Datasets:

---

DanielPFlorian

/

Transformers-Github-Issues

like 0

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=h1) Report\n> Merging [#4619](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/842588c12ffbbe3502c5ab4a18646ad31d9c1e34&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4619 +/- ##\n=======================================\n Coverage 78.02% 78.03% \n=======================================\n Files 124 124 \n Lines 20626 20626 \n=======================================\n+ Hits 16093 16095 +2 \n+ Misses 4533 4531 -2 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4619/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.53% <0.00%> (+0.11%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4619/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=footer). Last update [842588c...8588578](https://codecov.io/gh/huggingface/transformers/pull/4619?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "@julien-c Can you please see this ?", "@sgugger Can you please see it ?" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=h1) Report\n> Merging [#4618](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/842588c12ffbbe3502c5ab4a18646ad31d9c1e34&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `50.00%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4618 +/- ##\n==========================================\n- Coverage 78.02% 78.01% -0.01% \n==========================================\n Files 124 124 \n Lines 20626 20635 +9 \n==========================================\n+ Hits 16093 16099 +6 \n- Misses 4533 4536 +3 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/hf\\_argparser.py](https://codecov.io/gh/huggingface/transformers/pull/4618/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9oZl9hcmdwYXJzZXIucHk=) | `61.11% <0.00%> (-0.87%)` | :arrow_down: |\n| [src/transformers/trainer.py](https://codecov.io/gh/huggingface/transformers/pull/4618/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90cmFpbmVyLnB5) | `39.47% <0.00%> (ø)` | |\n| [src/transformers/training\\_args.py](https://codecov.io/gh/huggingface/transformers/pull/4618/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90cmFpbmluZ19hcmdzLnB5) | `76.40% <58.33%> (-2.61%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4618/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4618/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.85% <0.00%> (+0.41%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=footer). Last update [842588c...eae844b](https://codecov.io/gh/huggingface/transformers/pull/4618?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "You should use `--evaluate_during_training` which should do mostly what you're looking for", "@prajjwal1 , you should be able to achieve this with `--evaluate_during_training` provided you set `--save_steps` to `number_of_samples/batch_size`. However, I'm currently having trouble achieving this with that option when using both `run_language_modeling.py` and `run_glue.py` as I specify in https://github.com/huggingface/transformers/issues/4630. Any ideas @julien-c ? Thanks in advance.", "There's a problem with MNLI though. In the example, arguments are changed from `mnli` to `mnli-mm`, so running evaluation after each epoch will happen on MNLI and not the mismatched one with the current implementation. " ]

[]

[ { "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-c/followers", "following_url": "https://api.github.com/users/julien-c/following{/other_user}", "gists_url": "https://api.github.com/users/julien-c/gists{/gist_id}", "starred_url": "https://api.github.com/users/julien-c/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/julien-c/subscriptions", "organizations_url": "https://api.github.com/users/julien-c/orgs", "repos_url": "https://api.github.com/users/julien-c/repos", "events_url": "https://api.github.com/users/julien-c/events{/privacy}", "received_events_url": "https://api.github.com/users/julien-c/received_events", "type": "User", "site_admin": false } ]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=h1) Report\n> Merging [#4616](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a9aa7456ac824c9027385b149f405e4f5649273f&el=desc) will **increase** coverage by `0.01%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4616 +/- ##\n==========================================\n+ Coverage 78.02% 78.04% +0.01% \n==========================================\n Files 124 124 \n Lines 20626 20626 \n==========================================\n+ Hits 16093 16097 +4 \n+ Misses 4533 4529 -4 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4616/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.53% <0.00%> (+0.11%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4616/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.83% <0.00%> (+0.32%)` | :arrow_up: |\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4616/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.85% <0.00%> (+0.41%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=footer). Last update [a9aa745...4d147b2](https://codecov.io/gh/huggingface/transformers/pull/4616?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "This works well however I noticed some discrepancy in answers generated with pipeline and without pipeline\r\n\r\nfor this example\r\n```\r\nquestion = 'Who was Jim Henson?'\r\ntext = 'Jim Henson was a nice puppet.'\r\n```\r\n\r\npipeline produces `nice puppet.`\r\n without pipeline `a nice puppet`\r\n\r\n@patrickvonplaten is this expected or there's something wrong ? ", "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=h1) Report\n> Merging [#4615](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/8cc6807e8997b8b7404c07037bd02c578da98baf&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4615 +/- ##\n==========================================\n- Coverage 78.03% 78.02% -0.01% \n==========================================\n Files 124 124 \n Lines 20647 20647 \n==========================================\n- Hits 16111 16110 -1 \n- Misses 4536 4537 +1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/data/processors/squad.py](https://codecov.io/gh/huggingface/transformers/pull/4615/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9kYXRhL3Byb2Nlc3NvcnMvc3F1YWQucHk=) | `28.66% <ø> (ø)` | |\n| [src/transformers/trainer.py](https://codecov.io/gh/huggingface/transformers/pull/4615/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90cmFpbmVyLnB5) | `39.24% <0.00%> (-0.24%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=footer). Last update [8cc6807...8d7469b](https://codecov.io/gh/huggingface/transformers/pull/4615?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "@patrickvonplaten \r\nI seem to have figured out why this is happening.\r\n\r\nThis line https://github.com/huggingface/transformers/blob/master/src/transformers/data/processors/squad.py#L103\r\ntokenizes the doc text into individual tokens and then this line \r\nhttps://github.com/huggingface/transformers/blob/master/src/transformers/data/processors/squad.py#L134\r\n\r\nuses the list of those tokens for encoding\r\n\r\nWhile this works for other BERT models , for roberta and and longformer tokenizer, the final \r\nencoding results in this\r\n`'<s> Who was Jim Henson?</s></s>JimHensonwasanicepuppet</s>'`\r\n\r\nChanging `span_doc_tokens` with `example.context_text` at L134 seems to solve the problem. But I'm not sure if doing this will cause other things to break. ", "Thanks for the PR @patil-suraj. \r\n\r\nI will put this on hold for a week though since we will most likely do some major changes very soon here.\r\n\r\n1) I think the squad preprocessing functions will probably be refactoring building on the new `nlp` library @thomwolf \r\n\r\n2) IMO, the function `squad_convert_examples_to_features` should in general not be used in the `QuestionAnsweringPipeline` since we only need `input_ids` and `attention_mask` for inference and some other values for the `score`. Also, we should also be able to use the pipeline for `TriviaQA` (with Longformer, we now have a pretrained model that works very well on TriviaQA). The pipeline should not be dataset specific. I think it might be a good idea to do a bigger refactoring of `QuestionAnsweringPipeline` and make it independent from the `squad_convert_examples_to_features` function. What do you think @julien-c @thomwolf @LysandreJik ", "Hi @patrickvonplaten , I think we should fix this now, as the newly launched model inference api uses qa pipeline its failing or giving weird answers for longformer qa models on model hub. This might discourage the users from using them. ", "@patrickvonplaten I agree with you, especially because that method is made for heavy processing of large data, which is not the case with pipelines. It's slow to start, and uses multiprocessing by default, something we don't necessarily want with the pipelines.", "Also putting @mfuntowicz in cc here", "Ok we'll have conflicts here (#5496), we need to handle with care.\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", "I believe this issue with roberta models for QA never got fixed. Any plans to continue working here? \r\n\r\n> While this works for other BERT models, for roberta and and longformer tokenizer, the final\r\nencoding results in this \r\n< s> Who was Jim Henson?</s></s>JimHensonwasanicepuppet</s>\r\n\r\nAs mentioned by @patil-suraj, we don't respect whitespace before words in the passage. Therefore, we currently use the input id for \"ĠHenson\" in the question, but the one for \"Henson\" in the passage.\r\n\r\nThe current implementation also leads to quite poor results of our QA models. For example, F1 of `deepset/roberta-base-squad2` on SQuAD 2 dev is down to 0.69 whereas it gets 0.81 with \"whitespace preserving\" tokenization.\r\n\r\nA simple fix could be to add `add_prefix_space=True` here in the tokenizer call for Roberta tokenizers (or similar), but might not be the most elegant solution. \r\nhttps://github.com/huggingface/transformers/blob/28cf873036d078b47fb9dd38ac3421a7c874da44/src/transformers/data/processors/squad.py#L112\r\n\r\nI can to do a PR for this, if that's how you want to fix it. \r\n", "A PR would be very welcome :-) ", "@patrickvonplaten Added a PR https://github.com/huggingface/transformers/pull/7387" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=h1) Report\n> Merging [#4614](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a9aa7456ac824c9027385b149f405e4f5649273f&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4614 +/- ##\n=======================================\n Coverage 78.02% 78.03% \n=======================================\n Files 124 124 \n Lines 20626 20626 \n=======================================\n+ Hits 16093 16095 +2 \n+ Misses 4533 4531 -2 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4614/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.53% <0.00%> (+0.11%)` | :arrow_up: |\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4614/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.85% <0.00%> (+0.41%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=footer). Last update [a9aa745...cbab365](https://codecov.io/gh/huggingface/transformers/pull/4614?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Good catch, thanks!" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "They are embeddings generated from the model. (Bert -Base Model I guess. cause it has a hidden representation of 768 dim). You get 9 elements:- one contextual embedding for each word in your sequence. These values of embeddings represent some hidden features that are not easy to interpret.", "So the pipeline will just return the last layer encoding of Bert?\r\nSo what is the differance with a code like\r\n\r\n```\r\ninput_ids = torch.tensor(bert_tokenizer.encode(\"Hello, my dog is cute\")).unsqueeze(0) \r\noutputs = bert_model(input_ids)\r\nhidden_states = outputs[-1][1:] # The last hidden-state is the first element of the output tuple\r\nlayer_hidden_state = hidden_states[n_layer]\r\nreturn layer_hidden_state\r\n```\r\nAlso, does BERT encoding have similar traits as word2vec? e.g. similar word will be closer, France - Paris = England - London , etc?", "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", "> So the pipeline will just return the last layer encoding of Bert?\r\n> So what is the differance with a code like\r\n> \r\n> ```\r\n> input_ids = torch.tensor(bert_tokenizer.encode(\"Hello, my dog is cute\")).unsqueeze(0) \r\n> outputs = bert_model(input_ids)\r\n> hidden_states = outputs[-1][1:] # The last hidden-state is the first element of the output tuple\r\n> layer_hidden_state = hidden_states[n_layer]\r\n> return layer_hidden_state\r\n> ```\r\n> \r\n> Also, does BERT encoding have similar traits as word2vec? e.g. similar word will be closer, France - Paris = England - London , etc?\r\n\r\nHi @orko19,\r\nDid you understand the difference from 'hidden_states' vs. 'feature-extraction pipeline'? I'd like to understand it as well\r\nThanks!", "@merleyc I do not! Please share if you do :)", "The outputs between \"last_hidden_state\" and \"feature-extraction pipeline\" are same, you can try by yourself\r\n\r\n\"feature-extraction pipeline\" just helps us do some jobs from tokenize words to embedding " ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Hi, the pipeline doesn't offer such a functionality yet. You're better off using the model directly. Here's an example of how you would replicate the pipeline's behavior, and get a token score at the end:\r\n\r\n```py\r\nfrom transformers import AutoModelWithLMHead, AutoTokenizer\r\nimport torch\r\n\r\ntokenizer = AutoTokenizer.from_pretrained(\"distilroberta-base\")\r\nmodel = AutoModelWithLMHead.from_pretrained(\"distilroberta-base\")\r\n\r\nsequence = f\"Hugging Face is a French company based in {tokenizer.mask_token}\"\r\n\r\ninput_ids = tokenizer.encode(sequence, return_tensors=\"pt\")\r\nmask_token_index = torch.where(input_ids == tokenizer.mask_token_id)[1]\r\n\r\ntoken_logits = model(input_ids)[0]\r\nmask_token_logits = token_logits[0, mask_token_index, :]\r\nmask_token_logits = torch.softmax(mask_token_logits, dim=1)\r\n\r\ntop_5 = torch.topk(mask_token_logits, 5, dim=1)\r\ntop_5_tokens = zip(top_5.indices[0].tolist(), top_5.values[0].tolist())\r\n\r\nfor token, score in top_5_tokens:\r\n print(sequence.replace(tokenizer.mask_token, tokenizer.decode([token])), f\"(score: {score})\")\r\n\r\n# Get the score of token_id\r\nsought_after_token = \"London\"\r\nsought_after_token_id = tokenizer.encode(sought_after_token, add_special_tokens=False, add_prefix_space=True)[0] # 928\r\n\r\ntoken_score = mask_token_logits[:, sought_after_token_id]\r\nprint(f\"Score of {sought_after_token}: {mask_token_logits[:, sought_after_token_id]}\")\r\n```\r\n\r\nOutputs:\r\n\r\n```\r\nHugging Face is a French company based in Paris (score: 0.2310674488544464)\r\nHugging Face is a French company based in Lyon (score: 0.08198253810405731)\r\nHugging Face is a French company based in Geneva (score: 0.04769456014037132)\r\nHugging Face is a French company based in Brussels (score: 0.047622524201869965)\r\nHugging Face is a French company based in France (score: 0.04130581393837929)\r\nScore of London: tensor([0.0343], grad_fn=<SelectBackward>)\r\n```\r\n\r\nLet me know if it helps.", "@lavanyashukla Great thanks! \r\nAnd if I want a predicability of a whole sentence, the best way will be just to average all words scores?\r\n", "Yes, that's one way to do it.", "@LysandreJik I get an error:\r\n\r\n```\r\n\"NLP_engine.py\", line 120, in _word_in_sentence_prob\r\n mask_token_index = torch.where(input_ids == bert_tokenizer.mask_token_id)[1]\r\nTypeError: where(): argument 'condition' (position 1) must be Tensor, not bool\r\n\r\n```\r\nFor the code:\r\n ```\r\ndef _word_in_sentence_prob(self, sentence, word):\r\n\r\n sequence = f\"{sentence} {bert_tokenizer.mask_token}\"\r\n\r\n input_ids = bert_tokenizer.encode(sequence, bert_tokenizer=\"pt\")\r\n mask_token_index = torch.where(input_ids == bert_tokenizer.mask_token_id)[1]\r\n\r\n token_logits = bert_model(input_ids)[0]\r\n mask_token_logits = token_logits[0, mask_token_index, :]\r\n mask_token_logits = torch.softmax(mask_token_logits, dim=1)\r\n\r\n top_5 = torch.topk(mask_token_logits, 5, dim=1)\r\n top_5_tokens = zip(top_5.indices[0].tolist(), top_5.values[0].tolist())\r\n\r\n for token, score in top_5_tokens:\r\n print(sequence.replace(bert_tokenizer.mask_token, bert_tokenizer.decode([token])), f\"(score: {score})\")\r\n\r\n # Get the score of token_id\r\n sought_after_token = word\r\n sought_after_token_id = bert_tokenizer.encode(sought_after_token, add_special_tokens=False, add_prefix_space=True)[\r\n 0] # 928\r\n\r\n token_score = mask_token_logits[:, sought_after_token_id]\r\n print(f\"Score of {sought_after_token}: {mask_token_logits[:, sought_after_token_id]}\")\r\n return token_score\r\n```\r\n\r\nAny idea why?", "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", "@LysandreJik I also get the error:\r\n```\r\n mask_token_index = torch.where(input_ids == bert_tokenizer.mask_token_id)[1]\r\nTypeError: where(): argument 'condition' (position 1) must be Tensor, not bool\r\n```\r\nfor this code.\r\nI have torch version 1.7.1\r\nAny idea what is the problem? Might it be version-related?\r\nIf so, what changes should be made in the code? Or what version should I downgrade to?", "For those still searching for a solution (after nearly 3 years), just convert the condition into a tensor:\r\n\r\n`mask_token_index = torch.where(torch.tensor(input_ids == tokenizer.mask_token_id))[1]`" ]

[]

[]

[ "Looks like your training script is out of sync with the library. Can you install the library from source, as documented in https://github.com/huggingface/transformers/tree/master/examples#important-note ?", "Thanks, @julien-c building from source, solves the issue. ", "I also find a problem about this......\r\nIf we set the lables_name =\"labels\" in the TrainerAugments, it would be wrong.\r\nBecause lables_name must be a list in TrainerAugments. If we set the labels_name = \"labels\", the function prediction_steps() in Trainer will set has_lables equal to None. For this line:\r\nhas_labels = all(inputs.get(k) is not None for k in self.label_names) in Trainer.py 1462." ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=h1) Report\n> Merging [#4610](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a9aa7456ac824c9027385b149f405e4f5649273f&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4610 +/- ##\n=======================================\n Coverage 78.02% 78.02% \n=======================================\n Files 124 124 \n Lines 20626 20626 \n=======================================\n+ Hits 16093 16094 +1 \n+ Misses 4533 4532 -1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4610/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=footer). Last update [a9aa745...3db079d](https://codecov.io/gh/huggingface/transformers/pull/4610?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Usually, the input is simply cut in this case. Bart cuts the input to 1024 tokens when training on CNN Daily Mail. T5 cuts the input to 512 tokens when training on CNN Daily Mail.", "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" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=h1) Report\n> Merging [#4608](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a9aa7456ac824c9027385b149f405e4f5649273f&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4608 +/- ##\n=======================================\n Coverage 78.02% 78.02% \n=======================================\n Files 124 124 \n Lines 20626 20626 \n=======================================\n+ Hits 16093 16094 +1 \n+ Misses 4533 4532 -1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4608/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.85% <0.00%> (+0.41%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=footer). Last update [a9aa745...470e98f](https://codecov.io/gh/huggingface/transformers/pull/4608?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=h1) Report\n> Merging [#4607](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a9aa7456ac824c9027385b149f405e4f5649273f&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4607 +/- ##\n=======================================\n Coverage 78.02% 78.02% \n=======================================\n Files 124 124 \n Lines 20626 20626 \n=======================================\n+ Hits 16093 16094 +1 \n+ Misses 4533 4532 -1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4607/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=footer). Last update [a9aa745...efcdd13](https://codecov.io/gh/huggingface/transformers/pull/4607?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Great – [model page](https://huggingface.co/dkleczek/bert-base-polish-cased-v1)" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": 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", "Are there any updates on this issue?" ]

[]

[]

[]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "Great logo!", "link seems broken on huggingface.co but I'll fix directly", "Thank you Julien!" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "That is awesome, thank you" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=h1) Report\n> Merging [#4602](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/5ddd8d6531c8c49fdd281b55b93f6c81c9826f4b&el=desc) will **increase** coverage by `0.08%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4602 +/- ##\n==========================================\n+ Coverage 78.03% 78.11% +0.08% \n==========================================\n Files 124 124 \n Lines 20647 20647 \n==========================================\n+ Hits 16111 16128 +17 \n+ Misses 4536 4519 -17 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4602/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4602/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n| [src/transformers/data/processors/squad.py](https://codecov.io/gh/huggingface/transformers/pull/4602/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9kYXRhL3Byb2Nlc3NvcnMvc3F1YWQucHk=) | `34.07% <0.00%> (+5.41%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=footer). Last update [5ddd8d6...a5049de](https://codecov.io/gh/huggingface/transformers/pull/4602?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "@blizda did you find an answer to your query - \"Only Bert can be using as encoder and decoder? If so, can you add list of available models for encoder-decoder in documentation?\"? ", "There's an error-message saying which models can be used now; \r\n\"... Model type should be one of CamembertConfig, XLMRobertaConfig, RobertaConfig, BertConfig, OpenAIGPTConfig, GPT2Config, TransfoXLConfig, XLNetConfig, XLMConfig, CTRLConfig, ReformerConfig, BertGenerationConfig, XLMProphetNetConfig, ProphetNetConfig.\"" ]

[]

[]

[]

[]

[]

[ "What is your `transformers` version? Do you have PyTorch installed?", "Thank you. I solved.", "Great to hear!", "> Thank you. I solved.\r\n\r\nHow did you solve it?", "I installed pytorch library. " ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Hi @flozi00,\r\n\r\nThanks for the PR! To be honest I don't think we should merge this. A couple of reasons:\r\n\r\n1. In contrast to `num_buckets` which is said in the paper to be always around ~ `2 * sequence_length / chunk_length` `axial_pos_shape` can be freely set by the user. \r\n2. We are trying to add as little automatic settings that are not visible to the user as possible (also @thomwolf here). The reason is that it can later lead to errors that are hard to understand for the user. In this case, I don't think the user should use AxialPositionEmbeddings before having read the docs and understood how it works. Automatically setting the `num_buckets` is already suboptimal in this sense. ", "BTW, I just answered your email - sorry I forgot about this" ]

[]

[]

[]

[]

[]

[ "I'm not sure I understand what you're trying to do. Do you mind explaining? Showing the code you're using would help as well.", "```\r\ntraining_args = dict( num_cores= 8, model_name_or_path= 't5-base',\r\n max_len= 512 ,target_max_len= 2, output_dir= './models',\r\n overwrite_output_dir= True,\r\n per_gpu_train_batch_size= 8,\r\n per_gpu_eval_batch_size= 8,\r\n gradient_accumulation_steps= 4,\r\n learning_rate= 1e-4,\r\n tpu_num_cores= 8,\r\n logging_dir='/log',\r\n do_train= True, weight_decay=0.00,\r\n device='xla',local_rank=-1,\r\n max_steps=10000, adam_epsilon=1e-8,\r\n warmup_steps=0,\r\n train_batch_size=8,\r\n eval_batch_size=8,\r\n num_train_epochs=1,\r\n early_stop_callback=False,\r\n fp_16=False, # if you want to enable 16-bit training then install apex and set this to true\r\n opt_level='O1', # you can find out more on optimisation levels here https://nvidia.github.io/apex/amp.html#opt-levels-and-properties\r\n max_grad_norm=1.0, # if you enable 16-bit training then set this to a sensible value, 0.5 is a good default\r\n seed=42, fp16=False, n_gpu=0,SummaryWriter=None)\r\n from transformers import Trainer\r\ntrainer = Trainer(\r\n model=model,\r\n args=argparse.Namespace(**training_args),\r\n train_dataset=train_dataset,\r\n data_collator=T2TDataCollator(),\r\n prediction_loss_only=True, tb_writer=None\r\n )\r\n\r\ntrainer.train()\r\n\r\n````", "here is my code and i got \r\nAttributeError: 'Namespace' object has no attribute 'to_json_string'", "Trainer's args should be a `TrainingArguments` instance, not a dict or a Namespace.\r\n\r\nTry:\r\n```python\r\nfrom transformers import TrainingArguments\r\n```" ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Hi @gustavscholin, \r\n\r\nThanks for you issue!\r\nCould you please provide a code example so that I can reproduce the error? ", "Hi, @patrickvonplaten @gustavscholin \r\nFor me, setting \"base_model_prefix\" in modeling_encoder_decoder.py fixed this problem, as finding params is based on self.base_model_prefix. \r\n\r\nIs it fundamental solution? or just short-sighted? ", "@patrickvonplaten, here's a colab notebook to reproduce the error:\r\n\r\nhttps://colab.research.google.com/drive/102U7pJJcyw__Yq0PERxAKvKPSx3bvNSi?usp=sharing", "> Hi, @patrickvonplaten @gustavscholin\r\n> For me, setting \"base_model_prefix\" in modeling_encoder_decoder.py fixed this problem, as finding params is based on self.base_model_prefix.\r\n> \r\n> Is it fundamental solution? or just short-sighted?\r\n\r\nNo that was the right solution :-) I did exactly the same in this fix: #4680", "> @patrickvonplaten, here's a colab notebook to reproduce the error:\r\n> \r\n> https://colab.research.google.com/drive/102U7pJJcyw__Yq0PERxAKvKPSx3bvNSi?usp=sharing\r\n\r\nA saved encoder-decoder model will always only be saved in a single folder. A single folder can always be loaded with `.from_pretrained()`. So to make your notebook work, you simply have to replace this line:\r\n\r\n```python \r\nsaved_model = EncoderDecoderModel.from_encoder_decoder_pretrained('test_run', 'test_run')\r\n```\r\n\r\nby \r\n\r\n```python \r\nsaved_model = EncoderDecoderModel.from_pretrained('test_run')\r\n```" ]

[]

[]

[ "hi, i tried again, and write the class of ClsNerModel to model_tf_bert.py, and then reinstall transformers again. it works!!\r\nso, if i create my class outside the lib of transformers, what else should i do to make it work, and avoid the mistakes above\r\nthanks a lot!" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=h1) Report\n> Merging [#4593](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/b86e42e0ac1b59f21f0eccf351d3346bbe3ed4eb&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `100.00%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4593 +/- ##\n=======================================\n Coverage 78.09% 78.09% \n=======================================\n Files 123 123 \n Lines 20624 20625 +1 \n=======================================\n+ Hits 16106 16108 +2 \n+ Misses 4518 4517 -1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/configuration\\_longformer.py](https://codecov.io/gh/huggingface/transformers/pull/4593/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9jb25maWd1cmF0aW9uX2xvbmdmb3JtZXIucHk=) | `100.00% <ø> (ø)` | |\n| [src/transformers/modeling\\_longformer.py](https://codecov.io/gh/huggingface/transformers/pull/4593/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19sb25nZm9ybWVyLnB5) | `97.41% <100.00%> (+<0.01%)` | :arrow_up: |\n| [src/transformers/tokenization\\_longformer.py](https://codecov.io/gh/huggingface/transformers/pull/4593/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90b2tlbml6YXRpb25fbG9uZ2Zvcm1lci5weQ==) | `100.00% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4593/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.53% <0.00%> (+0.11%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=footer). Last update [b86e42e...d8d4187](https://codecov.io/gh/huggingface/transformers/pull/4593?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "Hi, what's your transformers and pytorch version? Is that the entire code you're using?\r\n\r\nYour code doesn't crash here but crashes at the line\r\n\r\n```py\r\nanswers = nlp(question=ques0, context=abstract, topk = 10)\r\n```\r\n\r\nDo you mind providing the question and the context you're using?", "Sorry, its a mistake on my part. Thanks for your reply." ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=h1) Report\n> Merging [#4591](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/b86e42e0ac1b59f21f0eccf351d3346bbe3ed4eb&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4591 +/- ##\n=======================================\n Coverage 78.09% 78.09% \n=======================================\n Files 123 123 \n Lines 20624 20624 \n=======================================\n+ Hits 16106 16107 +1 \n+ Misses 4518 4517 -1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4591/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=footer). Last update [b86e42e...1e39227](https://codecov.io/gh/huggingface/transformers/pull/4591?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "Yes, that is a GitHub-only (GFM) feature, while we use marked.js (https://marked.js.org/) for markdown parsing.\r\n\r\nYou'll have to use the actual emojis for now" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=h1) Report\n> Merging [#4589](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/b86e42e0ac1b59f21f0eccf351d3346bbe3ed4eb&el=desc) will **not change** coverage.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4589 +/- ##\n=======================================\n Coverage 78.09% 78.09% \n=======================================\n Files 123 123 \n Lines 20624 20624 \n=======================================\n Hits 16106 16106 \n Misses 4518 4518 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_longformer.py](https://codecov.io/gh/huggingface/transformers/pull/4589/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19sb25nZm9ybWVyLnB5) | `97.40% <ø> (ø)` | |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=footer). Last update [b86e42e...368ce8e](https://codecov.io/gh/huggingface/transformers/pull/4589?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Thanks for the PR! I actually just uploaded a pretrained question answering model from allen ai and changed the docs accordingly. So I think we don't need this PR anymore ;-). See #4593" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": 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" ]

[]

[]

[]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "I got the same issue - seems to happen with the larger models (t5 small is fine)", "I can reproduce the error - will investigate :-) ", "Okey this took me quite some time to figure out...\r\n\r\nSo what happens is the following. When setting **all** modules in half as is done in the code snippet above, the following happens. At some point in line:\r\nhttps://github.com/huggingface/transformers/blob/acaa2e6267ebfda9814795fa00b6ad86c35ea5d6/src/transformers/modeling_t5.py#L188\r\nthe tensor `layer_output` contains `inf` values and then later in:\r\nhttps://github.com/huggingface/transformers/blob/acaa2e6267ebfda9814795fa00b6ad86c35ea5d6/src/transformers/modeling_t5.py#L156\r\n`nan` values enter the game... \r\n\r\nI don't really think this is a bug in T5, but it's just due to T5's rather unstable architecture. `model.half()` essentially corresponds to an apex level O3: https://nvidia.github.io/apex/amp.html#o3-fp16-training which in itself tends to become unstable...\r\n\r\nSo using your code above and using the `apex` package instead of calling `half()` on the model, you can notice the following. The code snippet which is essentially the same as yours:\r\n\r\n```python\r\nfrom transformers import T5Model\r\nfrom apex import amp\r\nimport torch\r\n\r\nmodel = T5Model.from_pretrained(\"t5-base\").cuda().eval()\r\nmodel = amp.initialize(model, opt_level=\"O3\") \r\n\r\ninputs = torch.tensor([[37,423,215,1504,13,8,1186,10670,11,10449,49,1152,11363,15465,1514,5,4433,399,7863,24766,15,17,965,594,5386,14286,28,8,6,5,755,5781,32099,993,3744,21,8,2367,18,458,53,16616,32098,16,32097,7660,16409,77,19,3,107,13164,1054,32096,993,1970,9368,948,147,8,15465,5861,87,25481,788,12,8,32095,1300,61,37,423,215,1504,13,3,24151,40,3,19668,594,5386,14286,28,8,3,115,13164]]).cuda()\r\ndecoder_input_ids = torch.tensor([[21820, 296, 55]]).cuda()\r\n\r\nout = model(input_ids=inputs, decoder_input_ids=decoder_input_ids)\r\n# encoder outputs\r\nout[2][:,:2] # nan output\r\n```\r\n\r\nyields the same output consisting of `nan` values. The same happens for `opt_level` O2. \r\nUsing the recommended O1 level of optimization:\r\n\r\n```python\r\nfrom transformers import T5Model\r\nfrom apex import amp\r\nimport torch\r\n\r\nmodel = T5Model.from_pretrained(\"t5-base\").cuda().eval()\r\nmodel = amp.initialize(model, opt_level=\"O1\") \r\n\r\ninputs = torch.tensor([[37,423,215,1504,13,8,1186,10670,11,10449,49,1152,11363,15465,1514,5,4433,399,7863,24766,15,17,965,594,5386,14286,28,8,6,5,755,5781,32099,993,3744,21,8,2367,18,458,53,16616,32098,16,32097,7660,16409,77,19,3,107,13164,1054,32096,993,1970,9368,948,147,8,15465,5861,87,25481,788,12,8,32095,1300,61,37,423,215,1504,13,3,24151,40,3,19668,594,5386,14286,28,8,3,115,13164]]).cuda()\r\ndecoder_input_ids = torch.tensor([[21820, 296, 55]]).cuda()\r\n\r\nout = model(input_ids=inputs, decoder_input_ids=decoder_input_ids)\r\n# encoder outputs\r\nout[2][:,:2] # valid output\r\n```\r\n\r\nhowever does not produce any `nan` values. As far as I know O1 is also the recommended setting: https://nvidia.github.io/apex/amp.html#o1-mixed-precision-recommended-for-typical-use .\r\nAs far as I know O1 can already greatly speed up your calculations and save quite some memory, so that I would recommend going for this.\r\n\r\nAlso pinging @mfuntowicz, @julien-c and @LysandreJik for verification", "@patrickvonplaten Even with O1 I tried fine-tuning T5-base, and in less than 100 iterations, it will converge to nan values quickly. Seems like the stability of this model is poor. Perhaps first few iterations of fine-tuning require FP32.", "~I am having issues even in fp32 with everything besides t5-small.~\r\nI am having issues in `O1` with t5-large and t5-base.\r\n\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", "Having the same issue with loss going to `nan` when fine-tuning tf-base with fp16. tf-small works fine though.", "Ran into this issue and found a workaround to get FP16 training working. \r\nT5DenseGatedGeluDense doesn't play nice with FP16, specifically the final dense layer to resize from d_ff to d_model.\r\nI used pytorch's autocast/gradscaler mixed precision implementation and created an exception for that specific dense layer.\r\n\r\n```\r\nclass T5DenseGatedGeluDense(nn.Module):\r\n def __init__(self, config):\r\n super().__init__()\r\n self.wi_0 = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wi_1 = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wo = nn.Linear(config.d_ff, config.d_model, bias=False)\r\n self.dropout = nn.Dropout(config.dropout_rate)\r\n self.gelu_act = ACT2FN[\"gelu_new\"]\r\n\r\n def forward(self, hidden_states):\r\n hidden_gelu = self.gelu_act(self.wi_0(hidden_states))\r\n hidden_linear = self.wi_1(hidden_states)\r\n hidden_states = hidden_gelu * hidden_linear\r\n hidden_states = self.dropout(hidden_states)\r\n with autocast(enabled=False):\r\n hidden_states = self.wo(hidden_states)\r\n return hidden_states\r\n```", "@leecming Have you also tried the fix with `T5DenseReluDense`?", "Great qusetion @j-min - I actually didn't find the time yet to test the \"new\" t5 model with fp16. It might very well be that the following models work fine with fp16:\r\nhttps://huggingface.co/models?search=mt5\r\nand \r\nhttps://huggingface.co/models?search=t5-v1", "@patrickvonplaten @leecming I'm trying the fix as below.\r\n```python3\r\nclass T5DenseReluDense(nn.Module):\r\n def __init__(self, config):\r\n super().__init__()\r\n self.wi = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wo = nn.Linear(config.d_ff, config.d_model, bias=False)\r\n self.dropout = nn.Dropout(config.dropout_rate)\r\n\r\n def forward(self, hidden_states):\r\n hidden_states = self.wi(hidden_states)\r\n hidden_states = F.relu(hidden_states)\r\n hidden_states = self.dropout(hidden_states)\r\n with autocast(enabled=False):\r\n hidden_states = self.wo(hidden_states)\r\n return hidden_states\r\n\r\n\r\nclass T5DenseGatedGeluDense(nn.Module):\r\n def __init__(self, config):\r\n super().__init__()\r\n self.wi_0 = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wi_1 = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wo = nn.Linear(config.d_ff, config.d_model, bias=False)\r\n self.dropout = nn.Dropout(config.dropout_rate)\r\n self.gelu_act = ACT2FN[\"gelu_new\"]\r\n\r\n def forward(self, hidden_states):\r\n hidden_gelu = self.gelu_act(self.wi_0(hidden_states))\r\n hidden_linear = self.wi_1(hidden_states)\r\n hidden_states = hidden_gelu * hidden_linear\r\n hidden_states = self.dropout(hidden_states)\r\n with autocast(enabled=False):\r\n hidden_states = self.wo(hidden_states)\r\n return hidden_states\r\n```\r\n\r\nBtw it results in the error `expected scalar type Half but found Float`, since `hidden_states` parameters are float while self.wo parameters are half.\r\nCould you please guide how I bypass the error?\r\n```python3\r\nimport torch\r\nfrom torch.cuda.amp import autocast\r\nfrom transformers import T5Model\r\n\r\nmodel = T5Model.from_pretrained(\"t5-base\").cuda().eval()\r\ninputs = torch.tensor([[37,423,215,1504,13,8,1186,10670,11,10449,49,1152,11363,15465,1514,5,4433,399,7863,24766,15,17,965,594,5386,14286,28,8,6,5,755,5781,32099,993,3744,21,8,2367,18,458,53,16616,32098,16,32097,7660,16409,77,19,3,107,13164,1054,32096,993,1970,9368,948,147,8,15465,5861,87,25481,788,12,8,32095,1300,61,37,423,215,1504,13,3,24151,40,3,19668,594,5386,14286,28,8,3,115,13164]]).cuda()\r\ndecoder_input_ids = torch.tensor([[21820, 296, 55]]).cuda()\r\n\r\nout = model(input_ids=inputs, decoder_input_ids=decoder_input_ids)\r\n# encoder outputs\r\nout[2][:,:2]\r\n\r\nwith autocast():\r\n out = model(input_ids=inputs, decoder_input_ids=decoder_input_ids)\r\n loss = out.last_hidden_state.exp().mean()\r\n```\r\n\r\n", "Oh adding `hidden_states = hidden_states.to(torch.float32)` worked, never mind.\r\nIs there a more concrete script to check if this fixes T5's fp16 training? @patrickvonplaten \r\n\r\n```python3\r\nclass T5DenseReluDense(nn.Module):\r\n def __init__(self, config):\r\n super().__init__()\r\n self.wi = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wo = nn.Linear(config.d_ff, config.d_model, bias=False)\r\n self.dropout = nn.Dropout(config.dropout_rate)\r\n\r\n def forward(self, hidden_states):\r\n hidden_states = self.wi(hidden_states)\r\n hidden_states = F.relu(hidden_states)\r\n hidden_states = self.dropout(hidden_states)\r\n with autocast(enabled=False):\r\n hidden_states = hidden_states.to(torch.float32)\r\n hidden_states = self.wo(hidden_states)\r\n return hidden_states\r\n\r\n\r\nclass T5DenseGatedGeluDense(nn.Module):\r\n def __init__(self, config):\r\n super().__init__()\r\n self.wi_0 = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wi_1 = nn.Linear(config.d_model, config.d_ff, bias=False)\r\n self.wo = nn.Linear(config.d_ff, config.d_model, bias=False)\r\n self.dropout = nn.Dropout(config.dropout_rate)\r\n self.gelu_act = ACT2FN[\"gelu_new\"]\r\n\r\n def forward(self, hidden_states):\r\n hidden_gelu = self.gelu_act(self.wi_0(hidden_states))\r\n hidden_linear = self.wi_1(hidden_states)\r\n hidden_states = hidden_gelu * hidden_linear\r\n hidden_states = self.dropout(hidden_states)\r\n with autocast(enabled=False):\r\n hidden_states = hidden_states.to(torch.float32)\r\n hidden_states = self.wo(hidden_states)\r\n return hidden_states\r\n```\r\n\r\n```python3\r\nimport torch\r\nfrom torch.cuda.amp import autocast\r\nfrom transformers import T5Model\r\n\r\nmodel = T5Model.from_pretrained(\"t5-base\").cuda().eval()\r\ninputs = torch.tensor([[37,423,215,1504,13,8,1186,10670,11,10449,49,1152,11363,15465,1514,5,4433,399,7863,24766,15,17,965,594,5386,14286,28,8,6,5,755,5781,32099,993,3744,21,8,2367,18,458,53,16616,32098,16,32097,7660,16409,77,19,3,107,13164,1054,32096,993,1970,9368,948,147,8,15465,5861,87,25481,788,12,8,32095,1300,61,37,423,215,1504,13,3,24151,40,3,19668,594,5386,14286,28,8,3,115,13164]]).cuda()\r\ndecoder_input_ids = torch.tensor([[21820, 296, 55]]).cuda()\r\n\r\nout = model(input_ids=inputs, decoder_input_ids=decoder_input_ids)\r\n# encoder outputs\r\nout[2][:,:2]\r\n\r\nwith autocast():\r\n out = model(input_ids=inputs, decoder_input_ids=decoder_input_ids)\r\n loss = out.last_hidden_state.exp().mean()\r\n\r\nprint(loss)\r\n>>> tensor(1.1017, device='cuda:0', grad_fn=<MeanBackward0>)\r\n```\r\n", "This is actually a topic I wanted to look into more closely and didn't manage to do so time-wise...maybe next week. \r\n\r\nBut in short, one should try to train a whole T5 model with your suggested fix.\r\n\r\nWhat I would recommend doing is to take your guys' fix from above and open a PR with it. Then with this PR we should fine-tune a whole t5 model on some task, *e.g.* using the Seq2SeqTrainer.\r\n\r\nE.g. one could adapt this script:https://colab.research.google.com/drive/1Ekd5pUeCX7VOrMx94_czTkwNtLN32Uyu?usp=sharing and instead of using a `Bert2Bert` model one could just use a `google/t5v1_1-small` or base model and see whether there are any problem in training. \r\n\r\nalso cc @patil-suraj in case he has better pointers/ideas", "I'll try to do a run next week though :-) ", "It’s not a good fix since it relies on a specific AMP implementation (autocast) and wouldn’t work on others (e.g., Nvidia APEX). It also uses more memory than a clean AMP implementation.\r\n\r\nA cleaner quick fix would be to copy BERT’s gradient checkpointing code and train in FP32 mode with checkpointing. \r\n\r\nAlso, Nvidia with the latest Ampere cards has started supporting bf16 which is good news.", "I am having the same issue with mt5-small getting nan with deepspeed, I really appreciate any advice on this. I am having really a hard time with it, thanks a lot \r\n@patrickvonplaten @patil-suraj @sgugger Do you mind sharing the current state of mt5 training with fp16? thanks a lot", "see: https://github.com/huggingface/transformers/issues/10830", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "anyone coming after some years, try this https://huggingface.co/google/umt5-small instead", "no luck with https://huggingface.co/google/umt5-small as well even though I was training using `FP32`", "I got into this w/ T5-3b https://huggingface.co/t5-3b/tree/main, using the more recent T5ForSequenceClassification head. I thought it was due to that newer head but now I'm seeing the issue's been more profound. \r\n\r\nI'll see what my fp32 fine-tuning gives tomorrow, as I believe no other comprehensive solution has been put into place just yet. " ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=h1) Report\n> Merging [#4585](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/3e3e552125e86824239e445dd3c659df0aea4db9&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `94.11%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4585 +/- ##\n==========================================\n- Coverage 78.09% 78.09% -0.01% \n==========================================\n Files 123 123 \n Lines 20624 20622 -2 \n==========================================\n- Hits 16106 16104 -2 \n Misses 4518 4518 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/tokenization\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4585/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy90b2tlbml6YXRpb25fdXRpbHMucHk=) | `89.84% <93.93%> (+0.32%)` | :arrow_up: |\n| [src/transformers/\\_\\_init\\_\\_.py](https://codecov.io/gh/huggingface/transformers/pull/4585/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9fX2luaXRfXy5weQ==) | `99.13% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4585/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.01% <0.00%> (-0.66%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4585/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.53% <0.00%> (+0.11%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=footer). Last update [3e3e552...7f19c32](https://codecov.io/gh/huggingface/transformers/pull/4585?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "@LysandreJik I pushed a initial version of `test_np_encode_plus_sent_to_model` which converts input to numpy tensor. \r\n\r\nFor the moment we don't have any model to forward through (JAX/Flax PR is not merged). I added a note to complete the unittests when we have the full pipeline available." ]

[]

[]

[ "Failure is `tests/test_hf_api.py::HfApiEndpointsTest::test_presign_and_upload`, which seems unrelated, so going to merge." ]

[]

[]

[ "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", "unstale", "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", "unstale", "@sshleifer I was wondering whether there is any activity on this issue. I have trained some models with MarianMT, but I am really interested in training a model from scratch with the transformers library. ", "This isn't supported by default, but is definitely possible.\r\n\r\n\r\nRough steps would be:\r\n1) Make a local directory with your intialized model and tokenizer.\r\n\r\n2) Run a command like [this](https://github.com/huggingface/transformers/blob/master/examples/research_projects/seq2seq-distillation/distil_marian_no_teacher.sh) where `$m` is the path your your local dir.\r\n\r\ncc @patil-suraj", "@sshleifer could you please repost the command that the web-page does not exist anymore?", "https://github.com/huggingface/transformers/blob/master/examples/research_projects/seq2seq-distillation/distil_marian_no_teacher.sh", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "unstale", "unstale" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=h1) Report\n> Merging [#4581](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/4c6b21805647f3a96737a50390a4c3e9463d8ef7&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4581 +/- ##\n=======================================\n Coverage 78.09% 78.09% \n=======================================\n Files 123 123 \n Lines 20617 20596 -21 \n=======================================\n- Hits 16100 16084 -16 \n+ Misses 4517 4512 -5 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/4581/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19jdHJsLnB5) | `98.64% <ø> (+0.83%)` | :arrow_up: |\n| [src/transformers/modeling\\_gpt2.py](https://codecov.io/gh/huggingface/transformers/pull/4581/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19ncHQyLnB5) | `86.21% <ø> (+0.01%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/4581/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9jdHJsLnB5) | `98.40% <ø> (ø)` | |\n| [src/transformers/modeling\\_tf\\_gpt2.py](https://codecov.io/gh/huggingface/transformers/pull/4581/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9ncHQyLnB5) | `95.66% <ø> (+0.22%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4581/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=footer). Last update [4c6b218...8350637](https://codecov.io/gh/huggingface/transformers/pull/4581?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "> LGTM, awesome!\r\n> \r\n> Should we fix some examples or pipeline accordingly?\r\n\r\nThe generation method works fine, since `prepare_input_ids` for GPT2 and CTRL only took the last input_ids anyways. So all methods relying on `generate()` are fine including the pipeline and `run_generation` examples => so we should be good!" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=h1) Report\n> Merging [#4580](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/8cc6807e8997b8b7404c07037bd02c578da98baf&el=desc) will **increase** coverage by `0.02%`.\n> The diff coverage is `92.85%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4580 +/- ##\n==========================================\n+ Coverage 78.03% 78.05% +0.02% \n==========================================\n Files 124 124 \n Lines 20647 20688 +41 \n==========================================\n+ Hits 16111 16148 +37 \n- Misses 4536 4540 +4 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/\\_\\_init\\_\\_.py](https://codecov.io/gh/huggingface/transformers/pull/4580/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9fX2luaXRfXy5weQ==) | `99.13% <ø> (ø)` | |\n| [src/transformers/modeling\\_auto.py](https://codecov.io/gh/huggingface/transformers/pull/4580/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19hdXRvLnB5) | `78.57% <ø> (ø)` | |\n| [src/transformers/modeling\\_longformer.py](https://codecov.io/gh/huggingface/transformers/pull/4580/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19sb25nZm9ybWVyLnB5) | `96.85% <92.85%> (-0.56%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4580/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.50% <0.00%> (-0.17%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=footer). Last update [8cc6807...a9afa7b](https://codecov.io/gh/huggingface/transformers/pull/4580?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "This is great. Thanks, @patil-suraj", "Is it better to use pooled output for sequence classification like in BertForSequenceClassification? @ibeltagy @patil-suraj \r\n\r\n```\r\npooled_output = outputs[1]\r\npooled_output = self.dropout(pooled_output)\r\nlogits = self.classifier(pooled_output)\r\n```", "@leslyarun \r\n`LongformerClassificationHead` does the pooling ", "> @leslyarun\r\n> `LongformerClassificationHead` does the pooling\r\n\r\nThat's great. Fine then 👍 ", "Awesome thanks @patil-suraj! \r\n\r\nMerging", "@patil-suraj Thanks for this! I'm working on a multi-task version of `LongformerForSequenceClassification`. For my context, why did you decide to implement pooling separately from the [pooling done in `LongformerModel`](https://github.com/huggingface/transformers/blob/main/src/transformers/models/longformer/modeling_longformer.py#L1377-L1383)? It seems like the key differences between the pooling done in `LongformerClassificationHead` vs. `LongformerPooler` are:\r\n\r\n1. a dropout layer before the dense layer ([source](https://github.com/huggingface/transformers/blob/9ade58f0555430cec851e307c83c3a56c4a77d0b/src/transformers/models/longformer/modeling_longformer.py#L2004))\r\n2. additional dropout and dense layers ([source](https://github.com/huggingface/transformers/blob/9ade58f0555430cec851e307c83c3a56c4a77d0b/src/transformers/models/longformer/modeling_longformer.py#L2007-L2008))\r\n\r\nI see that this mimics the [`RobertaForSequenceClassification` implementation](https://github.com/huggingface/transformers/blob/main/src/transformers/models/roberta/modeling_roberta.py#L1449-L1468). Is the goal to avoid the pooler parameters learned during pre-training a `LongformerModel`? I see that this topic has been discussed in general (https://github.com/huggingface/transformers/issues/1328), but I am curious to learn more specifically for Longformer!" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "There is a method to save tokenizer. Check this notebook: https://github.com/huggingface/blog/blob/master/notebooks/01_how_to_train.ipynb\r\n\r\n", "Thats what I am using.\r\nits saving it in the `dataset` variable not in any file. ByTokenize data I mean pretraining data.", "You can look at serialization practices, you should able to do with torch at least.\r\nhttps://huggingface.co/transformers/serialization.html#serialization-best-practices", "well thats for all required model files. I am not getting how to save pretraining data", "Once you have your data you can pickle it or use `torch.save` to save it to your disk and reload it later.", "That worked @LysandreJik thanks! \r\nI still not getting how you can prepare pretraining data on the fly while training. I got large training data and don't want to wait until it gets prepared for training.", "Have you taken a look at PyTorch's Dataset/Dataloader utilities? I recommend taking a look at [loading hude data functionality](https://discuss.pytorch.org/t/loading-huge-data-functionality/346) or [how to use a dataset larger than memory](https://discuss.pytorch.org/t/how-to-use-dataset-larger-than-memory/37785/8).\r\n\r\nI personnally prefer using IterableDatasets when loading large files, as I find the API easier to use to limit large memory usage. This [tutorial](https://medium.com/swlh/how-to-use-pytorch-dataloaders-to-work-with-enormously-large-text-files-bbd672e955a0) is interesting on that subject.", "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" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=h1) Report\n> Merging [#4578](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4578 +/- ##\n==========================================\n- Coverage 77.87% 77.86% -0.01% \n==========================================\n Files 123 123 \n Lines 20566 20566 \n==========================================\n- Hits 16016 16014 -2 \n- Misses 4550 4552 +2 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4578/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4578/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.41% <0.00%> (-0.12%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4578/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (ø)` | |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=footer). Last update [a34a989...03b9ffc](https://codecov.io/gh/huggingface/transformers/pull/4578?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "I think it's not implemented yet.\r\n\r\n@julien-c any suggestion/thoughts for pretraining with wwm?", "NVIDIA/Megatron-LM does wwm on the fly in __ getitem __\r\n\r\nWe can do something similar in DataCollatorForLanguageModeling or in the dataset \r\n\r\nhttps://github.com/NVIDIA/Megatron-LM/blob/22c0e300670672e4e0a8604bd6ab89bc28c970a6/megatron/data/bert_dataset.py#L148", "Thanks for the suggestion, I'll look into it.", "@usuyama The Megatron example is for the BERT dataset which uses wordpiece tokenization. Any suggestions how to do wwm for GPT2 tokenizer?", "related #6491", "Check if still looking for an answer:\r\nhttps://github.com/huggingface/transformers/blob/07708793f20ec3a949ccab32cc4fe0c7272dcc4c/src/transformers/data/data_collator.py#L301" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": 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" ]

[]

[]

[ "You can try newer version of PyTorch (like 1.3~1.5). The problem shall be resolved.", "Hi thanks for the reply\r\nI tried with pytorch 1.4 but I got another error on the below\r\n\r\ndo you have any idea about this one?\r\nthanks!\r\n", "@amy-hyunji, this option (use_external_data_format) need PyTorch 1.5. This option is not needed for model < 2GB.\r\n\r\nIf you do not want to upgrade to PyTorch 1.5. You can install transformers from source, and modify the convert_graph_to_onnx.py (by removing the parameter during calling onnx.export function).", "@tianleiwu Thanks a lot :)" ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=h1) Report\n> Merging [#4574](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/b86e42e0ac1b59f21f0eccf351d3346bbe3ed4eb&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `100.00%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4574 +/- ##\n==========================================\n- Coverage 78.09% 78.08% -0.01% \n==========================================\n Files 123 123 \n Lines 20624 20624 \n==========================================\n- Hits 16106 16105 -1 \n- Misses 4518 4519 +1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_longformer.py](https://codecov.io/gh/huggingface/transformers/pull/4574/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19sb25nZm9ybWVyLnB5) | `97.40% <100.00%> (ø)` | |\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4574/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=footer). Last update [b86e42e...8528090](https://codecov.io/gh/huggingface/transformers/pull/4574?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[]

[]

[ "I ended up using the vanilla way to train", "I think compute_metrics should return a dictionary string to metric values. That is how it is written in the docstring of the train function" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": 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" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "Hello !\r\n\r\nWhich version of the lib do you use?", "Hello!\r\n\r\nI think it's a problem with your GPU rather than from transformers. Check your TensorFlow whether \"Failed to load the native TensorFlow runtime.\" appears once it is imported.", "That's probably because you don't have `tensorflow>=2.0` installed while you're trying to load a TensorFlow 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" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=h1) Report\n> Merging [#4570](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4570 +/- ##\n==========================================\n- Coverage 77.87% 77.87% -0.01% \n==========================================\n Files 123 123 \n Lines 20566 20566 \n==========================================\n- Hits 16016 16015 -1 \n- Misses 4550 4551 +1 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4570/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4570/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.41% <0.00%> (-0.12%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4570/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.83% <0.00%> (+0.16%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=footer). Last update [a34a989...78cf772](https://codecov.io/gh/huggingface/transformers/pull/4570?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": 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" ]

[]

[ { "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/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "Good catch. There is no mask token in the second checkpoint. I believe that is the same as the authors' implementation.\r\n\r\nCompletely off topic: if you still have the xsum data you used I would love a copy. I'm sam [at] huggingface.co . ", "Thanks for your fast answer ! \r\n\r\nDo you know why there is no mask token in the second checkpoint ? And if it has any impact on score ?", "I have a hunch the there is no `<mask>` token because of fairseq's `--find-unused-parameters` clarg, but I'm not certain.\r\n\r\nI would guess no impact on score because `<mask>` does not show up in the finetuning data.\r\n" ]

[]

[ { "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/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "Great Q! The point of using them is that `MarianMTModel`, which inherits from `BartForConditionalGeneration` uses them. You're correct for the bart checkpoints they stay 0. If you think there is a comment or different approach that would be clearer, I'm very open to PR/other ideas." ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=h1) Report\n> Merging [#4566](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `100.00%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4566 +/- ##\n=======================================\n Coverage 77.87% 77.87% \n=======================================\n Files 123 123 \n Lines 20566 20569 +3 \n=======================================\n+ Hits 16016 16019 +3 \n Misses 4550 4550 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_albert.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19hbGJlcnQucHk=) | `77.20% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19jdHJsLnB5) | `97.82% <100.00%> (+<0.01%)` | :arrow_up: |\n| [src/transformers/modeling\\_gpt2.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19ncHQyLnB5) | `86.25% <100.00%> (+0.04%)` | :arrow_up: |\n| [src/transformers/modeling\\_openai.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19vcGVuYWkucHk=) | `81.84% <100.00%> (+0.06%)` | :arrow_up: |\n| [src/transformers/modeling\\_tf\\_albert.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9hbGJlcnQucHk=) | `78.74% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_bert.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9iZXJ0LnB5) | `96.54% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_ctrl.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9jdHJsLnB5) | `98.40% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_electra.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9lbGVjdHJhLnB5) | `99.06% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_gpt2.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9ncHQyLnB5) | `95.43% <100.00%> (ø)` | |\n| [src/transformers/modeling\\_tf\\_openai.py](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9vcGVuYWkucHk=) | `95.86% <100.00%> (ø)` | |\n| ... and [3 more](https://codecov.io/gh/huggingface/transformers/pull/4566/diff?src=pr&el=tree-more) | |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=footer). Last update [a34a989...e31fdfa](https://codecov.io/gh/huggingface/transformers/pull/4566?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Hi, I'm not sure we want to change this. I do agree that `adder` is more explicit than `attention_mask` and better models what it does. However, imo the API isn't limited to high-level modules but even to lower level modules such as `AlbertAttention` and `AlbertTransformer`.\r\n\r\nThese modules may be used by users in their specific applications: `from transformers.modeling_albert import AlbertTransformer`.\r\n\r\nI think the gain here is not worth the loss of compatibility with all previous versions. What do you think @patrickvonplaten, @thomwolf, @julien-c ?", "Yes, I agree...sorry @NSanjay I didn't think this fully through when answering here: https://github.com/huggingface/transformers/issues/4141#issuecomment-629875496 :-/ ", "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" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 2052904485, "node_id": "MDU6TGFiZWwyMDUyOTA0NDg1", "url": "https://api.github.com/repos/huggingface/transformers/labels/reformer", "name": "reformer", "color": "5319e7", "default": false, "description": "Everything related to the reformer model" } ]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "I would not recommend to set `axial_pos_shape` to (512 * 1024). In the notebook I just used that to demonstrate how far the limits can be pushed for Reformer. Half a million token is extremely long and usually unnecessary. \r\n\r\nMake sure you have read and understood how AxialPostionEmbeddings work: https://huggingface.co/transformers/model_doc/reformer.html#axial-positional-encodings . \r\n\r\nFor \"normal\" language modeling it might make much more sense to start from the Reformer-wiken8 model and finetune it: https://huggingface.co/google/reformer-enwik8", "Greetings,\r\nWould fine tuning https://huggingface.co/google/reformer-enwik8 work normally with run_language_modeling.py script?\r\nThanks", "Hmm, for the most part but you will have to define your own tokenzer function as can be seen here: https://huggingface.co/google/reformer-enwik8#reformer-language-model-on-character-level-and-trained-on-enwik8\r\n", "So instead of sticking to the script, I would recommend slightly changing this notebook: https://github.com/patrickvonplaten/notebooks/blob/master/PyTorch_Reformer.ipynb. Instead of creating the dataset by using a tokenizer, you should use the function linked above. Does that make sense? Also linking: https://github.com/huggingface/transformers/pull/4480. If someone has an easy script for Reformer Char LM it'd be great to post it here or add a notebook. ", "Ok, thanks. So the function _flatten_and_tokenize_ (in the notebook) shall be replaced by the _encode_\r\nfunction (in the enwik8 model card), am I following right?", "> I would not recommend to set `axial_pos_shape` to (512 * 1024). In the notebook I just used that to demonstrate how far the limits can be pushed for Reformer. Half a million token is extremely long and usually unnecessary.\r\n> \r\nI've been using 'google/reformer-crime-and-punishment' model from [https://huggingface.co/transformers/model_doc/reformer.html#reformermodelwithlmhead](url)\r\n\r\nI get this error after I padded the sequence lengths to be a multiple of least common multiple chunk_length 64.\r\n\r\n```\r\n...\r\nfor epoch in range(EPOCHS):\r\n print(f\"EPOCH {epoch} started\" + '=' * 30)\r\n for idx,article in tqdm_notebook(enumerate(article_loader)):\r\n \r\n article_tens = tokenizer.encode(article[0], return_tensors='pt').to(device)\r\n \r\n print(article_tens.shape)\r\n #multiple of least common multiple chunk_length 64.\r\n pads_to_be_filled=getNoOfPads(article_tens.size()[1])\r\n \r\n padded_tens= torch.cat((article_tens[0],Variable(torch.zeros((pads_to_be_filled),dtype=torch.long).cuda())) )\r\n \r\n print(padded_tens.unsqueeze(0).shape)\r\n\r\n outputs = model(padded_tens.unsqueeze(0), labels=padded_tens.unsqueeze(0))[0]\r\n ...\r\n```\r\n```\r\n\r\nEPOCH 0 started==============================\r\n0/? [00:00<?, ?it/s]\r\ntorch.Size([1, 131])\r\ntorch.Size([1, 192])\r\n\r\n---------------------------------------------------------------------------\r\nAssertionError Traceback (most recent call last)\r\n<ipython-input-11-81c445097515> in <module>()\r\n 29 print(padded_tens.unsqueeze(0).shape)\r\n 30 \r\n---> 31 outputs = model(padded_tens.unsqueeze(0), labels=padded_tens.unsqueeze(0))[0]\r\n 32 print(outputs)\r\n 33 \r\n\r\n7 frames\r\n/usr/local/lib/python3.6/dist-packages/transformers/modeling_reformer.py in forward(self, position_ids)\r\n 127 reduce(mul, self.axial_pos_shape) == sequence_length\r\n 128 ), \"If training, make sure that config.axial_pos_shape factors: {} multiply to sequence length. Got prod({}) != sequence_length: {}. You might want to consider padding your sequence length to {} or changing config.axial_pos_shape.\".format(\r\n--> 129 self.axial_pos_shape, self.axial_pos_shape, sequence_length, reduce(mul, self.axial_pos_shape)\r\n 130 )\r\n 131 if self.dropout > 0:\r\n\r\nAssertionError: If training, make sure that config.axial_pos_shape factors: (512, 1024) multiply to sequence length. Got prod((512, 1024)) != sequence_length: 192. You might want to consider padding your sequence length to 524288 or changing config.axial_pos_shape.\r\n\r\n```\r\n\r\n>If training, make sure that config.axial_pos_shape factors: (512, 1024) multiply to sequence length. Got prod((512, 1024)) != sequence_length: 384. You might want to consider padding your sequence length to 524288 or changing config.axial_pos_shape.\r\n\r\nSo I guess that is because its default set to (512, 1024), and if so, how can I change it to a smaller value? \r\n\r\nReformerConfig {\r\n \"architectures\": [\r\n \"ReformerModelWithLMHead\"\r\n ],\r\n \"attention_head_size\": 64,\r\n \"attention_probs_dropout_prob\": 0.1,\r\n \"attn_layers\": [\r\n \"local\",\r\n \"lsh\",\r\n \"local\",\r\n \"lsh\",\r\n \"local\",\r\n \"lsh\"\r\n ],\r\n \"axial_norm_std\": 1.0,\r\n \"axial_pos_embds\": true,\r\n \"axial_pos_embds_dim\": [\r\n 64,\r\n 192\r\n ],\r\n \"axial_pos_shape\": [\r\n 512,\r\n 1024\r\n ],\r\n \"chunk_size_feed_forward\": 0,\r\n \"chunk_size_lm_head\": 0,\r\n \"eos_token_id\": 2,\r\n \"feed_forward_size\": 512,\r\n \"hash_seed\": null,\r\n \"hidden_act\": \"relu\",\r\n \"hidden_dropout_prob\": 0.05,\r\n \"hidden_size\": 256,\r\n \"initializer_range\": 0.02,\r\n \"intermediate_size\": 3072,\r\n \"is_decoder\": true,\r\n \"layer_norm_eps\": 1e-12,\r\n \"local_attention_probs_dropout_prob\": 0.05,\r\n \"local_attn_chunk_length\": 64,\r\n \"local_num_chunks_after\": 0,\r\n \"local_num_chunks_before\": 1,\r\n \"lsh_attention_probs_dropout_prob\": 0.0,\r\n \"lsh_attn_chunk_length\": 64,\r\n \"lsh_num_chunks_after\": 0,\r\n \"lsh_num_chunks_before\": 1,\r\n \"max_position_embeddings\": 524288,\r\n \"model_type\": \"reformer\",\r\n \"num_attention_heads\": 2,\r\n \"num_buckets\": [\r\n 64,\r\n 128\r\n ],\r\n \"num_chunks_after\": 0,\r\n \"num_chunks_before\": 1,\r\n \"num_hashes\": 1,\r\n \"num_hidden_layers\": 6,\r\n \"output_past\": true,\r\n \"pad_token_id\": 0,\r\n \"task_specific_params\": {\r\n \"text-generation\": {\r\n \"do_sample\": true,\r\n \"max_length\": 100\r\n }\r\n },\r\n \"vocab_size\": 320\r\n}\r\n\r\nGiven above is the default configuration of the model before training/finetuning\r\n\r\n> For \"normal\" language modeling it might make much more sense to start from the Reformer-wiken8 model and finetune it: https://huggingface.co/google/reformer-enwik8\r\n\r\nWill try that too.\r\n\r\nThank you.\r\n\r\n", "yeah the google/crime-and-punishment is not a good model for fine-tuning. It assumes you use a sequence length of > 500K tokens, which is not really reasonable.", "> Ok, thanks. So the function _flatten_and_tokenize_ (in the notebook) shall be replaced by the _encode_\r\n> function (in the enwik8 model card), am I following right?\r\n\r\nexactly. You should be able to just enwik8 function I linked above. The enwik8 model has a maximum length of ~65K tokens, which is very long but very feasible for reformer.", "> yeah the google/crime-and-punishment is not a good model for fine-tuning. It assumes you use a sequence length of > 500K tokens, which is not really reasonable.\r\n\r\nOh okay. Thank you very much for the clarification. Will try finetuning reformer-enwik8.", "It would be awesome if you could upload your training script here - people seem very interested in it :-) ", "\r\n @patrickvonplaten, Sure, will do when everything is sorted.", "> \r\n> \r\n> > Ok, thanks. So the function _flatten_and_tokenize_ (in the notebook) shall be replaced by the _encode_\r\n> > function (in the enwik8 model card), am I following right?\r\n> \r\n> exactly. You should be able to just enwik8 function I linked above. The enwik8 model has a maximum length of ~65K tokens, which is very long but very feasible for reformer.\r\n\r\nFrom the notebook am struggling to adapt the DataCollator, how to define it properly in this context? \r\nThanks", "Someone effectivelly fine tune on enwiki8 pre trained model? using colab with P100 gpu i was not able to load model yet due to memory limitation", "Unfortunately facing the same issue now.", "Can you add a link to your notebook here @lucashueda ?", "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", "@lucashueda Did you manage to fine-tune enwiki8 pre trained model ? or other datasets ? Would you mind sharing your Colab? ", "@epetros did you manage to perform the fine-tuning ? ", "@patrickvonplaten any update on this? \r\n\r\nOs is there a notebook where we can pre-train this model from wiki or huge corpus ourselves and then fine-tune it to downstream tasks? ", "Hi, I'm trying to fine-tune the ReformerModelWithLMHead (google/reformer-enwik8) for NER. I used the padding sequence length same as in the encode method (max_length = max([len(string) for string in list_of_strings])) along with attention_masks. And I got this error: \r\n\r\n**ValueError:** If training, make sure that config.axial_pos_shape factors: (128, 512) multiply to sequence length. Got prod((128, 512)) != sequence_length: 2248. You might want to consider padding your sequence length to 65536 or changing config.axial_pos_shape.\r\n\r\n1) When I changed the sequence length to 65536, my colab session crashed by getting all the inputs of 65536 lengths. \r\n2) According to the second option(changing config.axial_pos_shape), I cannot change it. \r\n\r\nI would like to know, Is there any chance to change config.axial_pos_shape while fine-tuning the model? Or I'm missing something in encoding the input strings for reformer-enwik8? Are there any additional steps to forward the input to the model after encoding?\r\n\r\nThanks!\r\n\r\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=h1) Report\n> Merging [#4564](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **decrease** coverage by `0.01%`.\n> The diff coverage is `14.28%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4564 +/- ##\n==========================================\n- Coverage 77.87% 77.86% -0.02% \n==========================================\n Files 123 123 \n Lines 20566 20569 +3 \n==========================================\n Hits 16016 16016 \n- Misses 4550 4553 +3 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/configuration\\_reformer.py](https://codecov.io/gh/huggingface/transformers/pull/4564/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9jb25maWd1cmF0aW9uX3JlZm9ybWVyLnB5) | `100.00% <ø> (ø)` | |\n| [src/transformers/modeling\\_reformer.py](https://codecov.io/gh/huggingface/transformers/pull/4564/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ19yZWZvcm1lci5weQ==) | `87.94% <14.28%> (-0.24%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4564/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.41% <0.00%> (-0.12%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4564/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (ø)` | |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=footer). Last update [a34a989...7b40493](https://codecov.io/gh/huggingface/transformers/pull/4564?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null } ]

[]

[ "So I can get the generation working well with distilgpt2, the thing is that I would like to do it multilingual using the light multilingual model DistilmBERT (distilbert-base-multilingual-cased), any tips? thank you :)\r\n\r\n```py\r\nimport torch\r\nfrom transformers import *\r\nfrom transformers import TFGPT2LMHeadModel, GPT2Tokenizer\r\n\r\ntokenizer = GPT2Tokenizer.from_pretrained('gpt2')\r\ninput_ids = torch.tensor(tokenizer.encode(\"Once upon a time\")).unsqueeze(0)\r\nmodel = GPT2LMHeadModel.from_pretrained(\"distilgpt2\", pad_token_id=tokenizer.eos_token_id)\r\ngreedy_output = model.generate(input_ids, max_length=50) #greedy search\r\n\r\nsample_outputs = model.generate(\r\n input_ids,\r\n do_sample=True, \r\n max_length=50, \r\n top_k=50, \r\n top_p=0.95, \r\n temperature=1,\r\n num_return_sequences=3\r\n)\r\n\r\nprint(\"Output:\\n\" + 100 * '-')\r\nfor i, sample_output in enumerate(sample_outputs):\r\n print(\"{}: {}\".format(i, tokenizer.decode(sample_output, skip_special_tokens=True)))\r\n```", "Hi, I took the liberty of editing your comments with triple backticks ```py\\`\\`\\` to be more readable.\r\n\r\nUnfortunately DistilmBERT can't be used for generation. This is due to the way the original BERT models were pre-trained, using masked language modeling (MLM). It therefore attends to both the left and right contexts (tokens on the left and right of the token you're trying to generate), while for generation the model only has access to the left context.\r\n\r\nGPT-2 was trained with causal language modeling (CLM), which is why it can generate such coherent sequences. We implement the `generation` method only for CLM models, as MLM models do not generate anything coherent. ", "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" ]

[]

[ { "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/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "@sshleifer might be able to help you here. @sshleifer - hope it's fine that I link you here :-) ", "This might also help: https://github.com/huggingface/transformers/pull/4539/files?short_path=3a2ba7b#diff-3a2ba7b492f00029d14cec3994b73ac7", "> This might also help: https://github.com/huggingface/transformers/pull/4539/files?short_path=3a2ba7b#diff-3a2ba7b492f00029d14cec3994b73ac7\r\n\r\nThank you very much. I will look and try to implement and let you know about the result!", "> This might also help: https://github.com/huggingface/transformers/pull/4539/files?short_path=3a2ba7b#diff-3a2ba7b492f00029d14cec3994b73ac7\r\n\r\nit seems working! thank you!" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=h1) Report\n> Merging [#4561](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **decrease** coverage by `0.01%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4561 +/- ##\n==========================================\n- Coverage 77.87% 77.86% -0.02% \n==========================================\n Files 123 123 \n Lines 20566 20566 \n==========================================\n- Hits 16016 16013 -3 \n- Misses 4550 4553 +3 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4561/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4561/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.50% <0.00%> (-0.17%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4561/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.41% <0.00%> (-0.12%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=footer). Last update [a34a989...b44c742](https://codecov.io/gh/huggingface/transformers/pull/4561?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Can you add the do_lower_case flag for all other instances in that README that use bert uncased? Thanks!", "Sure. Presumably for the tf version, the script will handle the casings automatically.", "From the code it would seem that way, but I am not sure actually. \r\n\r\ncc @thomwolf @LysandreJik: `do_lower_case` was missing in the commands to run squad with bert-base-uncased. Is this flag also necessary in the Tensorflow version? It's not present in the code, so I would assume not.\r\n\r\nThe other changes LGTM!", "Closed by #4245 (we still need to investigate why the lowercasing is not properly populated by the model's config)" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1834056635, "node_id": "MDU6TGFiZWwxODM0MDU2NjM1", "url": "https://api.github.com/repos/huggingface/transformers/labels/Core:%20Tokenization", "name": "Core: Tokenization", "color": "FF4446", "default": false, "description": "Internals of the library; Tokenization." } ]

[ { "login": "mfuntowicz", "id": 2241520, "node_id": "MDQ6VXNlcjIyNDE1MjA=", "avatar_url": "https://avatars.githubusercontent.com/u/2241520?v=4", "gravatar_id": "", "url": "https://api.github.com/users/mfuntowicz", "html_url": "https://github.com/mfuntowicz", "followers_url": "https://api.github.com/users/mfuntowicz/followers", "following_url": "https://api.github.com/users/mfuntowicz/following{/other_user}", "gists_url": "https://api.github.com/users/mfuntowicz/gists{/gist_id}", "starred_url": "https://api.github.com/users/mfuntowicz/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/mfuntowicz/subscriptions", "organizations_url": "https://api.github.com/users/mfuntowicz/orgs", "repos_url": "https://api.github.com/users/mfuntowicz/repos", "events_url": "https://api.github.com/users/mfuntowicz/events{/privacy}", "received_events_url": "https://api.github.com/users/mfuntowicz/received_events", "type": "User", "site_admin": false } ]

[ "Does this also happen when you use the slow tokenizers?\r\n\r\n```python\r\ntokenizer = AlbertTokenizer.from_pretrained(\"albert-base-v2\", use_fast=False)\r\n```", "Thanks for the suggestion @BramVanroy . I just tried this and the tokenizer has now been running for 70 minutes, so I think that's a yes, it also happens when I use slow mode.", "cc @mfuntowicz AlbertTokenizer seems to hang in fast mode but not in slow mode", "Hi, there is no fast mode for `AlbertTokenizer`, it's a SentencePiece based tokenizer which is not currently supported by `tokenizers`.\r\n\r\nDo you think you can find more information about where the tokenizer actually hang?\r\nCan you reproduce the behavior with a shorter input?", "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 recommend to change your preproccessing to:\r\n\r\n```python\r\nfrom transformers import XLNetTokenizer \r\nfrom keras.preprocessing.sequence import pad_sequences\r\ntokenizer = XLNetTokenizer.from_pretrained('xlnet-base-cased')\r\n\r\nencoded_sent = tokenizer.encode_plus(\r\n text, # Sentence to encode.\r\n return_tensors='pt',\r\n max_length=128)\r\n# ...\r\noutput = model(**encoded_sent )\r\n```\r\n\r\nHowever, the real problem is probably due to hyperparameters. You cannot simply use different models with the same hyperparameters and immediately expect results. You'll have to fiddle with the hyperparameters and find something that works for your case. \r\n", "I am converting the text to tensor in a later step.\r\nI also tried changing the lr ranging from 0.05 to 5e-8 but still loss did not changing and also applied a lr sheduler. Maybe I should try other optimizers other than AdamW?", "Since this is not a bug I am closing this. It is impossible for us to help with this any further since hyperparamter optimization is different for each task. This is something that you have to test yourself.\r\n\r\nFor a starting point, you can have a look at Table 8 in [the original paper](https://arxiv.org/pdf/1906.08237.pdf) where they suggest some good hyperparameter settings. But again, even then it depends on your specific case what would help and what wouldn't. Try and test!" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=h1) Report\n> Merging [#4558](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **decrease** coverage by `0.00%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4558 +/- ##\n==========================================\n- Coverage 77.87% 77.86% -0.01% \n==========================================\n Files 123 123 \n Lines 20566 20566 \n==========================================\n- Hits 16016 16014 -2 \n- Misses 4550 4552 +2 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4558/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4558/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.41% <0.00%> (-0.12%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4558/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.66% <0.00%> (ø)` | |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=footer). Last update [a34a989...ce96482](https://codecov.io/gh/huggingface/transformers/pull/4558?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Thanks! That's correct\r\n" ]

[]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=h1) Report\n> Merging [#4557](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **increase** coverage by `0.00%`.\n> The diff coverage is `86.95%`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4557 +/- ##\n=======================================\n Coverage 77.87% 77.87% \n=======================================\n Files 123 123 \n Lines 20566 20580 +14 \n=======================================\n+ Hits 16016 16027 +11 \n- Misses 4550 4553 +3 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_pytorch\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4557/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl9weXRvcmNoX3V0aWxzLnB5) | `89.30% <78.57%> (-0.63%)` | :arrow_down: |\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4557/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `88.70% <100.00%> (+0.03%)` | :arrow_up: |\n| [src/transformers/modeling\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4557/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ191dGlscy5weQ==) | `90.56% <100.00%> (+0.02%)` | :arrow_up: |\n| [src/transformers/file\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4557/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9maWxlX3V0aWxzLnB5) | `73.44% <0.00%> (-0.42%)` | :arrow_down: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=footer). Last update [a34a989...0323876](https://codecov.io/gh/huggingface/transformers/pull/4557?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n", "Good stuff!" ]

[ { "id": 1838412367, "node_id": "MDU6TGFiZWwxODM4NDEyMzY3", "url": "https://api.github.com/repos/huggingface/transformers/labels/model%20card", "name": "model card", "color": "92d5f4", "default": false, "description": "Related to pretrained model cards" } ]

[]

[ "# [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=h1) Report\n> Merging [#4556](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=desc) into [master](https://codecov.io/gh/huggingface/transformers/commit/a34a9896ac2a4a33ff9cd805c76eed914c8d8965&el=desc) will **increase** coverage by `0.05%`.\n> The diff coverage is `n/a`.\n\n[](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=tree)\n\n```diff\n@@ Coverage Diff @@\n## master #4556 +/- ##\n==========================================\n+ Coverage 77.87% 77.93% +0.05% \n==========================================\n Files 123 123 \n Lines 20566 20566 \n==========================================\n+ Hits 16016 16028 +12 \n+ Misses 4550 4538 -12 \n```\n\n\n| [Impacted Files](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=tree) | Coverage Δ | |\n|---|---|---|\n| [src/transformers/modeling\\_tf\\_utils.py](https://codecov.io/gh/huggingface/transformers/pull/4556/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9tb2RlbGluZ190Zl91dGlscy5weQ==) | `87.84% <0.00%> (-0.83%)` | :arrow_down: |\n| [src/transformers/data/processors/squad.py](https://codecov.io/gh/huggingface/transformers/pull/4556/diff?src=pr&el=tree#diff-c3JjL3RyYW5zZm9ybWVycy9kYXRhL3Byb2Nlc3NvcnMvc3F1YWQucHk=) | `34.07% <0.00%> (+5.41%)` | :arrow_up: |\n\n------\n\n[Continue to review full report at Codecov](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=continue).\n> **Legend** - [Click here to learn more](https://docs.codecov.io/docs/codecov-delta)\n> `Δ = absolute <relative> (impact)`, `ø = not affected`, `? = missing data`\n> Powered by [Codecov](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=footer). Last update [a34a989...433d479](https://codecov.io/gh/huggingface/transformers/pull/4556?src=pr&el=lastupdated). Read the [comment docs](https://docs.codecov.io/docs/pull-request-comments).\n" ]

[ { "id": 1834052333, "node_id": "MDU6TGFiZWwxODM0MDUyMzMz", "url": "https://api.github.com/repos/huggingface/transformers/labels/Ex:%20Question%20Answering", "name": "Ex: Question Answering", "color": "86FFCF", "default": false, "description": "" } ]

[]

[ "My results are different as well: \r\n\r\n\"exact_match\": 71.92999053926206\r\n\"f1\": 80.70949484221217\r\n\r\nMy guess is that this occurs because we are not using a fixed seed. The runs are not deterministic so difference _will_ occur.", "Possibly but the difference of 7~8 points in f1 and EM scores is way above the usual variance due to random seeds.", "Found the bug. `--do_lower_case` was missing in the script arguments.\r\n\r\nNow the results are pretty close to the ones mentioned in the tutorial.\r\n\r\n05/24/2020 23:50:04 - INFO - __main__ - Results: {'exact': 80.26490066225166, 'f1': 88.01726518927101, 'total': 10570, 'HasAns_exact': 80.26490066225166, 'HasAns_f1': 88.01726518927101, 'HasAns_total': 10570, 'best_exact': 80.26490066225166, 'best_exact_thresh': 0.0, 'best_f1': 88.01726518927101, 'best_f1_thresh': 0.0}", "> Possibly but the difference of 7~8 points in f1 and EM scores is way above the usual variance due to random seeds.\r\n\r\nUnfortunately not. Have a look at these experiments by my friends over at NLP Town. They did sentiment analyses and ran the experiments ten times (each time with a different seed). https://www.linkedin.com/posts/nlp-town_sentimentanalysis-camembert-xlm-activity-6605379961111007232-KJy3 \r\n\r\nThat being said, I do think you are right, good catch! ", "Closing this b/c #4245 was merged\r\n\r\n(we still need to investigate why the lowercasing is not properly populated by the model's config)\r\n" ]

[]

[ { "login": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false } ]

[ "Hey @drjosephliu - thanks so much for opening this PR! \r\nSorry for being lazy here - could you check out the comments I added in PR #4538 - I think they apply 1-to-1 the same way here.", "No problem. I'm a bit busy this week, but will aim to get it done by the end of the week.", "> No problem. I'm a bit busy this week, but will aim to get it done by the end of the week.\r\n\r\nSure, take your time :-) ", "I'm just in the middle of fixing up the tests and I've noticed that the `ReformerModel` has a different method signature than other models because it takes in `do_output_hidden_states`:\r\n\r\n```\r\n def forward(\r\n self,\r\n input_ids=None,\r\n attention_mask=None,\r\n position_ids=None,\r\n head_mask=None,\r\n inputs_embeds=None,\r\n num_hashes=None,\r\n do_output_hidden_states=False,\r\n do_output_attentions=False,\r\n ):\r\n```\r\n\r\nShould I be converting it to `output_hidden_states` here?", "> I'm just in the middle of fixing up the tests and I've noticed that the `ReformerModel` has a different method signature than other models because it takes in `do_output_hidden_states`:\r\n> \r\n> ```\r\n> def forward(\r\n> self,\r\n> input_ids=None,\r\n> attention_mask=None,\r\n> position_ids=None,\r\n> head_mask=None,\r\n> inputs_embeds=None,\r\n> num_hashes=None,\r\n> do_output_hidden_states=False,\r\n> do_output_attentions=False,\r\n> ):\r\n> ```\r\n> \r\n> Should I be converting it to `output_hidden_states` here?\r\n\r\nyes please!", "Before starting on the TF implementation, it might be a good idea how it is handled in `modeling_tf_bert.py` of PR: #4538 :-) ", "So all pytorch tests are passing, but I still haven't really figured out the TF ones. I copied what you did for `output_attention_heads` applied to `output_hidden_states` to TF Bert almost verbatim, but I'm still getting some failing tests", "Hey @drjosephliu,\r\n\r\nThis looks great already :-) We will probably have a lot of conflicts with master once #4538 is merged (I should have thought about this before ... ). Would it be ok for you to wait a couple of days on this PR until we merge #4538. Then I can rebase this PR to master and it will be easier to work from then on :-) \r\n\r\nCan you do the following change to the branch `hidden_states` of your fork, so that I can commit directly to your branch? :-)\r\n\r\nhttps://help.github.com/en/github/collaborating-with-issues-and-pull-requests/allowing-changes-to-a-pull-request-branch-created-from-a-fork\r\n\r\nIf this doesn't work, maybe just send me a collaboration invite to @patrickvonplaten :-)\r\n\r\nLooking forward to merge this soon :-) ", "Hey, the checkbox \"Allow edits by maintainer\" is already checked - I guess I should leave it checked right?", "Hey @drjosephliu,\r\n\r\nI tried to rebase to main, but it's just not reasonable. We have to many merge conflicts with `output_attentions` (in every file in every forward function) and not just for one commit, but for 7 commits :-/ At this point it would be much faster to just open a new PR. \r\n\r\nI should have seen this coming, so that you guys would not have started at the same time :-/ \r\nSuper sorry about that! \r\n\r\nThere are two solutions:\r\n1) You can open a new PR and sadly starting from scratch (Recommended):\r\nIf you look at the merged PR here: https://github.com/huggingface/transformers/pull/4538, you can see that we actually decided to keep `config.output_attentions` for now and only overwrite it with the arguments of the forward function. This means that the high-level forward functions all have the argument `output_hidden_states = None`. \r\n\r\n2) You can try to go through the whole rebase yourself (Not recommended):\r\n- In your repo if you run `git rebase main/master` from this branch, you see a bunch of merge conflicts arising (and that's just for the first commit of this branch). You could try to solve them correctly one-by-one, but you have to be careful to not introduce new bugs here. So I strongly advise against this.\r\n\r\n3) You don't feel like doing the same work again (Very much understandable :D). I would totally understand it, if you don't want to do the same work again. In this case I would re-open the issue to the community or do it myself. I would totally understand this - we have other \"good first issue\" tags.\r\n\r\nI hope, you give it a try with 1) :-) Very sorry about the merge conflicts again - let me know what you think", "Not a problem. I'm quite familiar with the codebase now so it shouldn't take too long.", "Saw your new PR @drjosephliu ! Thanks a lot - it's great that you tackle this :-) I will take a look tomorrow at the new PR :-) " ]