Datasets:

---

DanielPFlorian

/

Transformers-Github-Issues

like 0

[]

[]

[]

[]

[]

[ "Had to remove the following:\r\n\r\n```\r\ndevice = torch.device(\"cuda\" if torch.cuda.is_available() else \"cpu\")\r\nn_gpus = torch.cuda.device_count()\r\nif n_gpus > 1:\r\n model = nn.DataParallel(model)\r\nmodel.to(device)\r\n```\r\n\r\nThen everything is running for torch==1.7.1 for both GPUs. So `Trainer()` sorts everything by itself?" ]

[]

[]

[]

[]

[]

[]

[]

[]

[ "Your speech loading code is incorrect; instead try the following:\r\n\r\n```python\r\nfrom IPython.display import Audio\r\n\r\nspeech, rate = librosa.load(filename, sr=16000)\r\nAudio(speech, rate=rate)\r\n```", "When I run this line\r\n\r\n`processor = Speech2TextProcessor.from_pretrained(\"facebook/s2t-small-librispeech-asr\")`\r\n\r\nI am getting the following error: \"AttributeError: type object 'Speech2TextProcessor' has no attribute 'from_pretrained'\". Did this part was recently changed in the repository?\r\n\r\nEDIT: sorry, my mistake. The previous installation was causing trouble. After uninstalling everything and installing again it is working fine.", "As @elgeish said, the speech loading code was causing the issue. Glad to know that you resolved it!", "Success! Thanks\n\nOn Wed, Mar 10, 2021, 20:25 rodrigoheck ***@***.***> wrote:\n\n> When I run this line\n>\n> processor =\n> Speech2TextProcessor.from_pretrained(\"facebook/s2t-small-librispeech-asr\")\n>\n> I am getting the following error: \"AttributeError: type object\n> 'Speech2TextProcessor' has no attribute 'from_pretrained'\". Did this part\n> was recently changed in the repository?\n>\n> —\n> You are receiving this because you authored the thread.\n> Reply to this email directly, view it on GitHub\n> <https://github.com/huggingface/transformers/issues/10631#issuecomment-796383135>,\n> or unsubscribe\n> <https://github.com/notifications/unsubscribe-auth/AECXP3AO6IAKWJYWUGB42BTTDAS2PANCNFSM4Y6OTVBA>\n> .\n>\n", "I am using the maestro dataset(audio files transformed to Pytorch tensors). \r\n**Code:**\r\nif __name__ == '__main__':\r\n METADATA = \"data/processed.csv\"\r\n AUDIO_DIR = \"data\"\r\n SAMPLES = 16000\r\n SR = 16000\r\n if torch.cuda.is_available():\r\n device = \"cuda\"\r\n else:\r\n device = \"cpu\"\r\n\r\n ds = LoadDataset(metadata_file=METADATA,\r\n audio_dir=AUDIO_DIR,\r\n sample_rate=SR,\r\n num_samples=SAMPLES,\r\n device=device\r\n )\r\n dataloader = DataLoader(ds)\r\n dataiter = iter(dataloader)\r\n data = next(dataiter)\r\n features = data\r\n SR = 16000\r\n sample_rate = SR\r\n processor = AutoProcessor.from_pretrained(\r\n \"MIT/ast-finetuned-audioset-10-10-0.4593\")\r\n model = ASTModel.from_pretrained(\"MIT/ast-finetuned-audioset-10-10-0.4593\")\r\n\r\n # audio file is decoded on the fly\r\n inputs = processor(features, sampling_rate=sample_rate, return_tensors=\"pt\")\r\n with torch.no_grad():\r\n outputs = model(**inputs)\r\n\r\n last_hidden_states = outputs.last_hidden_state\r\n print(last_hidden_state) \r\n**Error Message:** \r\nCould not find image processor class in the image processor config or the model config. Loading based on pattern matching with the model's feature extractor configuration.\r\nSome weights of the model checkpoint at MIT/ast-finetuned-audioset-10-10-0.4593 were not used when initializing ASTModel: ['classifier.layernorm.bias', 'classifier.layernorm.weight', 'classifier.dense.weight', 'classifier.dense.bias']\r\n- This IS expected if you are initializing ASTModel from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\r\n- This IS NOT expected if you are initializing ASTModel from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\r\nTraceback (most recent call last):\r\n File \"model.py\", line 40, in <module>\r\n inputs = processor(features, sampling_rate=sample_rate, return_tensors=\"pt\")\r\n File \"/home/yadgire/.local/lib/python3.7/site-packages/transformers/models/audio_spectrogram_transformer/feature_extraction_audio_spectrogram_transformer.py\", line 183, in __call__\r\n features = [self._extract_fbank_features(waveform, max_length=self.max_length) for waveform in raw_speech]\r\n File \"/home/yadgire/.local/lib/python3.7/site-packages/transformers/models/audio_spectrogram_transformer/feature_extraction_audio_spectrogram_transformer.py\", line 183, in <listcomp>\r\n features = [self._extract_fbank_features(waveform, max_length=self.max_length) for waveform in raw_speech]\r\n File \"/home/yadgire/.local/lib/python3.7/site-packages/transformers/models/audio_spectrogram_transformer/feature_extraction_audio_spectrogram_transformer.py\", line 105, in _extract_fbank_features\r\n frame_shift=10,\r\n File \"/home/yadgire/.local/lib/python3.7/site-packages/torchaudio/compliance/kaldi.py\", line 592, in fbank\r\n waveform, channel, sample_frequency, frame_shift, frame_length, round_to_power_of_two, preemphasis_coefficient\r\n File \"/home/yadgire/.local/lib/python3.7/site-packages/torchaudio/compliance/kaldi.py\", line 143, in _get_waveform_and_window_properties\r\n window_size, len(waveform)\r\nAssertionError: choose a window size 400 that is [2, 1]\r\n@xjdeng Can you please help me out with this?", "@yadgire7\r\n\r\nI no longer use this model for speech to text, use [Whisper](https://huggingface.co/models?other=whisper) instead.\r\n\r\nFor music genre classification, [try converting the audio into spectrograms and training an image classifier on the spectrograms.](https://towardsdatascience.com/audio-deep-learning-made-simple-sound-classification-step-by-step-cebc936bbe5)\r\n\r\nI think you might be able to build a multimodal model that takes both the spectrogram and the transcribed text and tries to classify the music using both inputs, at least I think you could with the Fastai library by defining a text block with an image block though I haven't tried it before.\r\n\r\n" ]

[]

[]

[ "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." ]

[]

[]

[ "Pinging @sgugger ", "Yes, the default data collator always change `label` to `labels` because Hugging Face models expect that argument while Hugging Face datasets usually have `label`. You can work around this by using the default data collator of PyTorch and pass it to the `Trainer`, but you should make your model work like the one in Transformers to avoid any other issues we didn't think of (so use a `labels` argument and always return tuples).\r\n\r\nI'll see what I can do to avoid this specific bug in the future.", "Thanks a lot @sgugger, @LysandreJik. Should I close this issue?\r\n\r\nEDIT:\r\n\r\nI think that this could be mentioned in the docs where custom Trainer/TrainingArguments are discussed. What do you think?", "Like I said, will try to solve the bug in itself. My recommendation was more in general to avoid any other bugs :-) ", "Thanks again @sgugger :) " ]

[]

[]

[ "Sounds reasonable. Would you like to make a PR with this change?", "Can do.\r\nI should expand the path in TrainingArguments right? for logging_dir also?", "Yes, `output_dir` and `logging_dir`, preferable in the postinit of `TrainingArguments` so it's done as early as possible." ]

[]

[]

[ "Indeed, we could use that when the line by line option is set (otherwise there is just no padding). Would you like to make a PR with this?" ]

[]

[]

[ "Hello, thanks for opening an issue! We try to keep the github issues for bugs/feature requests.\r\nCould you ask your question on the [forum](https://discusss.huggingface.co) instead?\r\n\r\nThanks!" ]

[]

[]

[ "No, DeBERTa-v2 is not available in v4.3.3, it's only available from source as of now. Version v4.4.0 should be released end of this week or early next week, and will have DeBERTa-v2.", "Yes, exactly what i think! Thanks Lysandre for confirming this." ]

[]

[]

[ "Love it!\r\n\r\nMaybe a good practice to link to a sample of the related commits on hf.co: for instance here https://huggingface.co/facebook/bart-base/commit/c2469fb7e666a5c5629a161f17c9ef23c85217f7", "I think I did around 50 of them in various repos to move all the tokenizers files, so a bit hard to keep track of all of them.", "Yep just link one, or a small sample.\r\n\r\nMakes it easier to see what this PR entails on hf-hub side" ]

[]

[]

[ "Hi, not sure why this happens, cc @julien-c.\r\n\r\nA workaround is to load the model using `model = TapasModel.from_pretrained(\"google/tapas-base\")` and then use `model.save_pretrained(\"./\")` to save the `config.json` and `pytorch_model.bin` file to a local directory. ", "I don't remember how those models were uploaded so not sure why this is happening.\r\n\r\ncc'ing @Pierrci for visibility\r\n\r\nIn the meantime you can just rename the file to .bin", "@saichandrapandraju the file on tapas-large is a bin file, but since pytorch 1.6.0, bin files are now zip-based. You can check the documentation of [torch.save](https://pytorch.org/docs/stable/generated/torch.save.html) which states:\r\n\r\n> The 1.6 release of PyTorch switched torch.save to use a new zipfile-based file format. torch.load still retains the ability to load files in the old format. If for any reason you want torch.save to use the old format, pass the kwarg _use_new_zipfile_serialization=False.\r\n\r\nAre you using a torch version inferior to 1.6.0 to try to load the models?", "As pointed out by @julien-c, it is actually downloaded as a zip file, which seems to be the case for several models (my guess is that it does that when the file is zip-based, like all models saved with torch >1.6).\r\n\r\nDownloading through git doesn't have that issue.", "Ok will close this as:\r\n- we found the root cause but it's outside our control (`lfs` adds an auto content-type)\r\n- there are several workarounds like `git clone`ing the repo or just renaming the file (`from_pretrained` also works as usual)\r\n\r\nThanks for investigating @Pierrci and @LysandreJik 🥇 " ]

[]

[]

[ "My workaround right now is to keep a reference to the original `tokenizer.unique_no_split_tokens` before adding tokens then restoring it afterwards:\r\n\r\n```python\r\nfrom transformers import Wav2Vec2Processor\r\n\r\ntokenizer = Wav2Vec2Processor.from_pretrained('facebook/wav2vec2-base').tokenizer\r\nunique_no_split_tokens = tokenizer.unique_no_split_tokens\r\ntokenizer.add_tokens('x')\r\ntokenizer.unique_no_split_tokens = unique_no_split_tokens\r\ntoken_ids = tokenizer('C x A').input_ids\r\ndecoded = tokenizer.decode(token_ids)\r\nprint(decoded, token_ids)\r\n# C x A [19, 4, 32, 4, 7]\r\n```", "Hey @elgeish,\r\n\r\nSorry for replying that late! Yes, you are absolutely right here :-)\r\nI think we should overwrite the `add_tokens(self, ...)` function in \r\n`src/transformers/models/wav2vec2/tokenization_wav2vec2.py` with the \"hack\" just as you did:\r\n\r\n```python\r\ndef _add_tokens(self, new_tokens: Union[List[str], List[AddedToken]], special_tokens: bool = False) -> int:\r\n # copy past the function from `src/transformers/tokenization_utils.py` \r\n # + add the \"hack\": \r\n unique_no_split_tokens = tokenizer.unique_no_split_tokens\r\n tokenizer.unique_no_split_tokens = unique_no_split_tokens\r\n```\r\n\r\nIf you want and have some time, it would be amazing if you could open a PR :-)\r\n", "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.", "Hey @patrickvonplaten,\r\nShall I open a PR for this issue.", "Hey @Muktan,\r\n\r\nyes this would be great :-)" ]

[]

[]

[]

[]

[ { "login": "sgugger", "id": 35901082, "node_id": "MDQ6VXNlcjM1OTAxMDgy", "avatar_url": "https://avatars.githubusercontent.com/u/35901082?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sgugger", "html_url": "https://github.com/sgugger", "followers_url": "https://api.github.com/users/sgugger/followers", "following_url": "https://api.github.com/users/sgugger/following{/other_user}", "gists_url": "https://api.github.com/users/sgugger/gists{/gist_id}", "starred_url": "https://api.github.com/users/sgugger/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sgugger/subscriptions", "organizations_url": "https://api.github.com/users/sgugger/orgs", "repos_url": "https://api.github.com/users/sgugger/repos", "events_url": "https://api.github.com/users/sgugger/events{/privacy}", "received_events_url": "https://api.github.com/users/sgugger/received_events", "type": "User", "site_admin": false } ]

[ "Fixed by #10621 \r\nThanks for flagging and fixing :-)" ]

[]

[]

[ "Can you try this?\r\n\r\n```python\r\nfrom transformers import Wav2Vec2Processor\r\n\r\ntokenizer = Wav2Vec2Processor.from_pretrained('facebook/wav2vec2-base').tokenizer\r\ntokenizer.decode(tokenizer('CARRY').input_ids, group_tokens=False)\r\n# CARRY\r\n```\r\n\r\nBecause we need to decode the predicted tokens with CTC, `\"RR\"` is decoded to `\"R\"` by default. See this blog post for more information: https://distill.pub/2017/ctc/", "Yeah looks good! Thanks!", "By the way, I totally get why it's needed for CTC, I mistook it for the tokenizer used to decode the final results but noticed it wasn't the case. The final results work as expected. Sorry for the false alarm!" ]

[]

[]

[ "This is weird and linked to your environment somehow. \r\n@stas00 Was this the error you encountered when `dataclasses` is installed in Python 3.7 or was it a different one?", "no, that was not that error. I tested `run_qa.py` w/ dataclasses on py38 and it didn't fail.\r\n\r\nthe datasets error was: `AttributeError: module 'typing' has no attribute '_ClassVar'`\r\n\r\nhttps://github.com/huggingface/transformers/issues/8638", "I just tried this on 2 new servers with a fresh conda environment and reproduced behavior. \r\nSteps.\r\n```bash\r\nconda create -n test python=3.8\r\nconda activate test\r\npip install transformers datasets torch\r\npython run_qa.py --model_name_or_path bert-base-uncased --dataset_name squad --do_train --per_device_train_batch_size 8 --learning_rate 3e-5 --max_seq_length 384 --doc_stride 128 --output_dir bert-base-uncased-qa/ --overwrite_output_dir --cache_dir cache --preprocessing_num_workers 4 --seed 42 --num_train_epochs 1\r\n```\r\n", "I have also reproed with venv and regular environment on multiple machines", "The suggested commands work fine on my side, so can't reproduce the issue. ", "I have pushed a fix (on master by mistake but it's pretty harmless) a tentative fix to remove the line that caused you problem and replace it by a regex. Let me know if it fixes your issue or not (I can't confirm myself since I can't reproduce).", "FWIW, I followed your new conda env steps and couldn't reproduce the problem.\r\n\r\n@spacemanidol, fyi I edited your comment to fix the conda create line as it had the commands reversed.", "Can confirm this works. " ]

[]

[]

[ "This is already there :-) Just pass along `--ignore_data_skip` in your script or `ignore_data_skip=True` in your `TrainingArguments`.", "Wow, that was fast! :) \r\n\r\nThat loads the model from the checkpoint and advances the dataset to the next sample that would have been trained in the original run?\r\n\r\nFrom my reading of the code I assumed that it reloaded the model and started the training over at the first sample of the dataset.\r\n", "Ah sorry I misunderstood your feature request. Indeed it starts from the first sample instead of iterating. What you ask is a bit more complicated and will require a minimum version of datasets. It's possible but will take some time.", "Sorry for the confusion from my writing! And thanks, as always, for your work on this amazing project.", "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." ]

[]

[]

[ "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." ]

[]

[]

[]

[]

[]

[ "It seems that T5 was added to the conversion scripts two months ago in https://github.com/huggingface/transformers/pull/9654 as you've mentioned.\r\n\r\nHowever, in your error there is no mention of \"t5\":\r\n```\r\nValueError: --model_type should be selected in the list [bert, gpt, gpt2, transfo_xl, xlnet, xlm]\r\n```\r\n\r\nBut on master it clearly shows there should be one:\r\n```\r\n\"--model_type should be selected in the list [bert, gpt, gpt2, t5, transfo_xl, xlnet, xlm, lxmert]\"\r\n```\r\n\r\nAre you certain you're launching the command in the correct environment? Could you share the result of `transformers-cli env`?\r\n\r\n---\r\n\r\nAlso, this conversion command is far from being complete, I think it could use some of your templating magic @sgugger if you ever feel like it :) \r\nIt isn't high priority as we already have the model-specific conversion scripts.", "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." ]

[]

[]

[ "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.", "This is a pretty big deal one would think. An almost 100% bloat of the model checkpoint when exporting compared to the model card...?", "This is impacting me as well, is it possible for us to reopen. I am happy to provide more relevant deatils", "This seems to be a duplicate of https://github.com/huggingface/transformers/issues/11222" ]

[ { "id": 1843244711, "node_id": "MDU6TGFiZWwxODQzMjQ0NzEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model", "name": "New model", "color": "fbca04", "default": false, "description": "" } ]

[]

[ "There are already some PRs regarding these models, I'm working on adding the Linformer (#10587), there's also a PR for the Performer (#9325, see further down the thread - people can already train T5 with Performer). " ]

[]

[]

[ "cc @stas00 `run_seq2seq` is left as is for now. At some point, if your tests migrate from that script to either the new ones or another one, we will remove it.", "> cc @stas00 `run_seq2seq` is left as is for now. At some point, if your tests migrate from that script to either the new ones or another one, we will remove it.\r\n\r\nI know we discussed to potentially leave the all-in-one script for performance testing, but it's very likely we will be using a different approach that Morgan created.\r\n\r\nTherefore please don't leave this on me - please sync the tests with these changes and remove the do-it-all script. Thank you.\r\n", "May I suggest that the examples are inconsistent script naming-wise, some are very abbreviated `run_clm.py`, others are the extreme opposite `run_summarization.py` - that's a way too much to type - won't `run_sum.py` be sufficient?\r\n\r\nFor \"to type\" I mean when referring to them in documents, Issues, etc. There is no file-completion there.", "> May I suggest that the examples are inconsistent with script naming-wise, some are very abbreviated `run_clm.py`, others are the extreme opposite `run_summarization.py` - that's a way too much to type - won't `run_sum.py` be sufficient?\r\n\r\nmy personal preference goes to clarity so `run_summarization.py` trumps `run_sum.py` - but consistency is also important - unsure about the best tradeoff here\r\n\r\nedit: if we were to shorten the script names, what would be the matching acronym for `run_translation.py`?", "`run_trans.py`\r\n\r\nand just to clarify, I'm just flagging the inconsistency and my preference to type less, and in no way suggesting to interfere with this process - if most of you prefer the long names - go for it. ", "Also while we are at it or perhaps after the split - different PR, this further proposed improvement could be applied:\r\nhttps://github.com/huggingface/transformers/issues/10164\r\n", "I'm not seeing the benefit of shortening to `run_sum` and `run_trans` when tab-complete will give you the full name. In the language-modeling folder, the acronym was used because `run_causal_language_modeling`/`run_masked_language_modeling` is really long, same for question-answering.\r\n\r\nIf summarization and translation took two words at least, we could use the acronym to shorten the name, but since they don't I would go for the full name.", "> I'm not seeing the benefit of shortening to `run_sum` and `run_trans` when tab-complete will give you the full name. In the language-modeling folder, the acronym was used because `run_causal_language_modeling`/`run_masked_language_modeling` is really long, same for question-answering.\r\n> \r\n> If summarization and translation took two words at least, we could use the acronym to shorten the name, but since they don't I would go for the full name.\r\n\r\nI can't see how one is much longer than the other:\r\n```\r\nrun_causal_language_modeling\r\nrun_summarization\r\n```\r\n\r\nWould `run_causal_lm` and `run_masked_lm` be perhaps a good middle ground if you want example names to be of the explicit type?\r\n\r\n", "> I'm not seeing the benefit of shortening to `run_sum` and `run_trans` when tab-complete will give you the full name. In the language-modeling folder, the acronym was used because `run_causal_language_modeling`/`run_masked_language_modeling` is really long, same for question-answering.\r\n> \r\n> If summarization and translation took two words at least, we could use the acronym to shorten the name, but since they don't I would go for the full name.\r\n\r\nMy current understanding is that consistency primes for now, and the change for more explicit script names should be done in a separate PR, if there is indeed consensus that explicit names are better here.", "Very nice PR overall! \r\n\r\nI don't really agree though with the `sum`, `trans` and the \"one-size-fits-it-all\" design choices. Maybe we can settle on displaying a warning for T5 @stas00 ?", "> I don't really agree though with the `sum`, `trans` \r\n\r\nAs long as other examples are consistently named then it works.\r\n\r\n> and the \"one-size-fits-it-all\" design choices. \r\n\r\nIt doesn't sound like we are reaching a consensus here. I outlined that entering the same data more than once in the same input is error-prone. Perhaps there is another way to fix this w/o \"one-size-fits-it-all\" \r\n\r\n> Maybe we can settle on displaying a warning for T5 @stas00 ?\r\n\r\nYes, please. At the very least.", "@theo-m, @stas00 - if it's fine for you maybe we can change the script names to `run_summarization.py` and `run_translation.py` then and replace the automatic setting of `prefix` for T5 with a warning instead. Would that work? I'm more than happy to merge this PR then", "If the group prefers it this way then it is fine as you propose.", "> If the group prefers it this way then it is fine as you propose.\r\n\r\nOkey, maybe @sgugger, @patil-suraj and @LysandreJik can give their final word then as well", "I would prefer explicit names (`run_summarization.py` and `run_translation.py`) and not handling T5 prefixes automatically.", "I have already given my opinion on the names. For the `source_prefix` I have no strong opinion as I don't see any \"good\" solution sadly. I'm fine with the warning.", "Seems to me this could be merged, last call for maintainers @patrickvonplaten @sgugger @LysandreJik (thumbs up on this message will be interpreted as a go 😉 )" ]

[]

[]

[]

[]

[]

[ "This issue is more suited for the forum, but maybe @sgugger has some hints to share!", "I'm not sure what you would want to sort a single text stream. The `Trainer` supports `--group_by_length` but that's when you have multiple texts.\r\n\r\nNote that `DataCollatorForLanguageModeling` only performs random masking on your prepared data, nothing more.", "Sorry, maybe I was imprecise here (still trying to wrap my head around a lot of these new concepts). Essentially, what I was wondering if batches for language modeling are constructed in a way that is similar to the approach that you described in [Chapter 10: \r\nNLP Deep Dive](https://github.com/fastai/fastbook/blob/master/10_nlp.ipynb) (section 'Putting Our Texts into Batches for a Language Model') of the book you co-authored. If I got it correctly, that would allow me to train the transformer over variable length input sequences without worrying about sequences being truncated due to the constraints imposed by `tokenizer.max_len_single_sentence` (overflowing parts would simply end up at the appropriate position in the nex mini-batch).", "I think you may be referring to the LM DataLoader. This kind of preprocessing is done using the `datasets` library on our side. Take a look at the [run_clm](https://github.com/huggingface/transformers/blob/master/examples/language-modeling/run_clm.py) or [run_mlm](https://github.com/huggingface/transformers/blob/master/examples/language-modeling/run_mlm.py) examples (in run_mlm the part that is not in the block \"line_by_line\") or the [language modeling notebook](https://colab.research.google.com/github/huggingface/notebooks/blob/master/examples/language_modeling.ipynb) to see how.", "Thanks a lot! Indeed, the `group_text()` function was exactly what I was looking for. ", "Hey @sgugger,\r\n\r\nsorry for reopening this one. I am still not 100% sure if I conveyed my issue properly the first time. Hence, let me briefly restate it using an example:\r\n\r\nLet's say I chunk my text stream of length 100 (let's say including 3 sentences) into blocks of `block_size=5`, resulting in 20 blocks. Now, I'd like to feed them into my model using a `batch_size=5`, resulting in 4 batches à 5 text blocks.\r\n\r\nI am still not 100% sure how they are fed into the model using the `DataCollatorForLanguageModeling` and `Trainer` API:\r\n```python\r\n# Variant A\r\n\r\nbatch_1 = [block1, block2, block3, block4, block5]\r\n# ...\r\nbatch_4 = [block16, block17, block18, block19, block20]\r\n``` \r\n\r\n```python\r\n# Variant B\r\n\r\nbatch_1 = [block1, block5, block9, block13, block17]\r\nbatch_2 = [block2, block6, block10, block14, block18]\r\n# ...\r\nbatch_4 = [block4, block8, block12, block16, block20]\r\n``` \r\n\r\nIf I got it correctly, the method presented in the book referred to in the previous comment relates to *variant B*.", "You will need to write your own data collator for that as this is not in the Transformers library: contrary to LSTMs, Transformers do not have a state so we don't care about the ordering across batches for those models.", "This makes entirely sense, thanks for lifting this barrier in my head! " ]

[]

[]

[]

[]

[]

[ "Hello! I can do `model = AutoModelForMaskedLM.from_pretrained(\"EMBEDDIA/sloberta\")` without any issues on my end.\r\n\r\nCould it be linked to a connection issue?\r\n\r\n```py\r\n>>> from transformers import AutoModelForMaskedLM\r\n>>> model = AutoModelForMaskedLM.from_pretrained(\"EMBEDDIA/sloberta\")\r\nDownloading: 100%|██████████| 520/520 [00:00<00:00, 215kB/s]\r\nDownloading: 100%|██████████| 443M/443M [00:38<00:00, 11.5MB/s]\r\n\r\n```", "Tested from two computers on two different networks, didn't work. Managed to load it on Google Colab, though.\r\n\r\nAnyway, it does work now. I was foiled by conda loading an old version of transformers instead of a newer one. Thanks!" ]

[]

[]

[]

[]

[ { "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 }, { "login": "patil-suraj", "id": 27137566, "node_id": "MDQ6VXNlcjI3MTM3NTY2", "avatar_url": "https://avatars.githubusercontent.com/u/27137566?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patil-suraj", "html_url": "https://github.com/patil-suraj", "followers_url": "https://api.github.com/users/patil-suraj/followers", "following_url": "https://api.github.com/users/patil-suraj/following{/other_user}", "gists_url": "https://api.github.com/users/patil-suraj/gists{/gist_id}", "starred_url": "https://api.github.com/users/patil-suraj/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patil-suraj/subscriptions", "organizations_url": "https://api.github.com/users/patil-suraj/orgs", "repos_url": "https://api.github.com/users/patil-suraj/repos", "events_url": "https://api.github.com/users/patil-suraj/events{/privacy}", "received_events_url": "https://api.github.com/users/patil-suraj/received_events", "type": "User", "site_admin": false } ]

[ "Hi @patrickvonplaten & @patil-suraj, my PR does not pass one test, however, I am not able to reproduce this error on my local (I can't even find the file at `src/transformers/models/new_enc_dec/modeling_new_enc_dec.py` in the repo, which is the one where should be a problem with a copy inconsistency)", "Hi @stancld thank you for your work! The issue is because you have updated all model files (thank you!!), but you haven't updated the template. The template is used when adding a new model, it's available [here](https://github.com/huggingface/transformers/tree/master/templates/adding_a_new_model).\r\n\r\nFor example, these lines should probably be updated to include the `cross_head_mask`:\r\n\r\nhttps://github.com/huggingface/transformers/blob/44f64132a5f50726f9de4467ed745421c3b11ab3/templates/adding_a_new_model/cookiecutter-template-%7B%7Bcookiecutter.modelname%7D%7D/modeling_%7B%7Bcookiecutter.lowercase_modelname%7D%7D.py#L2070-L2085", "@LysandreJik - Thank you very much for the clarification :)", "Wonderful, the tests pass! Thanks for handling the templates.", "@patil-suraj Thank you for your review and the suggestions. I agree with the change in variable naming. The new one is more accurate. I will change it everywhere accordingly :) ", "Hi @patrickvonplaten, I've just rebased this branch to the current master to keep this PR up to date. Could you please review this one? Thanks a lot! :) ", "Hey @stancld, Patrick is off for a couple of weeks but will take a look at this as soon as he's back :)", "The PR looks good to me :-) \r\n\r\nThink we just need to fix the docstring. @stancld - let me know if you need help regarding the docstring", "@patrickvonplaten The docstring should be fixed now. I forgot a line from a conflict there.. :)" ]

[]

[]

[]

[]

[]

[ "Hello! You're loading a model called `albert_chinese_base`; my guess is that this model only contains the base transformer model and not the sequence classification head that you need.\r\n\r\nDoes that make sense? You should use a model fine-tuned on sequence classification, and not a base model, if you want to do sequence classification. Of course, that model will be fine-tuned to a specific sequence classification task so it can't be used in any context.", "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." ]

[]

[]

[ "Ya @stas00,\nActually I figured that and tested visually with the example tamplate not the model template" ]

[]

[]

[ "This single change made the tests on the TF CI pass from 6+ hours (non slow) to 29/31 minutes: https://github.com/huggingface/transformers/actions/runs/635711849\r\n\r\nWill look for a way to reduce their time." ]

[]

[]

[]

[]

[]

[ "Hi @ymfa , thanks a lot for the PR!\r\n\r\n- The `generate` method does allow you to pass `input_embeds` as a keyword argument (`**model_kwargs` arg). If `input_embeds` is passed then T5 or any other model will use those instead of `input_ids`. If you look at the `generate` method signature, you can see that `input_ids` is optional.\r\n\r\n- I think we can allow to pass `encoder_outputs` directly to `generate`. Pinging @patrickvonplaten ", "Hey @ymfa thanks for your PR and your thoughtful explanations in the description! \r\n\r\nThe design philosophy of `generate()` is that the 99% cases should be covered by `generate()` and for specific cases the sub-generated generate functions should be called directly as it is the case in the examples here:\r\nhttps://github.com/huggingface/transformers/blob/0d909f6bd8ca0bc1ec8f42e089b64b4fffc4d230/src/transformers/generation_utils.py#L1592\r\n\r\nAs you can see, you only need to add a couple more lines when directly using `beam_search(...)` instead `generate(...)`. Could this solve your use case?\r\n\r\n\r\n\r\n", "Hi @patil-suraj , thanks for your comment. \r\n\r\nI do realise that I can pass additional arguments via `**model_kwargs`. However, it doesn't work if I pass `inputs_embeds` instead of `input_ids`, because there are a number of places in `generate()` that depend on `input_ids`. \r\n\r\nSpecifically, this is the error I got (model is T5ForConditionalGeneration):\r\n```\r\n>>> model.generate(inputs_embeds=input_embeded)\r\nTraceback (most recent call last):\r\n...\r\nValueError: `bos_token_id` has to be defined when no `input_ids` are provided.\r\n```\r\n\r\nIf I do pass `bos_token_id` (which shouldn't be necessary), this is the error (it's because the `input_ids` are created from `bos_token_id` and passed to the encoder):\r\n```\r\n>>> model.generate(inputs_embeds=input_embeded, bos_token_id=0)\r\nTraceback (most recent call last):\r\n...\r\nValueError: You cannot specify both inputs and inputs_embeds at the same time\r\n```\r\n\r\nSo in the end, I am now using the method in this PR to achieve this purpose.", "Thanks @patrickvonplaten . \r\n\r\nTo be honest I haven't found a way to make `beam_search()` work even for the simple use case of passing ids only. In this example (model is T5ForConditionalGeneration), `input_ids_batch` is the just 5 identical `input_ids` stacked together for the beam size.\r\n```\r\n>>> input_ids_batch.shape\r\ntorch.Size([5, 15])\r\n>>> model.beam_search(input_ids_batch, beam_scorer)\r\nTraceback (most recent call last):\r\n...\r\n File \"...python3.7/site-packages/transformers/models/t5/modeling_t5.py\", line 871, in forward\r\n raise ValueError(f\"You have to specify either {err_msg_prefix}inputs or {err_msg_prefix}inputs_embeds\")\r\nValueError: You have to specify either inputs or inputs_embeds\r\n```\r\n\r\nWhen I pass the `encoder_outputs`:\r\n```\r\n>>> model.beam_search(decoder_input_ids_batch, beam_scorer, encoder_outputs=encoder_outputs)\r\nTraceback (most recent call last):\r\n...\r\n File \"...python3.7/site-packages/transformers/models/t5/modeling_t5.py\", line 498, in forward\r\n scores += position_bias\r\nRuntimeError: The size of tensor a (3) must match the size of tensor b (15) at non-singleton dimension 3\r\n```\r\n\r\nI don't think it would be a simpler PR if I change `beam_search()` instead of `generate()`. It would also be less general as it doesn't apply to the other decoding methods, and it would require the user to prepare `beam_scorer` and `decoder_input_ids_batch`. ", "> scores += position_bias\r\n\r\nI think you just need to broadcast your `decoder_input_ids_batch` to be of the same size as `encoder_outputs`, then it should work. \r\n\r\nOverall, I'm not really in favor of adding this new special use case. To me it's not naturally to do `model.generate(None, decoder_input_ids=..., encoder_outputs=encoder_outputs)` lot of people won't understand that the first argument has in fact to be `None` since it corresponds to the encoder input ids. For such a case it should be relatively easy to make it work with `beam_search` to be honest. The philosophy here is that that both `decoder_input_ids` and `encoder_outputs.last_hidden_state` have to be of the sample batch dimensions, so `encoder_outputs.shape][0] == decoder_input_ids.shape[0]` . Then the command is as simple as:\r\n\r\n```\r\nmodel.beam_search(decoder_input_ids, beam_scorer, encoder_outputs)\r\n```\r\n\r\nThe problem is that if we allow to many specific use cases for `generate()` the method becomes quite cluttered with if-statements again and I would like to avoid it. In this case, I think it's much cleaner to directly call the `beam_search` method tbh.", "What do you think @patil-suraj ?", "I've made a change so that the method signature of `generate()` remains the same as before. There is not \"clutterness\" added into this method now. This PR can be regarded as a fix, by properly handling the case when `encoder_outputs` is passed as one of the `model_kwargs`.\r\n\r\nYou're right that decoder_input_ids and encoder_outputs.last_hidden_state have to be of the sample batch dimensions. However, this means both of them need to be broadcast according to the beam size, which is not done automatically. A user would have to broadcast these tensors and objects manually in order to call `beam_search()`.\r\n\r\nTo be honest, I have been using the patch I submitted here for doing both beam search and sampling. The easy-to-use generation utility is one of my main reasons to choose the transformers package. I don't think it is a good idea to limit the potential based on \"cleanness.\" On the contrary, I even suggest refactoring `generate()` more systematically, so that `input_ids` is no longer used like a central \"currency\" in this method.", "Hey @ymfa, \r\n\r\nIt's actually a good point that changing between different generation methods while using `encoder_outputs` is not very user-friendly and \"pre-computing\" encoder_outputs is quite a common use case for all seq2seq models. Also since only the \"helper\" methods `_prepare...` are changed, I think I'm fine with the PR now! Thanks for being persistent here!\r\n\r\n@patil-suraj @LysandreJik - it would be nice you could take a look as well.", "I think `generate` now deserves its own doc page where we could explain this and maybe some more details like what the method supports, what are its limitations and what features won't be supported etc. It's changed significantly after the refactor and the design now follows a strict philosophy. It would be better to document that. What do you think @patrickvonplaten?\r\n\r\n@ymfa The PR is good to merge, I'll merge it once you add helpful comments in `_prepare_input_ids_for_generation` as suggested by Patrick and Lysandre." ]

[]

[]

[]

[]

[]

[ "Are you talking about this model? https://huggingface.co/mrm8488/roberta-large-finetuned-wsc", "> Are you talking about this model? https://huggingface.co/mrm8488/roberta-large-finetuned-wsc\r\n\r\nYes, this model i can not call by transformers. So how can i use it?", "You can't call it from transformers? Do you get an error?\r\nIt seems I can load it:\r\n\r\n```py\r\n>>> from transformers import AutoModelForMaskedLM\r\n>>> model = AutoModelForMaskedLM.from_pretrained(\"mrm8488/roberta-large-finetuned-wsc\")\r\n```", "> You can't call it from transformers? Do you get an error?\r\n> It seems I can load it:\r\n> \r\n> ```python\r\n> >>> from transformers import AutoModelForMaskedLM\r\n> >>> model = AutoModelForMaskedLM.from_pretrained(\"mrm8488/roberta-large-finetuned-wsc\")\r\n> ```\r\n\r\nI couldn't load it from \"AutoModel\". Thanks for your snippet! Anyways, should i finetune this for text classification task by removing Language Modeling head on top? Just for experiments!", "You can load it in a text-classification auto model in order to fine-tune it to text-classification:\r\n\r\n```py\r\n>>> from transformers import AutoModelForSequenceClassification\r\n... model = AutoModelForSequenceClassification.from_pretrained(\"mrm8488/roberta-large-finetuned-wsc\")\r\n```\r\n\r\nIt tells you that the LM head layers were discarded, and that it initialized randomly the layers for text-classification:\r\n\r\n```\r\nSome weights of the model checkpoint at mrm8488/roberta-large-finetuned-wsc were not used when initializing RobertaForSequenceClassification: ['lm_head.bias', 'lm_head.dense.weight', 'lm_head.dense.bias', 'lm_head.layer_norm.weight', 'lm_head.layer_norm.bias', 'lm_head.decoder.weight', 'lm_head.decoder.bias', 'roberta.pooler.dense.weight', 'roberta.pooler.dense.bias']\r\n- This IS expected if you are initializing RobertaForSequenceClassification from the checkpoint of a model trained on another task or with another architecture (e.g. initializing a BertForSequenceClassification model from a BertForPreTraining model).\r\n- This IS NOT expected if you are initializing RobertaForSequenceClassification from the checkpoint of a model that you expect to be exactly identical (initializing a BertForSequenceClassification model from a BertForSequenceClassification model).\r\nSome weights of RobertaForSequenceClassification were not initialized from the model checkpoint at mrm8488/roberta-large-finetuned-wsc and are newly initialized: ['classifier.dense.weight', 'classifier.dense.bias', 'classifier.out_proj.weight', 'classifier.out_proj.bias']\r\nYou should probably TRAIN this model on a down-stream task to be able to use it for predictions and inference.\r\n```\r\n\r\nYou should now fine-tune this on a text-classification dataset so that the randomly initialized layers may be trained!", "Oh wow, i don't fine tune for classification by this way. I directly removed the LM head though. But still thank you Lysandre!", "Happy to help!" ]

[ { "id": 2803958109, "node_id": "MDU6TGFiZWwyODAzOTU4MTA5", "url": "https://api.github.com/repos/huggingface/transformers/labels/fairscale", "name": "fairscale", "color": "A94273", "default": false, "description": "" } ]

[]

[]

[]

[]

[]

[]

[]

[]

[]

[]

[]

[]

[]

[ "Hi! I don't think torch supports CUDA 11.2 yet. See https://github.com/pytorch/pytorch/issues/50232#issuecomment-777703998", "I had a similar issue with torch 1.8 and solved it by downgrading to 1.7.1", "> Hi! I don't think torch supports CUDA 11.2 yet. See [pytorch/pytorch#50232 (comment)](https://github.com/pytorch/pytorch/issues/50232#issuecomment-777703998)\r\n\r\nThanks for the quick response. I just tested the script with CUDA11.1 and it worked just fine.", "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." ]

[]

[]

[]

[]

[]

[]

[]

[ { "login": "LysandreJik", "id": 30755778, "node_id": "MDQ6VXNlcjMwNzU1Nzc4", "avatar_url": "https://avatars.githubusercontent.com/u/30755778?v=4", "gravatar_id": "", "url": "https://api.github.com/users/LysandreJik", "html_url": "https://github.com/LysandreJik", "followers_url": "https://api.github.com/users/LysandreJik/followers", "following_url": "https://api.github.com/users/LysandreJik/following{/other_user}", "gists_url": "https://api.github.com/users/LysandreJik/gists{/gist_id}", "starred_url": "https://api.github.com/users/LysandreJik/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/LysandreJik/subscriptions", "organizations_url": "https://api.github.com/users/LysandreJik/orgs", "repos_url": "https://api.github.com/users/LysandreJik/repos", "events_url": "https://api.github.com/users/LysandreJik/events{/privacy}", "received_events_url": "https://api.github.com/users/LysandreJik/received_events", "type": "User", "site_admin": false } ]

[ "Hi! In order to use either of those weights you'll have to convert them to a HuggingFace format.\r\n\r\nFor that you have two available scripts:\r\n- From [TF1](https://github.com/huggingface/transformers/blob/master/src/transformers/models/bert/convert_bert_original_tf_checkpoint_to_pytorch.py)\r\n- From [TF2](https://github.com/huggingface/transformers/blob/master/src/transformers/models/bert/convert_bert_original_tf2_checkpoint_to_pytorch.py)", "Hi @LysandreJik! Thanks for always being kind and active to solve our questions, I have checked both TF1 and TF2 convert script which TF1 works perfectly to convert bert_model.ckpt to pytorch_model.bin, while the TF2 will always give error message as picture here:\r\n\r\nTo add on, in my task I keep training checkpoint model on my dataset by running [run_pretraining.py ](https://github.com/tensorflow/models/blob/master/official/nlp/bert/run_pretraining.py), so according to the script description on TF2 convert script, I believe the problem is mlm head which I have to add something above the convert script to convert model right? or keep running run_pretraining.py won't give such header on it?\r\n\r\nIf the header exist, how can I solve this problem? Even though I tried BERT for months but I am not that confident to say I understand how it works. In #9941 I have tried to add elif condition on m_name but seems it's a bad approach on this question. \r\n\r\nAgain, thanks for the reply! Really appreciated for giving me such direction to it!\r\n\r\n", "Hmmm, I see this is an issue indeed! Could you let me know how you obtained your TF2 checkpoint so that I may check it on my side?\r\n\r\nYou're welcome, happy to help :)", "Sure! Due to the [repository](https://github.com/tensorflow/models/tree/master/official/nlp/bert#access-to-pretrained-checkpoints) haven't release chinese model yet, so the way I obtained this model is by these steps, \r\n1. Download [bert-base-chinese](https://storage.googleapis.com/bert_models/2018_11_03/chinese_L-12_H-768_A-12.zip) from google-bert\r\n2. Use this [script](https://github.com/tensorflow/models/blob/master/official/nlp/bert/tf2_encoder_checkpoint_converter.py) to convert tf1 checkpoint to tf2 with following args\r\n`python tf2_encoder_checkpoint_converter.py --checkpoint_to_convert=$BASE_DIR/bert_model.ckpt --converted_checkpoint_path=tmp/ --bert_config_file=$BASE_DIR/bert_config.json`\r\nnote: $BASE_DIR is the path to the model directory (which is the model done by step 1 & 2)\r\n3. Using [create_pretraining_data.py](https://github.com/tensorflow/models/blob/master/official/nlp/data/create_pretraining_data.py) to create the dataset that I want to keep training by its domain, the simple data I use can be found [here](https://drive.google.com/file/d/163IAWfgQZ1TIN6WYzWdoRwmFXDzGIRxj/view?usp=sharing).\r\nThe args I use are like this\r\n`python create_pretraining_data.py --input_file=./sample.txt --output_file=ModelRecord --vocab_file=./$BASE_DIR/vocab.txt --max_seq_length=128 --max_predictions_per_seq=19 --masked_lm_prob=0.15 --random_seed=46 --dupe_factor=1`\r\n4. Now we have ModelRecord, use tf2 version [run_pretraining.py](https://github.com/tensorflow/models/blob/master/official/nlp/bert/run_pretraining.py) to training model on these instances in ModelRecord. The args I use look like this\r\n`python run_pretraining.py --input_file=ModelRecord --output_dir=tmp/Model_L128_B32 --bert_config_file=bert_config.json --max_seq_length=128 --max_predictions_per_seq=19 --do_train=True --train_batch_size=32 --num_train_steps=2000000 --num_warmup_steps=2000 --learning_rate=1e-5 --save_checkpoints_step=60000 --keep_checkpoint_max=4`\r\n5. After training, the directory in Model_L128_B32(where model saved) containing these checkpoint,\r\n \r\nI load the checkpoint in directory \"pretrained\", and rename the checkpoint I want to evaluate to model.ckpt.data-00000-of-00001, ckpt_index to model.ckpt.index, with config and vocab in the directory. \r\n\r\n### System Info\r\nMy environments are: (using pipreqs)\r\nPython version: 3.7.5\r\nCUDA used to build PyTorch: cuda_11.0_bu.TC445_37.28845127_0\r\nOS: Ubuntu 18.04.5 LTS\r\nGCC version: 7.5.0\r\n\r\n### Versions of relevant libraries: (using pipreqs)\r\nnumpy==1.19.5\r\ntransformers==4.2.2\r\nsix==1.15.0\r\ntorch==1.7.1+cu110\r\ntensorflow_gpu==2.2.0\r\ngin_config==0.1.1\r\nabsl_py==0.11.0\r\ntensorflow_hub==0.11.0\r\nsentencepiece==0.1.94\r\nabsl==0.0\r\ntorchvision==0.8.2+cu110\r\nbert4keras==0.9.9\r\ngin==0.1.006\r\ntensorflow==2.4.1\r\n\r\n\r\nI think that's all of it, if I miss any step please inform me with no hesitate, I might get wrong :P \r\nBest regards!\r\n", "I see, the conversion script should work for that use-case. Is there a way for you to share the checkpoints you have obtained so that I can take a look? You can share them through the hub under your username.", "I test for a while and also follow what #8504 did, but now sure why not success\r\n\r\n\r\ninstead, I upload to [mydrive](https://drive.google.com/drive/folders/1e1xHXZQSEpHBI0YF6xLpUY2Ne8Zi7Cjl?usp=sharing), I will try it again tomorrow see if I miss something.", "Ah, it's probably because you didn't install git-lfs/didn't track the files! Doing this in the repo should help:\r\n```\r\ngit-lfs install\r\ngit-lfs track <name_of_your_large_file>\r\n```\r\nYou can check it's being correctly tracked:\r\n```\r\ngit-lfs track\r\n```\r\ncheck for the name of your large file and ensure it's being tracked by git-lfs:\r\n```\r\nObjects to be committed:\r\n\r\n\t.gitattributes (Git: 31aaf10)\r\n\tREADME.md (Git: 358442a)\r\n\tconfig.json (Git: 57a54a8)\r\n \r\n[...] \r\n\tpytorch_model.bin (LFS: 6a9a9a5)\r\n ^^^\r\n[...]\r\n\tspecial_tokens_map.json (Git: e3ec7ab)\r\n\ttokenizer_config.json (Git: ab033df)\r\n\tvocab.txt (Git: 4d96f93)\r\n\r\n```\r\nThen you should be able to push without any issue", "LGTM! The only issue I encounter is I don't actually know how these file being add. \r\n\r\n``` \r\n\tpytorch_model.bin (LFS: 6a9a9a5)\r\n ^^^\r\n```\r\nMy occasion for track and commit is\r\n\r\n\r\n \r\nI assume this will be shown at the git commit -m \"comment\" phase? Anyway I finally upload my [model](https://huggingface.co/rmxkyz/zh_tf2/tree/main), weeee! \r\n", "Fantastic! I don't have time to look at it today, but I'll try and do that on Monday. Thanks!", "Thanks for your kindly support! I am not rush with this experiment, so do it ease when you're free :D", "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." ]

[]

[]

[ "Hello, thanks for opening an issue! We try to keep the github issues for bugs/feature requests.\r\nCould you ask your question on the [forum](https://discusss.huggingface.co) instead?\r\n\r\nThanks!" ]

[ { "id": 2796628563, "node_id": "MDU6TGFiZWwyNzk2NjI4NTYz", "url": "https://api.github.com/repos/huggingface/transformers/labels/WIP", "name": "WIP", "color": "234C99", "default": false, "description": "Label your PR/Issue with WIP for some long outstanding Issues/PRs that are work in progress" } ]

[ { "login": "patil-suraj", "id": 27137566, "node_id": "MDQ6VXNlcjI3MTM3NTY2", "avatar_url": "https://avatars.githubusercontent.com/u/27137566?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patil-suraj", "html_url": "https://github.com/patil-suraj", "followers_url": "https://api.github.com/users/patil-suraj/followers", "following_url": "https://api.github.com/users/patil-suraj/following{/other_user}", "gists_url": "https://api.github.com/users/patil-suraj/gists{/gist_id}", "starred_url": "https://api.github.com/users/patil-suraj/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patil-suraj/subscriptions", "organizations_url": "https://api.github.com/users/patil-suraj/orgs", "repos_url": "https://api.github.com/users/patil-suraj/repos", "events_url": "https://api.github.com/users/patil-suraj/events{/privacy}", "received_events_url": "https://api.github.com/users/patil-suraj/received_events", "type": "User", "site_admin": false } ]

[ "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", "This Linformer's ForSequenceClassification is not work. I finished mlm pretrain with 1.14loss(100epoch, wiki-cn). But in afqmf, acc only reach 0.6899", "Hi @luoda888,\r\n\r\ngreat to see you're already using Linformer! Did you pretrain Linformer on several GPUs using the `run_mlm.py` script? What is afqmf?\r\n\r\nNote that in the [original paper](https://arxiv.org/abs/2006.04768), they state that \"All models are pretrained with the masked-language-modeling (MLM) objective, and the training for all experiments are parallelized across 64 Tesla V100 GPUs with 250k updates.\"", "@NielsRogge since the weights for `linformer` are not yet released, maybe we can add this model in the the `research_projects` dir, similar to [performer](https://github.com/huggingface/transformers/tree/master/examples/research_projects/performer) and add a training/fine-tuning script there.", "> Hi @luoda888,\r\n> \r\n> great to see you're already using Linformer! Did you pretrain Linformer on several GPUs using the `run_mlm.py` script? What is afqmf?\r\n> \r\n> Note that in the [original paper](https://arxiv.org/abs/2006.04768), they state that \"All models are pretrained with the masked-language-modeling (MLM) objective, and the training for all experiments are parallelized across 64 Tesla V100 GPUs with 250k updates.\"\r\n\r\nThanks reply.\r\n1. I pre-trained the linformer model from scratch, using Wikipedia Chinese corpus. Set text maxlen=128, k=64, on 10 nodes (each with 8*32g V100) (which means I use 80*v100), 2.5 hours of pre-training. Using LinkerForMaskedLM, mlm loss reached 1.14 when training stopped.\r\n2. I use the trained model for validation on downstream tasks, and afqmf is the semantic similarity (text matching) task in Chinese (https://github.com/CLUEbenchmark/CLUE). According to the report, the accuracy of bert-base-chinese is 74.16% when the parameter settings of bs=16, lr=2e-5, maxlen=128, epoch=3 are fine-tuned. I used the parameters of bs=256, lr=2e-5, maxlen=128, epoch=3 and the bert-base-chinese accuracy was 72.81%. However, when the LinformerForSequenceClassifier is used for classification, the AUC is only 0.5, and the accuracy is only 0.6899 (that is, all predictions are 0 categories).\r\n3. For the specific training details, I use the Trainer interface, distributed pre-training. The config of the informer is seq_length=128, LinformerConfig(vocab_size=vocab_size, seq_length=seq_length, share_projection=True, k=seq_length//2) . I use BertWordPieceTokenizer for tokenizer, random mask (DataCollatorForLanguageModeling). with FP16 & train_batch_size=256, gradient_accumulation_steps=1, dataloader_num_workers=4, model.save_pretrained(output_files).\r\n\r\n_**I analyze possible causes.**_\r\n1. The fine-tuning script can run properly, and the bert-base-chinese open source weight is executed as expected.\r\n2. Is it possible that the tokenizer is faulty? Inconsistent in pre-training and fine-tuning.\r\n3. Press BertConfig again to perform pre-training. The fine-tuning effect is also AUC = 0.5. So I'm guessing that Bert Word Piece Tokenizer has some problems with Chinese.\r\n\r\nIf you don't mind, I'd like to share the code with you to locate the problem.", "> @NielsRogge since the weights for `linformer` are not yet released, maybe we can add this model in the the `research_projects` dir, similar to [performer](https://github.com/huggingface/transformers/tree/master/examples/research_projects/performer) and add a training/fine-tuning script there.\r\n\r\nI think it's a very good idea, and it would be nice to have a torch-only version available instead of jax, flax. Also, I've observed that transformers_plus_performers reproduce the performer paper, and I don't know if you've noticed.", "I'd love you to come up with a pre-training script from the beginning on the Chinese Wikipedia and how to fine-tune the Chinese downstream tasks. In addition, it is possible to design more pre-training tasks (SpanMask + SBO, StuncBert(TGS), ASP, etc.) and design more fine-tuned multi-task learning (MT-DNN).\r\n\r\nBTW, nezha is also can add in models. https://github.com/lonePatient/NeZha_Chinese_PyTorch. I test it in Chinese~\r\nBTWWW, PGD & FGM & FreeAT & FreeLB & SMART also can add in modeling_language or fine-tune~~~~", "That's really interesting to read! Wow, 80 V100's, that's a lot.\r\n\r\nSome remarks:\r\n\r\n* I still need someone to verify my implementation of `LinformerSelfAttention` in `modeling_linformer.py`, as I'm currently the only one that implemented it, I'd like to have a second opinion to know for sure my implementation is correct. I used [this implementation](https://github.com/mlpen/Nystromformer/blob/main/code/attention_linformer.py) (from the author of Nyströmformer, who benchmarked several efficient self-attention implementations) as a reference. I also checked @lucidrains' implementation, available [here](https://github.com/lucidrains/linformer). However, as you see the loss going down, it might indeed be a tokenizer issue (I'm fairly sure my implementation is correct). \r\n* In both the original paper and the Nyströmformer implementation, it looks like they rely on a RoBERTa encoder rather than BERT, hence they also use `RobertaTokenizer` rather than `BertTokenizer`. So this is something that might have to be updated (as mentioned at the top of this PR). \r\n* Did you use `BertTokenizer.from_pretrained(\"bert-base-chinese\")`? Or did you train the tokenizer on Chinese Wikipedia before using it?\r\n\r\nAnd indeed, it would be great if we can provide the same functionality for Chinese, and introduce pre-training/fine-tuning scripts for Chinese. ", "> That's really interesting to read! Wow, 80 V100's, that's a lot.\r\n> \r\n> Some remarks:\r\n> \r\n> * I still need someone to verify my implementation of `LinformerSelfAttention` in `modeling_linformer.py`, as I'm currently the only one that implemented it, I'd like to have a second opinion to know for sure my implementation is correct. I used [this implementation](https://github.com/mlpen/Nystromformer/blob/main/code/attention_linformer.py) (from the author of Nyströmformer, who benchmarked several efficient self-attention implementations) as a reference. I also checked @lucidrains' implementation, available [here](https://github.com/lucidrains/linformer). However, as you see the loss going down, it might indeed be a tokenizer issue (I'm fairly sure my implementation is correct).\r\n> * In both the original paper and the Nyströmformer implementation, it looks like they rely on a RoBERTa encoder rather than BERT, hence they also use `RobertaTokenizer` rather than `BertTokenizer`. So this is something that might have to be updated (as mentioned at the top of this PR).\r\n> * Did you use `BertTokenizer.from_pretrained(\"bert-base-chinese\")`? Or did you train the tokenizer on Chinese Wikipedia before using it?\r\n> \r\n> And indeed, it would be great if we can provide the same functionality for Chinese, and introduce pre-training/fine-tuning scripts for Chinese.\r\n\r\n```\r\nlimit_alphabat = 30000\r\nspecial_tokens = [\"[PAD]\", \"[UNK]\", \"[CLS]\", \"[SEP]\", \"[MASK]\"]\r\n\r\ntokenizer = tokenizers.BertWordPieceTokenizer(clean_text=True, handle_chinese_chars=True, strip_accents=True, lowercase=True)\r\ntokenizer.train(\r\n files,\r\n vocab_size = args.vocab_size,\r\n min_frequency = args.min_frequency,\r\n show_progress = True,\r\n special_tokens = special_tokens,\r\n limit_alphabet = limit_alphabat,\r\n wordpieces_prefix = \"##\",\r\n )\r\ntry:\r\n os.mkdir(output_files)\r\nexcept:\r\n pass\r\ntokenizer.save_model(output_files)\r\n```\r\nMultiProcessing for tokenizer (1.3GB corpus is too large). Copy K times for dp mask (follow roberta)\r\n```\r\ndef _convert_to_transformer_inputs(question, answer, tokenizer, max_sequence_length):\r\n \"\"\"Converts tokenized input to ids, masks and segments for transformer (including bert)\"\"\"\r\n \r\n def return_id(str1, str2, truncation_strategy, length):\r\n\r\n inputs = tokenizer.encode_plus(str1, str2,\r\n add_special_tokens=True,\r\n max_length=length,\r\n truncation_strategy=truncation_strategy,\r\n truncation=True\r\n )\r\n \r\n input_ids = inputs[\"input_ids\"]\r\n input_masks = [1] * len(input_ids)\r\n input_segments = inputs[\"token_type_ids\"]\r\n padding_length = length - len(input_ids)\r\n padding_id = tokenizer.pad_token_id\r\n input_ids = input_ids + ([padding_id] * padding_length)\r\n input_masks = input_masks + ([0] * padding_length)\r\n input_segments = input_segments + ([0] * padding_length)\r\n \r\n return [input_ids, input_masks, input_segments]\r\n \r\n input_ids_q, input_masks_q, input_segments_q = return_id(\r\n question, answer, 'longest_first', max_sequence_length)\r\n \r\n return input_ids_q\r\n\r\ndef compute_input_arrays(lines, tokenizer, max_sequence_length):\r\n \r\n input_ids = Parallel(backend='multiprocessing', n_jobs=args.n_jobs, batch_size=512)\\\r\n (delayed(_convert_to_transformer_inputs)(line, None, tokenizer, max_sequence_length) for line in tqdm(lines))\r\n \r\n return [i for i in np.asarray(input_ids, dtype=np.int32)]\r\n\r\ntokenizer = tfs.BertTokenizer.from_pretrained(output_files + '/vocab.txt', maxlen=512)\r\nwith open(files, encoding=\"utf-8\") as f:\r\n lines = [line for line in tqdm(f.read().splitlines()) if (len(line) > 0 and not line.isspace())]\r\ndataset = compute_input_arrays(lines, tokenizer, args.maxlen)\r\n\r\ndp_mask = args.dp_mask\r\nshuffle = deepcopy(dataset)\r\nfor i in tqdm(range(dp_mask)):\r\n random.shuffle(shuffle)\r\n dataset.extend(shuffle)\r\n\r\nnp.save(files + \"-dpmask.npy\", dataset)\r\nlogger.info(\"Sentence Cut Finish\")\r\n```\r\n\r\n1. In many Chinese pre-training models, Roberta Tokenizer is often used as Bert Tokenizer, such as hfl:/roberta-chinese-wwm.\r\n\r\n2. The same problem occurs when I use the BertTokenizer to train the Bert. After the same parameters are trained, the AUC is still 0.5.\r\n\r\n```\r\nclass LineByLineTextDataset(Dataset):\r\n \"\"\"\r\n This will be superseded by a framework-agnostic approach soon.\r\n \"\"\"\r\n\r\n def __init__(self, tokenizer: PreTrainedTokenizer, file_path: str, block_size: int):\r\n assert os.path.isfile(file_path), f\"Input file path {file_path} not found\"\r\n # Here, we do not cache the features, operating under the assumption\r\n # that we will soon use fast multithreaded tokenizers from the\r\n # `tokenizers` repo everywhere =)\r\n\r\n lines = np.load(files + '-dpmask.npy')\r\n if args.debug:\r\n lines = lines[:20000]\r\n \r\n self.examples = [{\"input_ids\": torch.tensor(e, dtype=torch.long)} for e in lines]\r\n\r\n def __len__(self):\r\n return len(self.examples)\r\n\r\n def __getitem__(self, i) -> Dict[str, torch.tensor]:\r\n return self.examples[i]\r\n\r\n\r\nvocab_size = 21128\r\nseq_length = args.max_seq_length\r\nconfig = LinformerConfig(vocab_size=vocab_size, seq_length=seq_length, share_projection=True, k=seq_length//2)\r\nmodel = LinformerForMaskedLM(config=config)\r\n\r\ntokenizer = tfs.BertTokenizer.from_pretrained(output_files + '/vocab.txt', maxlen=seq_length)\r\ndataset = LineByLineTextDataset(tokenizer=tokenizer, file_path=files, block_size=seq_length)\r\ndatacol = tfs.DataCollatorForLanguageModeling(tokenizer=tokenizer, mlm=True, mlm_probability=0.15)\r\n\r\ntrain_args = tfs.TrainingArguments(\r\n output_dir = output_files,\r\n overwrite_output_dir = True,\r\n num_train_epochs = args.train_epochs,\r\n per_device_train_batch_size = args.train_batch_size,\r\n gradient_accumulation_steps = args.gradient_accumulation_steps,\r\n save_steps = 10000,\r\n logging_steps = 500,\r\n save_total_limit = 10,\r\n fp16 = args.fp16,\r\n prediction_loss_only = True,\r\n dataloader_num_workers = args.num_workers,\r\n local_rank = args.local_rank,\r\n disable_tqdm = False,\r\n)\r\n\r\nlogger.info(\"TrainingArguments Init\")\r\nlogger.info(train_args)\r\ndecay_parameters = get_parameter_names(model, [torch.nn.LayerNorm])\r\ndecay_parameters = [name for name in decay_parameters if \"bias\" not in name]\r\noptimizer_grouped_parameters = [\r\n { \"params\": [p for n, p in model.named_parameters() if n in decay_parameters], \"weight_decay\": 0.0,},\r\n { \"params\": [p for n, p in model.named_parameters() if n not in decay_parameters], \"weight_decay\": 0.0,}\r\n]\r\n\r\noptimizer_cls = torch.optim.Adam(optimizer_grouped_parameters, lr=args.learning_rate)\r\nlr_cls = torch.optim.lr_scheduler.CosineAnnealingLR(optimizer_cls, T_max=5, eta_min=0)\r\n\r\ntrainer = tfs.Trainer(\r\n model = model,\r\n args = train_args,\r\n data_collator = datacol,\r\n train_dataset = dataset,\r\n optimizers = (optimizer_cls, lr_cls),\r\n)\r\n\r\ntrainer.train()\r\nmodel.save_pretrained(output_files)\r\n```", "> I still need someone to verify my implementation of LinformerSelfAttention in modeling_linformer.py, as I'm currently the only one that implemented it, I'd like to have a second opinion to know for sure my implementation is correct.\r\n\r\nI would be happy to take a look now. Would you mind opening another PR and adding this in `research_projects/linformer` dir ? Just the modeling and config file should be enough.\r\n\r\nAlso, would it be possible to make the implem more similar to the official implem in `fairseq` ? That way we could compare apples to apples. We could then train a small model in `fairseq` and then port and compare it with this implem , that would make it a bit easier to verify the model.", "> > I still need someone to verify my implementation of LinformerSelfAttention in modeling_linformer.py, as I'm currently the only one that implemented it, I'd like to have a second opinion to know for sure my implementation is correct.\r\n> \r\n> I would be happy to take a look now. Would you mind opening another PR and adding this in `research_projects/linformer` dir ? Just the modeling and config file should be enough.\r\n> \r\n> Also, would it be possible to make the implem more similar to the official implem in `fairseq` ? That way we could compare apples to apples. We could then train a small model in `fairseq` and then port and compare it with this implem , that would make it a bit easier to verify the model.\r\n\r\nlol. I can provide GPUs for testing.", "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", "What happened to this? \r\nasking for a friend :D ", "Hi,\r\n\r\nWell actually I wonder if we could investigate the issue with fine-tuning LinFormer. \r\n\r\n@luoda888 would you be able to pre-train the model on English data (Books/Wikipedia)?", "Hello,\r\nSounds preety cool tbh, \r\ni'll try to tinker with. lets see if i go anywhere", "> Hi,\r\n> \r\n> Well actually I wonder if we could investigate the issue with fine-tuning LinFormer.\r\n> \r\n> @luoda888 would you be able to pre-train the model on English data (Books/Wikipedia)?\r\n\r\nI will try it" ]

[]

[]

[ "Hi @vimarshc, thank you for opening this PR! Could you:\r\n- rebase your PR on the most recent master so that the failing tests don't fail anymore\r\n- run `make fixup` at the root of your repository to fix your code quality issue (More information related to this on step 5 of [this document](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md#start-contributing-pull-requests)", "Awesome!\r\n\r\nWould you like to attempt to add a test for this check? \r\n\r\nWe need to use tiny models so it's fast and I made the suggestions here:\r\nhttps://github.com/huggingface/transformers/issues/10293#issuecomment-784630105\r\n\r\nIf you're not sure how to do it please let me know and I will add a test.\r\n", "Hi @stas00, \r\nI'd like to add the tests myself if that's ok. I have to add the same checks for the `from_pretrained` for Tokenizer however it's not as straightforward. The Tokenizer's `from_pretrained` is written with some assumptions in mind and I'm not entirely sure where to add the check. Here's the `from_pretrained` method for Tokenizers. \r\n\r\n\r\nRegardless, I shall try to add the test for this assertion I've already added and the changes mentioned by @LysandreJik in the next 24 hours. \r\n", "OK, so your change works for the model and the config:\r\n```\r\nPYTHONPATH=src python -c 'from transformers import PegasusForConditionalGeneration; PegasusForConditionalGeneration.from_pretrained(\"patrickvonplaten/t5-tiny-random\")'\r\nTraceback (most recent call last):\r\n File \"<string>\", line 1, in <module>\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/modeling_utils.py\", line 975, in from_pretrained\r\n config, model_kwargs = cls.config_class.from_pretrained(\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/configuration_utils.py\", line 387, in from_pretrained\r\n assert (\r\nAssertionError: You tried to initiate a model of type 'pegasus' with a pretrained model of type 't5'\r\n```\r\nsame for:\r\n```\r\nPYTHONPATH=src python -c 'from transformers import PegasusConfig; PegasusConfig.from_pretrained(\"patrickvonplaten/t5-tiny-random\")'\r\n```\r\n\r\nAs you discovered - and I didn't know - the tokenizer doesn't seem to need the config file, so it doesn't look there is a way to check that the tokenizer being downloaded is of the right kind. I will ask.\r\n\r\nAnd yes, it's great if you can add the test - thank you.\r\n\r\nI restyled your PR to fit our style guide - we don't use `format` and you need to run the code through `make fixup` or `make style` (slower) before committing - otherwise CIs may fail. Which is what @LysandreJik was requesting.\r\nhttps://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md#start-contributing-pull-requests\r\n\r\nSo please `git pull` your branch to get my updates.", "Hi @stas00, \r\nThanks for the update. \r\nWill take a pull, add the test and go through the checklist before pushing the changes. \r\nWill try to push in a few hours. ", "I'm puzzled. why did you undo my fix? If you want to restore it, it was:\r\n\r\n```\r\n--- a/src/transformers/configuration_utils.py\r\n+++ b/src/transformers/configuration_utils.py\r\n@@ -384,6 +384,9 @@ class PretrainedConfig(object):\r\n\r\n \"\"\"\r\n config_dict, kwargs = cls.get_config_dict(pretrained_model_name_or_path, **kwargs)\r\n+ assert (\r\n+ config_dict[\"model_type\"] == cls.model_type\r\n+ ), f\"You tried to initiate a model of type '{cls.model_type}' with a pretrained model of type '{config_dict['model_type']}'\"\r\n return cls.from_dict(config_dict, **kwargs)\r\n\r\n @classmethod\r\n```", "Hi, \r\nApologies. \r\nI rebased my branch and assumed had to force push which deleted your changes. ", "Hi, \r\nI have added the tests. \r\nEverything seems to be working fine. \r\n\r\nHowever, I pushed after taking a pull from the master, and yet it's showing a merge conflict. Not sure how that got there. ", "you messed up your PR branch - so this PR now contains dozens of unrelated changes.\r\n\r\nYou can do a soft reset to the last good sha, e.g.:\r\n```\r\ngit reset --soft d70a770\r\ngit commit\r\ngit push -f\r\n```\r\n\r\nJust save somewhere your newly added test code first.\r\n", "I think you picked the wrong sha and ended up with an even worse situation. Try `d70a770` as I suggested.", "OK, so looking at the errors - need to solve 2 issues:\r\n\r\n### Issue 1.\r\n```\r\n assert (\r\n> config_dict[\"model_type\"] == cls.model_type\r\n ), f\"You tried to initiate a model of type '{cls.model_type}' with a pretrained model of type '{config_dict['model_type']}'\"\r\nE KeyError: 'model_type'\r\n```\r\nso some models don't have the `model_type` key. \r\n\r\n@vimarshc, I suppose you need to edit the code to skip this assert if we don't have the data.\r\n\r\nYou can verify that your change works with this test:\r\n```\r\npytest -sv tests/test_trainer.py::TrainerIntegrationTest -k test_early_stopping_callback\r\n```\r\n\r\nI looked at the config.json generated by this test and it's:\r\n```\r\n{\r\n \"a\": 0,\r\n \"architectures\": [\r\n \"RegressionPreTrainedModel\"\r\n ],\r\n \"b\": 0,\r\n \"double_output\": false,\r\n \"transformers_version\": \"4.4.0.dev0\"\r\n}\r\n```\r\nso far from being complete.\r\n\r\n### Issue 2\r\n\r\nThis one looks trickier:\r\n```\r\nE AssertionError: You tried to initiate a model of type 'blenderbot-small' with a pretrained model of type 'blenderbot'\r\n```\r\n\r\nWe will ask for help with this one.", "@patrickvonplaten, @patil-suraj - your help is needed here.\r\n\r\nBlenderbotSmall has an inconsistency. It declares its model type as \"blenderbot-small\":\r\n```\r\nsrc/transformers/models/auto/configuration_auto.py: (\"blenderbot-small\", BlenderbotSmallConfig),\r\nsrc/transformers/models/auto/configuration_auto.py: (\"blenderbot-small\", \"BlenderbotSmall\"),\r\nsrc/transformers/models/blenderbot_small/configuration_blenderbot_small.py: model_type = \"blenderbot-small\"\r\n```\r\nbut the pretrained models all use `model_type: blenderbot`: https://huggingface.co/facebook/blenderbot-90M/blob/main/config.json\r\n\r\nSo this new sanity check this PR is trying to add fails. \r\n```\r\n config_dict, kwargs = cls.get_config_dict(pretrained_model_name_or_path, **kwargs)\r\n> assert (\r\n config_dict[\"model_type\"] == cls.model_type\r\n ), f\"You tried to initiate a model of type '{cls.model_type}' with a pretrained model of type '{config_dict['model_type']}'\"\r\nE AssertionError: You tried to initiate a model of type 'blenderbot-small' with a pretrained model of type 'blenderbot'\r\n```\r\n\r\nWhat shall we do?\r\n\r\nIt's possible that that part of the config object needs to be re-designed, so that there is a top architecture/type and then perhaps sub-types? \r\n", "Hi @stas00 \r\nWill add the check you mentioned today. \r\n", "Looks good, @vimarshc \r\n\r\nSo we are down to one failing test:\r\n\r\n```\r\ntests/test_modeling_blenderbot_small.py::Blenderbot90MIntegrationTests::test_90_generation_from_short_input\r\n```", "I wonder if we could sort of cheat and do:\r\n\r\n```\r\nif not cls.model_type in config_dict[\"model_type\"]: assert ...\r\n```\r\n\r\nso this will check whether the main type matches as a substring of a sub-type. It's not a precise solution, but will probably catch the majority of mismatches.\r\n\r\nActually for t5/mt5 it's reversed. `model_type` are t5 and mt5, but both may have `T5ForConditionalGeneration` as `architecture`. \r\nhttps://huggingface.co/google/mt5-base/blob/main/config.json#L16 since `MT5ForConditionalGeneration` is a copy of `T5ForConditionalGeneration` with the only difference of having `model_type = \"mt5\"`\r\n\r\nSo I think this check could fail in some situations. In which case we could perhaps check if one is a subset of another in either direction?\r\n\r\n```\r\nif not (cls.model_type in config_dict[\"model_type\"] or config_dict[\"model_type\"] in cls.model_type): assert ...\r\n```\r\n\r\nSo this proposes a sort of fuzzy-match.\r\n\r\n", ">BlenderbotSmall has an inconsistency. It declares its model type as \"blenderbot-small\":\r\n\r\n@stas00 You are right. Before the BART refactor all `blenderbot` models shared the same model class, but the config was not updated after the refactor. The `model_type` on the hub should be `blenderbot-small`. I will fix that.", "I updated the config https://huggingface.co/facebook/blenderbot-90M/blob/main/config.json.\r\n\r\nAnd actually, there's a new version of `blenderbot-90M` , https://huggingface.co/facebook/blenderbot_small-90M\r\n\r\nIt's actually the same model, but with the proper name. The blenderbot small test uses `blenderbot-90M` which should be changed to use this new model.", "Hi @stas00, \r\nThe fuzzy match approach will not work for the case 'distilbert' vs 'bert'. ", "> Hi @stas00,\r\n> The fuzzy match approach will not work for the case 'distilbert' vs 'bert'.\r\n\r\nThat's an excellent counter-example! As I proposed that it might mostly work ;)\r\n\r\nBut it looks like your original solution will now work after @patil-suraj fixing.\r\n\r\nsome unrelated test is failing - I rebased this branch - let's see if it will be green now.", "> I updated the config https://huggingface.co/facebook/blenderbot-90M/blob/main/config.json.\r\n> \r\n> And actually, there's a new version of `blenderbot-90M` , https://huggingface.co/facebook/blenderbot_small-90M\r\n> \r\n> It's actually the same model, but with the proper name. The blenderbot small test uses `blenderbot-90M` which should be changed to use this new model.\r\n\r\nThank you, Suraj!\r\n\r\nSince it's sort of related to this PR, do you want to push the change in here, or do it in another PR?", "Oh bummer, we have 2 more in TF land:\r\n```\r\nFAILED tests/test_modeling_tf_flaubert.py::TFFlaubertModelTest::test_compile_tf_model\r\nFAILED tests/test_modeling_tf_flaubert.py::TFFlaubertModelTest::test_save_load\r\n```\r\nsame issue for both tests:\r\n```\r\nE AssertionError: You tried to initiate a model of type 'xlm' with a pretrained model of type 'flaubert'\r\n```\r\n\r\n@LysandreJik, who can help resolving this one? Thank you!\r\n\r\n\r\n\r\n", "Yes, I'll take a look as soon as possible!", "I fixed the tests related to FlauBERT. Flax test is a flaky test that @patrickvonplaten is working on solving, and should not block this PR.", "Thank you for taking care of this, @LysandreJik \r\n\r\nI suppose we will take care of potentially doing the same for the Tokenizer validation in another PR.", "With the tokenizer it'll likely be a bit more complex, as it is perfectly possible to have decoupled models/tokenizers, e.g., a BERT model and a different tokenizer like it is the case in [BERTweet (config.json)](https://huggingface.co/vinai/bertweet-base/blob/main/config.json).", "Indeed, I think this will require a change where there is a required `tokenizer_config.json` which identifies itself which arch it belongs to, so while it should be possible to mix a model and tokenizer from different architectures, this shouldn't fail with random misleading errors like:\r\n\r\n```\r\npython -c 'from transformers import BartTokenizer; BartTokenizer.from_pretrained(\"prajjwal1/bert-tiny\")'\r\nTraceback (most recent call last):\r\n File \"<string>\", line 1, in <module>\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/tokenization_utils_base.py\", line 1693, in from_pretrained\r\n raise EnvironmentError(msg)\r\nOSError: Can't load tokenizer for 'prajjwal1/bert-tiny'. Make sure that:\r\n\r\n- 'prajjwal1/bert-tiny' is a correct model identifier listed on 'https://huggingface.co/models'\r\n\r\n- or 'prajjwal1/bert-tiny' is the correct path to a directory containing relevant tokenizer files\r\n```\r\n\r\nbut to indicate to the user that they got either the wrong tokenizer class or the the tokenizer identifier, since the above error is invalid - it's the correct identifier \r\n\r\nAs can be seen from:\r\n```\r\npython -c 'from transformers import BertTokenizer; BertTokenizer.from_pretrained(\"prajjwal1/bert-tiny\")'\r\n```\r\nwhich works.\r\n\r\n(and it erroneously says \"model identifier\" and there is no model here, but that's an unrelated minor issue).\r\n\r\nAnd of course there are many other ways I have seen this mismatch to fail, usually a lot noisier when it's missing some file.\r\n", "@LysandreJik, I rebased this PR and it looks good. v4.4.0 is out so we can probably merge this one now.\r\n\r\nThank you.", "Indeed, this is great! Thanks a lot @vimarshc and @stas00 for working on this.", "So should I create a new issue for doing the same for the Tokenizers? I think it'd be much more complicated since we don't save any tokenizer data at the moment that puts the tokenizer in any category/architecture.", "Hi, \r\nThanks, @stas00 for providing the guidance to close this issue. This is my first contribution to transformers so you can imagine my excitement. :D \r\nI understand that a similar change for Tokenizer will be a bit more complicated. Would love to take a shot at fixing that as well. :) " ]

[]

[]

[ "I'm going to merge this since the issue could be interfering with other PRs." ]

[]

[]

[ "It's still hanging waiting for something. If I'm not mistaken it's hanging in saving the model. but there is another thread that might be relevant:\r\n```\r\nThread 0x00007f6dd984b740 (most recent call first):\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/_utils.py\", line 45 in _type\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/cuda/__init__.py\", line 496 in type\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/storage.py\", line 72 in cpu\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/serialization.py\", line 488 in _save\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/serialization.py\", line 372 in save\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/modeling_utils.py\", line 835 in save_pretrained\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1535 in _save\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1496 in save_model\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1204 in _save_checkpoint\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1179 in _maybe_log_save_evaluate\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1088 in train\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/examples/seq2seq/run_seq2seq.py\", line 590 in main\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/examples/seq2seq/run_seq2seq.py\", line 654 in <module>\r\n\r\nThread 0x00007f59389d2740 (most recent call first):\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1552 in store_flos\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/src/transformers/trainer.py\", line 1132 in train\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/examples/seq2seq/run_seq2seq.py\", line 590 in main\r\n File \"/mnt/nvme1/code/huggingface/transformers-test-fix/examples/seq2seq/run_seq2seq.py\", line 654 in <module>\r\n```\r\nRight after:\r\n```\r\nSaving model checkpoint to /tmp/tmp4rujlsox/checkpoint-1 'epoch': 1.0}\r\nstderr: Configuration saved in /tmp/tmp4rujlsox/checkpoint-1/config.json\r\n```\r\n\r\nYou can add:\r\n\r\n```\r\n--- a/examples/seq2seq/run_seq2seq.py\r\n+++ b/examples/seq2seq/run_seq2seq.py\r\n@@ -649,4 +649,6 @@ def _mp_fn(index):\r\n\r\n\r\n if __name__ == \"__main__\":\r\n+ import faulthandler\r\n+ faulthandler.dump_traceback_later(20, repeat=True)\r\n main()\r\n```\r\n\r\nTo get the hanging bt.\r\n\r\nAnd it's best debug outside of pytest, otherwise some bt gets messed up it seems.", "Oh? It was running to completion on my side after the removal. So let's merge as is and I will look into the failure when I have time." ]

[]

[]

[]

[]

[]

[]

[]

[ { "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 } ]

[ "Thanks to a fix for `timit_asr` that @patrickvonplaten made, now I have some good results using `wav2vec2-base`:\r\n\r\n<img width=\"1050\" alt=\"timit_asr_pr_ 10581\" src=\"https://user-images.githubusercontent.com/6879673/110573935-0435e480-8111-11eb-8e0e-845af4e2eab7.png\">\r\n\r\nI'm running one for `arabic_speech_corpus` using `wav2vec2-base` as well. @patrickvonplaten let me know when you have the rest of configs for https://huggingface.co/facebook/wav2vec2-large-xlsr uploaded so I can try it as well (or a workaround). Thanks!", "Thanks for adding this functionality! Your script worked succesfully when I fine-tuned `wav2vec2-base` , `xlsr` (https://dl.fbaipublicfiles.com/fairseq/wav2vec/xlsr_53_56k.pt) and `wa2vec2-large-100k` (multilingual Large Model from https://github.com/facebookresearch/voxpopuli#pre-trained-models pre-trained on VoxPopuli dataset) on TIMIT dataset. If fine-tuning on some another custom dataset, is it enough to set `--orthography` to `timit` in `run_asr.py` if the transcriptions are lowercased and `librispeech` if they are uppercased?", "> Thanks for adding this functionality! Your script worked succesfully when I fine-tuned `wav2vec2-base` , `xlsr` (https://dl.fbaipublicfiles.com/fairseq/wav2vec/xlsr_53_56k.pt) and `wa2vec2-large-100k` (multilingual Large Model from https://github.com/facebookresearch/voxpopuli#pre-trained-models pre-trained on VoxPopuli dataset) on TIMIT dataset.\r\n\r\nThanks, @Getmany1 - you can do me a favor and run it with `arabic_speech_corpus` dataset and `--target_feature_extractor_sampling_rate --orthography buckwalter` on `xlsr` to verify it works with extended vocab. Unfortunately I can't fit `xlsr` on my machine.\r\n\r\n> If fine-tuning on some another custom dataset, is it enough to set `--orthography` to `timit` in `run_asr.py` if the transcriptions are lowercased and `librispeech` if they are uppercased?\r\nFor the most part, yes! Run it with `--verbose_logging` to see how the orthography rules pre-processed the text. Keep us posted!", "`XLSR `model I used didn't work with this setup: the training loss is _nan_ when I tried to fine-tune on Arabic corpus. If I got it correctly, with `--orthography buckwalter` you modify the tokenizer only. However, if you load e.g. \r\n`model = Wav2Vec2ForCTC.from_pretrained(\"facebook/wav2vec2-large-xlsr-53\")`\r\nand check it structure\r\n`print(model.state_dict)`\r\nyou'll see that the last layer of the network is the LM head with the default vocabulary size:\r\n`(lm_head): Linear(in_features=1024, out_features=32, bias=True)`\r\nIf I understand correctly, you need to convert the model manually if you want to have a letter vocabulary different from english.\r\nI converted the fairseq xlsr checkpoint using this script\r\n`transformers/src/transformers/models/wav2vec2/convert_wav2vec2_original_pytorch_checkpoint_to_pytorch.py`\r\ntogether with \"custom\", Swedish letter dictionary, and then succeeded to fine-tune it on a Swedish corpus.\r\nI guess you need to do the same for Arabic in order to have proper LM head on top of the model.", "Yes, thanks! I missed that step (which is similar to `PreTrainedModel.resize_token_embeddings()`). I'm adding a method for that: `Wav2Vec2ForCTC.resize_lm_head()` which simply calls `PreTrainedModel.model._get_resized_lm_head()` and updates model config. The LM head looks good when inspecting the return value. I'll add some unit tests as well.\r\n\r\nNow I'm fine-tuning it on `wav2vec2-base` but I'm not expecting great results given the phonetic differences with Arabic. If you can try it again with `xlsr`, I'd appreciate it!", "Thanks @elgeish, the method seems to work correctly. Now the loss is not `nan` any more during the fine-tuning of the `xlsr` model. Training loss:\r\n\r\n```\r\n{'loss': 584.0067, 'learning_rate': 8.333333333333333e-05, 'epoch': 0.28}\r\n{'loss': 291.7098, 'learning_rate': 0.00016666666666666666, 'epoch': 0.55}\r\n```\r\nValidation loss:\r\n```\r\n{'eval_loss': 374.4364929199219, 'eval_wer': 1.0, 'eval_runtime': 22.3424, 'eval_samples_per_second': 4.476, 'epoch': 0.28}\r\n{'eval_loss': 374.20855712890625, 'eval_wer': 1.0, 'eval_runtime': 22.8504, 'eval_samples_per_second': 4.376, 'epoch': 0.55}\r\n```\r\nPredictions after epoch 0.55:\r\n```\r\n03/12/2021 12:07:54 - DEBUG - __main__ - reference: \"wayaquwlu lEulamA'u <in~ahu min gayri lmuraj~aHi >an tuTaw~ira lbaktiyryA lmuEdiyapu muqAwamapan Did~a lEilAji ljadiyd >al~a*iy >aSbaHa mutAHan biAlfiEl fiy $akli marhamin lil>amrADi ljildiy~api\"\r\n03/12/2021 12:07:54 - DEBUG - __main__ - prediction: \"\"\r\n03/12/2021 12:07:54 - DEBUG - __main__ - reference: \"wayumkinuka lHuSuwlu EalY taTbiyqAtin lilt~adriybAti l>asAsiy~api maj~Anan\"\r\n03/12/2021 12:07:54 - DEBUG - __main__ - prediction: \"\"\r\n```\r\n\r\nUnfortunately, after that I always run out of memory in Colab. The recordings in the Arabic corpus are probably very long. As a result even with `batch_size=1` 16Gb of GPU memory is not enough:\r\n```\r\nRuntimeError: CUDA out of memory. Tried to allocate 80.00 MiB (GPU 0; 15.90 GiB total capacity; 14.80 GiB already allocated; 87.75 MiB free; 14.94 GiB reserved in total by PyTorch)\r\n 2% 1050/54390 [07:16<6:09:57, 2.40it/s]\r\n```\r\nIf you have more than 16Gb of GPU memory available, I can share the `xlsr` model to try it out on your machine.", "I added a `--max_duration_in_seconds` filter. I'm seeing ok results now fine-tuning `wav2vec2-base` after 500 steps, for example:\r\n```\r\nreference: \"wamin tilka ls~ilaE >al$~Ayu lS~iyniy~u wAlwaraqu wAlbAruwdu wAlbuwSilapu\"\r\nprediction: \"wamin tiloka Als~ila>a$~aAyu AS~iyniy~u walowaraqu waAlobaAruwdu waAlobuwSilapu\"\r\n```\r\nI've also started fine-tuning [wav2vec2-large-xlsr-53](https://huggingface.co/facebook/wav2vec2-large-xlsr-53) after adding missing files locally on my machine.\r\n\r\nFor future PRs, I'm thinking of:\r\n* Supporting other languages and examples from [Patrick's awesome blog post](https://huggingface.co/blog/fine-tune-xlsr-wav2vec2), which I need to read\r\n* CER and WER transformations for other languages (e.g., ignoring tashkil in Arabic)\r\n* Supporting [Lhotse](https://lhotse.readthedocs.io/en/latest/corpus.html) datasets, which provide a ton of speech-specific functionality\r\n\r\nBefore adding UTs, I want to check with @patrickvonplaten the code here is on the right track.\r\n@Getmany1 you've been super helpful, thank you!", "> Hey @elgeish,\r\n> \r\n> Thanks a lot for your PR! Sorry that I only reviewed it now. In general I like the changes you have made to `run_wav2vec2.py` :-)\r\n> \r\n> A couple of things, I'd like to change:\r\n> \r\n> * Remove the change from `modeling_wav2vec2.py`. I don't really want a resize lm head method in the model there -> it's too much of an edge case IMO. You can simply first instantiate the processor then use the tokenizer length to load the model with `Wav2Vec2ForCTC.from_pretrained(model_path, vocab_size=len(tokenizer))`. This way we don't need a `resize_lm_head()` method.\r\n> * Can you give me more context on the `buckwalter.json` file?\r\n> * Can you also add more text to the README.md that explains a bit how to use the `run_wav2vec2.py` script? E.g. what does the orthography class do, what is the buckwalter vocab?\r\n\r\nThank you! I responded inline. I'll update `README.md` as well. I think you mean `run_asr.py`, no?", "Great the PR looks good to me now! Thanks a lot for doing this :-) The failures seem unrelated, so I'll rerun the tests again to make the CI happy" ]

[]

[]

[ "I think you may be experiencing a bug that was fixed since then (I would need the whole error message to be sure) so before we dive further, could you see if an [install from source](https://huggingface.co/transformers/installation.html#installing-from-source) solves your problem?", "Hi @sgugger,\r\n\r\nThank you very much for the quick answer, I tested with the installation from source and he it worked !" ]

[ { "id": 2659267025, "node_id": "MDU6TGFiZWwyNjU5MjY3MDI1", "url": "https://api.github.com/repos/huggingface/transformers/labels/DeepSpeed", "name": "DeepSpeed", "color": "4D34F7", "default": false, "description": "" } ]

[ { "login": "stas00", "id": 10676103, "node_id": "MDQ6VXNlcjEwNjc2MTAz", "avatar_url": "https://avatars.githubusercontent.com/u/10676103?v=4", "gravatar_id": "", "url": "https://api.github.com/users/stas00", "html_url": "https://github.com/stas00", "followers_url": "https://api.github.com/users/stas00/followers", "following_url": "https://api.github.com/users/stas00/following{/other_user}", "gists_url": "https://api.github.com/users/stas00/gists{/gist_id}", "starred_url": "https://api.github.com/users/stas00/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/stas00/subscriptions", "organizations_url": "https://api.github.com/users/stas00/orgs", "repos_url": "https://api.github.com/users/stas00/repos", "events_url": "https://api.github.com/users/stas00/events{/privacy}", "received_events_url": "https://api.github.com/users/stas00/received_events", "type": "User", "site_admin": false } ]

[ "@dorost1234, thank you for the kind words. I'm glad to hear it was useful.\r\n\r\nIn general to answer the bulk of your questions - you will find the full documentation here:\r\nhttps://huggingface.co/transformers/master/main_classes/trainer.html#deepspeed\r\n\r\nPlease let me know if you still have any question after reading it.\r\n\r\n> [deepscale] OVERFLOW! Rank 0 Skipping step. Attempted loss scale: 131072.0, reducing to 65536.0\r\n\r\nI shall document this. This is just a warning that DeepSpeed prints when it delays the optimizer stepping due to fp16 dynamic scaling. I definitely want to document that since it's alarming and hard to understand. and how to get the optimizer kick on the first step.\r\n\r\nAlso I started working on a notebook, which you can try - should work on jupyter or colab:\r\nhttps://github.com/stas00/porting/blob/master/transformers/deepspeed/DeepSpeed_on_colab_CLI.ipynb\r\nPlease let me know what you'd like to be added to it.\r\n", "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." ]

[]

[]

[ "hi @akalieren \r\n\r\nthe `parse_dict` or `parse_args_into_dataclasses` methods always return a `tuple` of parsed arguments for each `dataclass` that was used to initialize `HfArgumentParser`. Here you're initializing it with just `TrainingArguments` so `parse_dict` returns a `tuple` of length 1.\r\n\r\nHope this helps.", "Thanks for quick response @patil-suraj,\r\n\r\nI am understanding it returns `tuple` of initializing given data classes; however, my question is that is there a reason why it returns `tuple` instead of just `*outputs`. \r\n\r\nI hope this is not kind of misunderstanding, and grateful about your time.\r\nThanks.\r\n\r\n", "`return *outputs` is not a valid python statement. Also `outputs` here is a list and it's just a good python practice to return return `tuple` instead of `list` when returning multiple values.", "I tried it and realized that you are right. Thanks a lot for your interest and kind responses.\r\n\r\nI am closing this issue. Wish healthy days." ]

[]

[]

[ "solved with installing sentencepiece, I appreciate adding a file mentioning requirements.txt thanks ", "Hi @dorost1234 ,\r\nGlad you resolved the issue. Your `transformers` version is old, we have now added the `sentencepiece` dependency in `requirements.txt`.\r\nhttps://github.com/huggingface/transformers/blob/master/examples/seq2seq/requirements.txt#L2" ]

[]

[]

[ "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." ]

[]

[]

[ "sorry my mistake to use last version of examples ", "hi how you solved this. i also face this error. i can't find \"is_offline_mode\" and \"get_full_repo_name\" under transformers.utils. i use https://github.com/huggingface/transformers/tree/main/examples/pytorch/summarization", "This should work if you are using `transformers` master/main ", "> https://github.com/huggingface/transformers/tree/main/examples/pytorch/summarization\r\n\r\nI miss the same issue. How do you solve it?" ]

[]

[]

[ "How did you create the `nlp_features` function?\r\n\r\nThe sequence length is different due to different tokenization. ", "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." ]

[]

[]

[]

[]

[]

[ "hi @amiyamandal-dev \r\nThank you for your interest in `S2T`. It's still a work in progress and not available on master yet. If you want to try it, you could checkout this PR #10175", "Hey @amiyamandal-dev ,\r\n\r\nThe model is now available on [master](https://huggingface.co/transformers/master/model_doc/speech_to_text.html)! You could install transformers from the source if you want to try it." ]

[]

[]

[ "Hey @HanGuo97,\r\n\r\nWe try to keep the GitHub issues for bug reports. Do you mind asking your question on the forum instead? Also there might already be similar questions on the forum, such as https://discuss.huggingface.co/t/create-a-custom-model-that-works-with-any-pretrained-transformer-body/4186. Thanks!", "Got it, thanks for letting me know!" ]

[]

[]

[]