to use `[UNK]` token or there is a way to add this token to vocabulary and thus to train embedding for it?
Also, I noticed that oov words by default return multiple tokens. In my task (sequence tagging) I would like ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1230/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1230/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1229 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1229/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1229/comments | https://api.github.com/repos/huggingface/transformers/issues/1229/events | https://github.com/huggingface/transformers/pull/1229 | 491,137,676 | MDExOlB1bGxSZXF1ZXN0MzE1NTY0ODYx | 1,229 | changes in evaluate function in run_lm_finetuning.py | {
"login": "SKRohit",
"id": 9626333,
"node_id": "MDQ6VXNlcjk2MjYzMzM=",
"avatar_url": "https://avatars.githubusercontent.com/u/9626333?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/SKRohit",
"html_url": "https://github.com/SKRohit",
"followers_url": "https://api.github.com/users/SKRohit/... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1229?src=pr&el=h1) Report\n> Merging [#1229](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1229?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/84d346b68707f3c43903b122baae76ae022ef420?... | 1,568 | 1,568 | 1,568 | CONTRIBUTOR | null | changed the return value of `evaluate` function from `results` to `result` and also removed unused empty dict `results` | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1229/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1229/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1229",
"html_url": "https://github.com/huggingface/transformers/pull/1229",
"diff_url": "https://github.com/huggingface/transformers/pull/1229.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1229.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1228 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1228/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1228/comments | https://api.github.com/repos/huggingface/transformers/issues/1228/events | https://github.com/huggingface/transformers/pull/1228 | 491,083,752 | MDExOlB1bGxSZXF1ZXN0MzE1NTIxMzI2 | 1,228 | Trying to fix the head masking test | {
"login": "thomwolf",
"id": 7353373,
"node_id": "MDQ6VXNlcjczNTMzNzM=",
"avatar_url": "https://avatars.githubusercontent.com/u/7353373?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/thomwolf",
"html_url": "https://github.com/thomwolf",
"followers_url": "https://api.github.com/users/thomw... | [] | closed | false | null | [] | [
"@LysandreJik was this still WIP or finished?",
"It solved the problems with head masking -> finished!"
] | 1,568 | 1,578 | 1,568 | MEMBER | null | Reviving this PR from @LysandreJik which tried to fix the head masking failing test by making random seed accessible anywhere within the common tests. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1228/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1228/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1228",
"html_url": "https://github.com/huggingface/transformers/pull/1228",
"diff_url": "https://github.com/huggingface/transformers/pull/1228.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1228.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1227 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1227/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1227/comments | https://api.github.com/repos/huggingface/transformers/issues/1227/events | https://github.com/huggingface/transformers/issues/1227 | 491,028,877 | MDU6SXNzdWU0OTEwMjg4Nzc= | 1,227 | class DistilBertForMultiLabelSequenceClassification() | {
"login": "emtropyml",
"id": 53178223,
"node_id": "MDQ6VXNlcjUzMTc4MjIz",
"avatar_url": "https://avatars.githubusercontent.com/u/53178223?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/emtropyml",
"html_url": "https://github.com/emtropyml",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,568 | 1,573 | 1,573 | NONE | null | ## 🚀 Feature
Distil BERT For Multi-Label Sequence Classification
## Motivation
To do multi-label text classification using DistilBERT
## Additional context
None | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1227/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1227/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1226 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1226/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1226/comments | https://api.github.com/repos/huggingface/transformers/issues/1226/events | https://github.com/huggingface/transformers/issues/1226 | 490,836,436 | MDU6SXNzdWU0OTA4MzY0MzY= | 1,226 | Question on the position embedding of DistilBERT | {
"login": "gpengzhi",
"id": 16913241,
"node_id": "MDQ6VXNlcjE2OTEzMjQx",
"avatar_url": "https://avatars.githubusercontent.com/u/16913241?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/gpengzhi",
"html_url": "https://github.com/gpengzhi",
"followers_url": "https://api.github.com/users/gpe... | [] | closed | false | null | [] | [
"Hello @gpengzhi \r\nYou're right, the name is quite confusing: the second matrix of embeddings that you're showing is actually initialized from `bert-base-uncased` (compare with `bert = BertModel.from_pretrained('bert-base-uncased'); print(bert.embeddings.position_embeddings.weight)`). Once initialized, these posi... | 1,567 | 1,568 | 1,568 | CONTRIBUTOR | null | ## ❓ Questions & Help
To the best of my knowledge, sinusoidal position embedding is used in the training procedure of DistilBERT, which is computed by [create_sinusoidal_embeddings](https://github.com/huggingface/pytorch-transformers/blob/master/pytorch_transformers/modeling_distilbert.py#L52).
When I compute the... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1226/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1226/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1225 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1225/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1225/comments | https://api.github.com/repos/huggingface/transformers/issues/1225/events | https://github.com/huggingface/transformers/issues/1225 | 490,810,773 | MDU6SXNzdWU0OTA4MTA3NzM= | 1,225 | Bert output last hidden state | {
"login": "ehsan-soe",
"id": 12740904,
"node_id": "MDQ6VXNlcjEyNzQwOTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/12740904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ehsan-soe",
"html_url": "https://github.com/ehsan-soe",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hello! I believe that you are currently computing values for your padding indices, resulting in your confusion. There is a parameter `attention_mask` to be passed to the `forward`/`__call__` method which will prevent the values to be computed for the padded indices!",
"@LysandreJik thanks for replying.\r\nConsid... | 1,567 | 1,657 | 1,573 | NONE | null | ## ❓ Questions & Help
Hi,
Suppose we have an utterance of length 24 (considering special tokens) and we right-pad it with 0 to max length of 64.
If we use Bert pertained model to get the last hidden states, the output would be of size [1, 64, 768].
Can we use just the first 24 as the hidden states of the utter... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1225/reactions",
"total_count": 2,
"+1": 2,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1225/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1224 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1224/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1224/comments | https://api.github.com/repos/huggingface/transformers/issues/1224/events | https://github.com/huggingface/transformers/issues/1224 | 490,780,821 | MDU6SXNzdWU0OTA3ODA4MjE= | 1,224 | Remove duplicate hidden_states of the last layer in BertEncoder in modeling_bert.py | {
"login": "usertomlin",
"id": 2401439,
"node_id": "MDQ6VXNlcjI0MDE0Mzk=",
"avatar_url": "https://avatars.githubusercontent.com/u/2401439?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/usertomlin",
"html_url": "https://github.com/usertomlin",
"followers_url": "https://api.github.com/users... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi! Indeed, the BERT-base only has 12 layers. The `all_hidden_states` is 13-dimensional however because it keeps track of the inputs as well.\r\n\r\nIn the code you have shown, the `hidden_states` variable is computed between the two underlined variables you mentioned. None of it is redundant :)! ",
"> Hi! Indee... | 1,567 | 1,573 | 1,573 | NONE | null | ## 🚀 Feature
<!-- A clear and concise description of the feature proposal. Please provide a link to the paper and code in case they exist. -->
For "class BertEncoder" in modeling_bert.py, remove duplicate hidden_states of the last layer
## Motivation
![image](https://user-images.githubusercontent.com/24... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1224/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1224/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1223 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1223/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1223/comments | https://api.github.com/repos/huggingface/transformers/issues/1223/events | https://github.com/huggingface/transformers/issues/1223 | 490,715,837 | MDU6SXNzdWU0OTA3MTU4Mzc= | 1,223 | [RuntimeError: sizes must be non-negative] : XLnet, Large and Base | {
"login": "pythonometrist",
"id": 4297337,
"node_id": "MDQ6VXNlcjQyOTczMzc=",
"avatar_url": "https://avatars.githubusercontent.com/u/4297337?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/pythonometrist",
"html_url": "https://github.com/pythonometrist",
"followers_url": "https://api.gith... | [] | closed | false | null | [] | [
"Can you post a simple example showing the behavior and a detailed error message?",
"-----------------------------------------------------------------------\r\nRuntimeError Traceback (most recent call last)\r\n<ipython-input-1-e5cfbf5c4eca> in <module>\r\n 127 t... | 1,567 | 1,568 | 1,568 | NONE | null | ## 🐛 Bug
<!-- Important information -->
Model I am using (Bert, XLNet....): XLNet
Language I am using the model on (English, Chinese....): English
The problem arise when using:
* [ ] the official example scripts: (give details)
* [x] my own modified scripts: (give details)
The tasks I am working... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1223/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1223/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1222 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1222/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1222/comments | https://api.github.com/repos/huggingface/transformers/issues/1222/events | https://github.com/huggingface/transformers/issues/1222 | 490,711,350 | MDU6SXNzdWU0OTA3MTEzNTA= | 1,222 | Citing DistilBERT | {
"login": "rishibommasani",
"id": 47439426,
"node_id": "MDQ6VXNlcjQ3NDM5NDI2",
"avatar_url": "https://avatars.githubusercontent.com/u/47439426?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rishibommasani",
"html_url": "https://github.com/rishibommasani",
"followers_url": "https://api.gi... | [] | closed | false | null | [] | [
"Hello @rishibommasani \r\nThank you for your question.\r\nFor citing DistilBERT, you're right, there is no formal write-up like an arXiv paper yet (it's definitely in our TODO stack). For the moment, I would recommend citing the blogpost as an URL.\r\nVictor"
] | 1,567 | 1,568 | 1,568 | NONE | null | Currently, my understanding is citing the repo/codebase should be done by via a link (i.e. in the paper as a footnote) as there is no citation (i.e. in BibTeX style) yet.
For citing DistilBERT (the released model and distillation approach), how should this be done? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1222/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1222/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1221 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1221/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1221/comments | https://api.github.com/repos/huggingface/transformers/issues/1221/events | https://github.com/huggingface/transformers/issues/1221 | 490,691,858 | MDU6SXNzdWU0OTA2OTE4NTg= | 1,221 | Hi there, is bert-large-uncased-whole-word-masking-finetuned-squad trained for Squad 1.0 or 2.0? | {
"login": "JianLiu91",
"id": 22347717,
"node_id": "MDQ6VXNlcjIyMzQ3NzE3",
"avatar_url": "https://avatars.githubusercontent.com/u/22347717?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/JianLiu91",
"html_url": "https://github.com/JianLiu91",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi! I believe this checkpoint originates from the training specified [there](https://huggingface.co/pytorch-transformers/examples.html#squad).\r\n\r\nThe SQuAD version would then be 1.1!",
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further a... | 1,567 | 1,573 | 1,573 | NONE | null | I mean, whether the training data contains examples with no answer? | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1221/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1221/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1220 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1220/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1220/comments | https://api.github.com/repos/huggingface/transformers/issues/1220/events | https://github.com/huggingface/transformers/issues/1220 | 490,668,010 | MDU6SXNzdWU0OTA2NjgwMTA= | 1,220 | RuntimeError: Gather got an input of invalid size: got [2, 3, 12, 256, 64], but expected [2, 4, 12, 256, 64] (gather at /opt/conda/conda-bld/pytorch_1544199946412/work/torch/csrc/cuda/comm.cpp:227) | {
"login": "ehsan-soe",
"id": 12740904,
"node_id": "MDQ6VXNlcjEyNzQwOTA0",
"avatar_url": "https://avatars.githubusercontent.com/u/12740904?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/ehsan-soe",
"html_url": "https://github.com/ehsan-soe",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This is a wild guess since I don't have access to your modified version, but I feel like this has to do with a mismatch in the batch size (expecting a batch size of 4 but receiving a batch size of 3).\r\n\r\nCould you check your input tensor and label tensor sizes and get back to me so I can try and reproduce it o... | 1,567 | 1,651 | 1,587 | NONE | null | ## ❓ Questions & Help
Hi,
I am running a modified version of ```run_lm_finetuning.py```, it was working fine and model checkpoints have been saved, until the last step of the first epoch (9677/9678), where I got this error:
```
Traceback (most recent call last):██████████████████████████████████████████████████... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1220/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1220/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1219 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1219/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1219/comments | https://api.github.com/repos/huggingface/transformers/issues/1219/events | https://github.com/huggingface/transformers/pull/1219 | 490,609,485 | MDExOlB1bGxSZXF1ZXN0MzE1MTc4NDMw | 1,219 | fix tokenize(): potential bug of splitting pretrained tokens with newly added tokens | {
"login": "askerlee",
"id": 1575461,
"node_id": "MDQ6VXNlcjE1NzU0NjE=",
"avatar_url": "https://avatars.githubusercontent.com/u/1575461?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/askerlee",
"html_url": "https://github.com/askerlee",
"followers_url": "https://api.github.com/users/asker... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/1219?src=pr&el=h1) Report\n> Merging [#1219](https://codecov.io/gh/huggingface/transformers/pull/1219?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/80faf22b4ac194061a08fde09ad8b202118c151e?src=pr&el=desc) will **i... | 1,567 | 1,584 | 1,584 | NONE | null | In the tokenizer base class, `split_on_token()` attempts to split input text by each of the added tokens. Because it uses `text.split(tok)`, it may accidentally split a token in the pretrained vocabulary at the middle.
For example a new token "ht" is added to the vocabulary. Then "light" will be split into `["lig", ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1219/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1219/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1219",
"html_url": "https://github.com/huggingface/transformers/pull/1219",
"diff_url": "https://github.com/huggingface/transformers/pull/1219.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1219.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1218 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1218/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1218/comments | https://api.github.com/repos/huggingface/transformers/issues/1218/events | https://github.com/huggingface/transformers/issues/1218 | 490,564,978 | MDU6SXNzdWU0OTA1NjQ5Nzg= | 1,218 | How to set the weight decay in other layers after BERT output? | {
"login": "g-jing",
"id": 44223191,
"node_id": "MDQ6VXNlcjQ0MjIzMTkx",
"avatar_url": "https://avatars.githubusercontent.com/u/44223191?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/g-jing",
"html_url": "https://github.com/g-jing",
"followers_url": "https://api.github.com/users/g-jing/fo... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"@RoderickGu. maybe try:\r\n```python\r\n bert_param_optimizer = list(model.bert.named_parameters())\r\n lstm_param_optimizer = list(model.bilstm.named_parameters())\r\n crf_param_optimizer = list(model.crf.named_parameters())\r\n linear_param_optimizer = list(model.classifier.named_parameters())\r\n ... | 1,567 | 1,573 | 1,573 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
I notice that we should set weight decay of bias and LayerNorm.weight to zero and set weight decay of other parameter in BERT to 0.01. But how to set the weight decay of other layer such as the classifier after BERT? Thanks | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1218/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1218/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1217 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1217/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1217/comments | https://api.github.com/repos/huggingface/transformers/issues/1217/events | https://github.com/huggingface/transformers/pull/1217 | 490,530,053 | MDExOlB1bGxSZXF1ZXN0MzE1MTE5Njc5 | 1,217 | Fixing head masking test | {
"login": "thomwolf",
"id": 7353373,
"node_id": "MDQ6VXNlcjczNTMzNzM=",
"avatar_url": "https://avatars.githubusercontent.com/u/7353373?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/thomwolf",
"html_url": "https://github.com/thomwolf",
"followers_url": "https://api.github.com/users/thomw... | [] | closed | false | null | [] | [
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1217?src=pr&el=h1) Report\n> Merging [#1217](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1217?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/pytorch-transformers/commit/ee027c89f2b8fd0338df39b7e0b48345ea132e99?... | 1,567 | 1,651 | 1,569 | MEMBER | null | Try to fix the Nan in head masking tests by removing them. | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1217/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1217/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1217",
"html_url": "https://github.com/huggingface/transformers/pull/1217",
"diff_url": "https://github.com/huggingface/transformers/pull/1217.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1217.patch",
"merged_at": n... |
https://api.github.com/repos/huggingface/transformers/issues/1216 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1216/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1216/comments | https://api.github.com/repos/huggingface/transformers/issues/1216/events | https://github.com/huggingface/transformers/issues/1216 | 490,468,131 | MDU6SXNzdWU0OTA0NjgxMzE= | 1,216 | Is there any sample code for fine-tuning BERT on sequence labeling tasks, e.g., NER on CoNLL-2003? | {
"login": "tuvuumass",
"id": 23730882,
"node_id": "MDQ6VXNlcjIzNzMwODgy",
"avatar_url": "https://avatars.githubusercontent.com/u/23730882?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/tuvuumass",
"html_url": "https://github.com/tuvuumass",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi @tuvuumass,\r\n\r\nIssue https://github.com/huggingface/pytorch-transformers/issues/64 is a good start for sequence labeling tasks. It also points to some repositories that show how to fine-tune BERT with PyTorch-Transformers (with focus on NER).\r\n\r\nNevertheless, it would be awesome to get some kind of fine... | 1,567 | 1,659 | 1,577 | CONTRIBUTOR | null | ## ❓ Questions & Help
Is there any sample code for fine-tuning BERT on sequence labeling tasks, e.g., NER on CoNLL-2003, using BertForTokenClassification?
<!-- A clear and concise description of the question. --> | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1216/reactions",
"total_count": 4,
"+1": 4,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1216/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1215 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1215/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1215/comments | https://api.github.com/repos/huggingface/transformers/issues/1215/events | https://github.com/huggingface/transformers/issues/1215 | 490,420,803 | MDU6SXNzdWU0OTA0MjA4MDM= | 1,215 | Cut off sequences of length greater than max_length= 512 for roberta | {
"login": "rush86999",
"id": 16848240,
"node_id": "MDQ6VXNlcjE2ODQ4MjQw",
"avatar_url": "https://avatars.githubusercontent.com/u/16848240?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/rush86999",
"html_url": "https://github.com/rush86999",
"followers_url": "https://api.github.com/users/... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Hi! Indeed RoBERTa has a max length of 512. Why don't you slice your text?",
"I was hoping the tokenizer could take care of it as a functionality? The actual hope is not to throw an error but allow training with it by increasing the positional encoding as a way to allow training on the whole text length. Is this... | 1,567 | 1,573 | 1,573 | NONE | null | ## 🚀 Feature
<!-- A clear and concise description of the feature proposal. Please provide a link to the paper and code in case they exist. -->
Roberta uses max_length of 512 but text to tokenize is variable length. Is there an option to cut off source text to maximum length during tokenization process?
## Motivat... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1215/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1215/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1214 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1214/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1214/comments | https://api.github.com/repos/huggingface/transformers/issues/1214/events | https://github.com/huggingface/transformers/pull/1214 | 490,410,833 | MDExOlB1bGxSZXF1ZXN0MzE1MDIyOTM3 | 1,214 | Better examples | {
"login": "LysandreJik",
"id": 30755778,
"node_id": "MDQ6VXNlcjMwNzU1Nzc4",
"avatar_url": "https://avatars.githubusercontent.com/u/30755778?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/LysandreJik",
"html_url": "https://github.com/LysandreJik",
"followers_url": "https://api.github.com/... | [] | closed | false | null | [] | [
"There were indeed quite a few artifacts. I fixed them in the two latest commits.",
"# [Codecov](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1214?src=pr&el=h1) Report\n> Merging [#1214](https://codecov.io/gh/huggingface/pytorch-transformers/pull/1214?src=pr&el=desc) into [master](https://codecov.i... | 1,567 | 1,576 | 1,568 | MEMBER | null | Refactored the examples section: removed old and not up-to-date examples and added new examples for fine-tuning and generation.
The `examples` file is not in the `/doc/source` folder anymore but in the `/examples` folder. It is therefore visible when users open the folder on GitHub.
Note: In order to generate th... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1214/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1214/timeline | null | false | {
"url": "https://api.github.com/repos/huggingface/transformers/pulls/1214",
"html_url": "https://github.com/huggingface/transformers/pull/1214",
"diff_url": "https://github.com/huggingface/transformers/pull/1214.diff",
"patch_url": "https://github.com/huggingface/transformers/pull/1214.patch",
"merged_at": 1... |
https://api.github.com/repos/huggingface/transformers/issues/1213 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1213/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1213/comments | https://api.github.com/repos/huggingface/transformers/issues/1213/events | https://github.com/huggingface/transformers/issues/1213 | 490,265,945 | MDU6SXNzdWU0OTAyNjU5NDU= | 1,213 | Fine-tuned RoBERTa models on CPU | {
"login": "avostryakov",
"id": 174194,
"node_id": "MDQ6VXNlcjE3NDE5NA==",
"avatar_url": "https://avatars.githubusercontent.com/u/174194?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/avostryakov",
"html_url": "https://github.com/avostryakov",
"followers_url": "https://api.github.com/user... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n"
] | 1,567 | 1,573 | 1,573 | NONE | null | ## ❓ Questions & Help
Because of using FusedLayerNorm as BertLayerNorm from apex library after finetuning a saved model isn't possible to use on CPU with apex and Cuda installed. What will be the easiest way to run finetuned models on a server without a GPU? I see that the easiest way to just use the python version ... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1213/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1213/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1212 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1212/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1212/comments | https://api.github.com/repos/huggingface/transformers/issues/1212/events | https://github.com/huggingface/transformers/issues/1212 | 490,232,022 | MDU6SXNzdWU0OTAyMzIwMjI= | 1,212 | LSTM returns nan after using the pretrained BERT embedding as input | {
"login": "laifi",
"id": 34584914,
"node_id": "MDQ6VXNlcjM0NTg0OTE0",
"avatar_url": "https://avatars.githubusercontent.com/u/34584914?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/laifi",
"html_url": "https://github.com/laifi",
"followers_url": "https://api.github.com/users/laifi/follow... | [] | closed | false | null | [] | [
"i found that the problem is related to the data\r\ni will close this issue , or if anyone has the permission he could delete this :) "
] | 1,567 | 1,568 | 1,567 | NONE | null | Hello ,
i'm using the pretrained Bert model (from pytorch-transformers) to get the contextual embedding of a written text , i summed the last 4 hidden layers outputs (i red that the concatenation of the last four layers usually produce the best results )
than i use a LSTM layer with attention to get the paragraph l... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1212/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1212/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1211 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1211/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1211/comments | https://api.github.com/repos/huggingface/transformers/issues/1211/events | https://github.com/huggingface/transformers/issues/1211 | 490,191,175 | MDU6SXNzdWU0OTAxOTExNzU= | 1,211 | How to fine tune small dataset? | {
"login": "g-jing",
"id": 44223191,
"node_id": "MDQ6VXNlcjQ0MjIzMTkx",
"avatar_url": "https://avatars.githubusercontent.com/u/44223191?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/g-jing",
"html_url": "https://github.com/g-jing",
"followers_url": "https://api.github.com/users/g-jing/fo... | [
{
"id": 1314768611,
"node_id": "MDU6TGFiZWwxMzE0NzY4NjEx",
"url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix",
"name": "wontfix",
"color": "ffffff",
"default": true,
"description": null
}
] | closed | false | null | [] | [
"Usually, you should train more epochs. Optimal batch size from 8 to 16. I'm not sure but maybe learning rate should be lower in this case. You can try",
"Thanks for your suggestion. I will try that. Besides, do you think I should modify the warmup step or just set it as 10% of total step just like in the origina... | 1,567 | 1,575 | 1,573 | NONE | null | ## ❓ Questions & Help
<!-- A clear and concise description of the question. -->
Most people test bert on large dataset, but when it comes to small dataset, I assume the fine tune process and batch size maybe different. Besides, the dataset domain is twitter domain, which is kind of different from BERT pretrained co... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1211/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1211/timeline | completed | null | null |
https://api.github.com/repos/huggingface/transformers/issues/1210 | https://api.github.com/repos/huggingface/transformers | https://api.github.com/repos/huggingface/transformers/issues/1210/labels{/name} | https://api.github.com/repos/huggingface/transformers/issues/1210/comments | https://api.github.com/repos/huggingface/transformers/issues/1210/events | https://github.com/huggingface/transformers/issues/1210 | 490,155,096 | MDU6SXNzdWU0OTAxNTUwOTY= | 1,210 | Finetuning distilbert-base-uncased | {
"login": "aah39",
"id": 54974926,
"node_id": "MDQ6VXNlcjU0OTc0OTI2",
"avatar_url": "https://avatars.githubusercontent.com/u/54974926?v=4",
"gravatar_id": "",
"url": "https://api.github.com/users/aah39",
"html_url": "https://github.com/aah39",
"followers_url": "https://api.github.com/users/aah39/follow... | [] | closed | false | null | [] | [
"Hi! Yes we're in the process of adding DistilBERT to the examples. Until then, you can simply edit the script to add it.\r\n\r\nPlease note that the `simple_lm_finetuning` script is now deprecated in favor of `run_lm_finetuning`.",
"Hi there @aah39 , I came across the same issue in run_glue.py when I tried to fi... | 1,567 | 1,575 | 1,575 | NONE | null | ## ❓ Questions & Help
When trying to finetune distilbert-base-uncased on my own dataset
I receive the following error message:
ERROR - pytorch_transformers.tokenization_utils - Model name 'distilbert-base-uncased' was not found in model name list (bert-base-uncased,bert-large-uncased, bert-base-cased, bert-large... | {
"url": "https://api.github.com/repos/huggingface/transformers/issues/1210/reactions",
"total_count": 0,
"+1": 0,
"-1": 0,
"laugh": 0,
"hooray": 0,
"confused": 0,
"heart": 0,
"rocket": 0,
"eyes": 0
} | https://api.github.com/repos/huggingface/transformers/issues/1210/timeline | completed | null | null |
Subsets and Splits
No community queries yet
The top public SQL queries from the community will appear here once available.