Datasets:

---

DanielPFlorian

/

Transformers-Github-Issues

like 0

[]

[]

[ "Thanks a ton!" ]

[]

[]

[]

[]

[]

[ "Ok I submitted #8121 for the typo in `_new_step`.\r\n\r\nAs for this PR, I'm thinking the logic should be:\r\n* use folder referenced by the last item returned from `_sorted_checkpoints`\r\n* in case it's empty, we should probably save the current checkpoint locally and upload it (since we specifically requested an upload to wandb)", "After experimenting a bit more:\r\n\r\n1. Should we upload a model only if `_sorted_checkpoints(…)` is non-empty?\r\n * we don't necessarily get the last model (eg save every 100 steps with 520 steps total)\r\n\r\n2. Should we just save current state model at end of training in `args.output_dir + \"\\wandb\"`\r\n * we need to have access to `Trainer.save_model` from `WandbCallback`\r\n * we could decide to use `state.best_model_checkpoint` when present instead\r\n * we ignore any checkpoint\r\n", "@sgugger Do you want me to make an attempt at giving access to the `Trainer` from callbacks or is it a pattern you want to avoid?", "Hi @borisdayma, sorry I took a bit of time to reply on this, we were waiting for the new version of the model hub to materialize before moving forward on this.\r\n\r\nSo! The callbacks aren't 2-way in Transformers because then you have to be very careful about the order of their execution. Here the design was to just allow for callbacks that can read the state, not write, and for any piece of code that needs the write access, users should subclass the `Trainer`. The circular reference is also problematic for memory management so we leave 2-way callbacks for libraries focused on training models, and keep our simple reporting callbacks as they are :-)\r\n\r\nLike you said, you have access to the state with `best_model_checkpoint`. You can also unpack the model from the kwargs and access it. What is in the `Trainer.save_model` method that you need? Worst case scenario, you can even isntantiate an empty Trainer with just the model and the training arguments, and use its `save_model` method.", "The issue with `best_model_checkpoint` is that it does not exist if there's no measurement metric set.\r\nIt could make sense to define it as the last checkpoint in that case.\r\n\r\nThe next issues would then be:\r\n* sometimes no model has been saved yet (maybe not enough epochs) while we want to log the model -> we could accept that it's an issue on the user side and give a warning\r\n* sometimes we may log every 100 steps and run for 180 steps. The last checkpoint is a bit old -> on this aspect I feel like the `Trainer` should automatically save the final step as a checkpoint\r\n\r\nWhat do you think?\r\n\r\nThe alternative would be to completely ignore that logic, let wandb save a model somewhere and upload it. I had not realized we could have access to `model` from the callback (though saving from `Trainer` is better as it handles TPU, save tokenizer, args and may also change in the future).", "I think the most logical is to save the final model, the intermediate checkpoints are there to resume training is something went wrong, or load the best model at the end (which is done before the `on_train_end` event). That's also why we don't always save the model at the end of training, leaving that part to the user in a script.\r\n\r\nIf you use the logic of unpacking the model from the kwargs, you can simply create a new `Trainer` with it which can then save it easily with the `Trainer.save_model` method. Normally the model you unpack is the reference to the real model, so you won't have a `DistributedDataParallel` or something like that, and everything should work smoothly.\r\n", "Finally getting closer!\r\n\r\nFew notes:\r\n* I import Trainer inside my function to avoid circular reference\r\n* I need to find a way to see if I need `Trainer` or `TfTrainer`, should I infer it through `type(model)`\r\n* I use `Trainer.state` as model metadata but maybe it's not that useful. The artifact is associated to a run which already has all the config parameters but it could be useful to relog it, or maybe I should just log the final metrics instead (that I can get through wandb)", "I think it's becoming pretty cool. Here is [an artifact](https://wandb.ai/borisd13/huggingface/artifacts/model/run-test-clm/2fc1855753d610244974) logged with this method.\r\n\r\nThe current limitation is that it only works with Pytorch for now.\r\nAre there any plan for more synergy between `Trainer` or `TfTrainer` or should they be considered independently?", "`TFTrainer` will be reworked in the near future and be a simple wrap around the Keras fit method (and the callbacks will be regular Keras callbacks).", "I adjusted the metadata when we use `load_best_model_at_end`.\r\nIn that case we don't want to log the last metrics but only flos and best metric.", "Small change:\r\n* force commit of last step\r\n* more robust way to get metadata, it will consider any data that has been logged and is a number\r\n\r\nI'm now ready on my side. Feel free to ping me!", "@LysandreJik let me know if you have any comments", "Happy new year everybody!\r\nSince it has already been a while since this PR was made, am I supposed to merge master and verify the tests are passing?\r\nLet me know if I need to do anything on my side." ]

[]

[]

[]

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

[ { "login": "Narsil", "id": 204321, "node_id": "MDQ6VXNlcjIwNDMyMQ==", "avatar_url": "https://avatars.githubusercontent.com/u/204321?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Narsil", "html_url": "https://github.com/Narsil", "followers_url": "https://api.github.com/users/Narsil/followers", "following_url": "https://api.github.com/users/Narsil/following{/other_user}", "gists_url": "https://api.github.com/users/Narsil/gists{/gist_id}", "starred_url": "https://api.github.com/users/Narsil/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Narsil/subscriptions", "organizations_url": "https://api.github.com/users/Narsil/orgs", "repos_url": "https://api.github.com/users/Narsil/repos", "events_url": "https://api.github.com/users/Narsil/events{/privacy}", "received_events_url": "https://api.github.com/users/Narsil/received_events", "type": "User", "site_admin": false }, { "login": "n1t0", "id": 1217986, "node_id": "MDQ6VXNlcjEyMTc5ODY=", "avatar_url": "https://avatars.githubusercontent.com/u/1217986?v=4", "gravatar_id": "", "url": "https://api.github.com/users/n1t0", "html_url": "https://github.com/n1t0", "followers_url": "https://api.github.com/users/n1t0/followers", "following_url": "https://api.github.com/users/n1t0/following{/other_user}", "gists_url": "https://api.github.com/users/n1t0/gists{/gist_id}", "starred_url": "https://api.github.com/users/n1t0/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/n1t0/subscriptions", "organizations_url": "https://api.github.com/users/n1t0/orgs", "repos_url": "https://api.github.com/users/n1t0/repos", "events_url": "https://api.github.com/users/n1t0/events{/privacy}", "received_events_url": "https://api.github.com/users/n1t0/received_events", "type": "User", "site_admin": false } ]

[ "Yes, we need to remove all the hardcoded configuration values of tokenizers in the transformers source code, and upload `tokenizer_config.json` files for all those models.\r\n\r\nAlso cc @n1t0 ", "Very strange, making some tests, the Rust implem is much slower than the Python one...\r\n\r\nMeasure done on my Mac (i7)\r\n\r\n```python\r\nimport time\r\n\r\nfrom transformers import AutoTokenizer\r\n\r\ntext = \"\"\"\r\nIl se déduit des arrêts de la Cour de justice de l’Union européenne du 27 avril 2017 (A-Rosa Flussschiff GmbH, n° C-620/15) et du 6 février 2018 (Ömer Altun, n° C-359/16) que le juge, lorsqu’il est saisi de poursuites pénales du chef de travail dissimulé, pour défaut de déclarations aux organismes de protection sociale, et que la personne poursuivie produit des certificats E101, devenus A1, à l’égard des travailleurs concernés, délivrés au titre de l’article 14, paragraphe 2, sous a, du règlement n° 1408/71, ne peut, à l’issue du débat contradictoire, écarter lesdits certificats que si, sur la base de l’examen des éléments concrets recueillis au cours de l’enquête judiciaire ayant permis de constater que ces certificats avaient été obtenus ou invoqués frauduleusement et que l’institution émettrice saisie s’était abstenue de prendre en compte, dans un délai raisonnable, il caractérise une fraude constituée, dans son élément objectif, par l’absence de respect des conditions prévues à la disposition précitée et, dans son élément subjectif, par l’intention de la personne poursuivie de contourner ou d’éluder les conditions de délivrance dudit certificat pour obtenir l’avantage qui y est attaché.\r\nDoit ainsi être cassé l’arrêt de la cour d’appel qui écarte les certificats E101 sans avoir, au préalable, recherché si l’institution émettrice desdits certificats avait été saisie d’une demande de réexamen et de retrait de ceux-ci sur la base des éléments concrets recueillis dans le cadre de l’enquête judiciaire permettant, le cas échéant, de constater que ces certificats avaient été obtenus ou invoqués de manière frauduleuse et que l’institution émettrice s’était abstenue, dans un délai raisonnable, de les prendre en considération aux fins de réexamen du bien-fondé de la délivrance desdits certificats, et dans l’affirmative, sans établir, sur la base de l’examen des éléments concrets et dans le respect des garanties inhérentes au droit à un procès équitable, l’existence d’une fraude de la part de la société poursuivie, constituée, dans son élément matériel, par le défaut, dans les faits de la cause, des conditions prévues à l’article 14, paragraphe 2, sous a, précité aux fins d’obtention ou d’invocation des certificats E101 en cause et, dans son élément moral, par l’intention de ladite société de contourner ou d’éluder les conditions de délivrance dudit certificat pour obtenir l’avantage qui y est attaché (arrêt n° 1, pourvoi 13-88.631, arrêt n° 2, pourvoi 13-88.632 et arrêt n° 3, pourvoi n° 15-80.735).\r\nEn revanche, prononce par des motifs conformes à la doctrine de la Cour de l’Union européenne précitée, la cour d’appel qui, pour relaxer les prévenues, sociétés d’aviation civile, énonce que l’enquête n’ a pas permis de constater les éléments de fraude et s’abstient, en conséquence, d’opérer une vérification relative aux certificats E101 produits par elles (arrêt n° 4, pourvoi n° 1581316).\r\n\"\"\"\r\n\r\nfast = False\r\nrepeat = 1000\r\n\r\n# use_fast\r\ntokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=\"camembert/camembert-base-ccnet\", use_fast=fast)\r\n\r\n_ = tokenizer(text)\r\nstart = time.time()\r\nfor _ in range(repeat):\r\n _ = tokenizer(text)\r\nprint(\"ccnet new\", time.time() - start)\r\n\r\n# CCNET Camembert saved few months ago\r\ntokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=\"output/model\", use_fast=fast)\r\n\r\n_ = tokenizer(text)\r\nstart = time.time()\r\nfor _ in range(repeat):\r\n _ = tokenizer(text)\r\nprint(\"ccnet old\", time.time() - start)\r\n\r\ntokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=\"camembert-base\", use_fast=fast)\r\n\r\n_ = tokenizer(text)\r\nstart = time.time()\r\nfor _ in range(repeat):\r\n _ = tokenizer(text)\r\nprint(\"camembert base\", time.time() - start)\r\n\r\n\r\n```\r\n\r\nfast = False\r\n```\r\nwandb: WARNING W&B installed but not logged in. Run `wandb login` or set the WANDB_API_KEY env variable.\r\nccnet new 2.104267120361328\r\nccnet old 2.3693552017211914\r\nToken indices sequence length is longer than the specified maximum sequence length for this model (684 > 512). Running this sequence through the model will result in indexing errors\r\ncamembert base 2.245959997177124\r\n\r\nProcess finished with exit code 0\r\n\r\n```\r\n\r\n\r\nfast = True\r\n```\r\nwandb: WARNING W&B installed but not logged in. Run `wandb login` or set the WANDB_API_KEY env variable.\r\nccnet new 2.7245991230010986\r\nccnet old 2.7714219093322754\r\ncamembert base 2.9007809162139893\r\n```\r\n\r\nIt appears that fast tokenizer... is much slower than Python implementation on a Mac (measures not done a Linux machine).", "Thank you for reporting this @pommedeterresautee. \r\n\r\nI think this is expected though: You are comparing a tokenizer that is based on SentencePiece (c++) with one in Rust. Our rust implementation is a bit slower than the SentencePiece when encoding a single sentence, but as soon as you are starting to encode batches with padding and post-processing, it gets faster!", "Thank you, I didn't understood that batching was a thing during tokenization too!\r\n\r\n```python\r\nimport time\r\n\r\nfrom transformers import AutoTokenizer\r\n\r\ntext = \"\"\"\r\nIl se déduit des arrêts de la Cour de justice de l’Union européenne du 27 avril 2017 (A-Rosa Flussschiff GmbH, n° C-620/15) et du 6 février 2018 (Ömer Altun, n° C-359/16) que le juge, lorsqu’il est saisi de poursuites pénales du chef de travail dissimulé, pour défaut de déclarations aux organismes de protection sociale, et que la personne poursuivie produit des certificats E101, devenus A1, à l’égard des travailleurs concernés, délivrés au titre de l’article 14, paragraphe 2, sous a, du règlement n° 1408/71, ne peut, à l’issue du débat contradictoire, écarter lesdits certificats que si, sur la base de l’examen des éléments concrets recueillis au cours de l’enquête judiciaire ayant permis de constater que ces certificats avaient été obtenus ou invoqués frauduleusement et que l’institution émettrice saisie s’était abstenue de prendre en compte, dans un délai raisonnable, il caractérise une fraude constituée, dans son élément objectif, par l’absence de respect des conditions prévues à la disposition précitée et, dans son élément subjectif, par l’intention de la personne poursuivie de contourner ou d’éluder les conditions de délivrance dudit certificat pour obtenir l’avantage qui y est attaché.\r\nDoit ainsi être cassé l’arrêt de la cour d’appel qui écarte les certificats E101 sans avoir, au préalable, recherché si l’institution émettrice desdits certificats avait été saisie d’une demande de réexamen et de retrait de ceux-ci sur la base des éléments concrets recueillis dans le cadre de l’enquête judiciaire permettant, le cas échéant, de constater que ces certificats avaient été obtenus ou invoqués de manière frauduleuse et que l’institution émettrice s’était abstenue, dans un délai raisonnable, de les prendre en considération aux fins de réexamen du bien-fondé de la délivrance desdits certificats, et dans l’affirmative, sans établir, sur la base de l’examen des éléments concrets et dans le respect des garanties inhérentes au droit à un procès équitable, l’existence d’une fraude de la part de la société poursuivie, constituée, dans son élément matériel, par le défaut, dans les faits de la cause, des conditions prévues à l’article 14, paragraphe 2, sous a, précité aux fins d’obtention ou d’invocation des certificats E101 en cause et, dans son élément moral, par l’intention de ladite société de contourner ou d’éluder les conditions de délivrance dudit certificat pour obtenir l’avantage qui y est attaché (arrêt n° 1, pourvoi 13-88.631, arrêt n° 2, pourvoi 13-88.632 et arrêt n° 3, pourvoi n° 15-80.735).\r\nEn revanche, prononce par des motifs conformes à la doctrine de la Cour de l’Union européenne précitée, la cour d’appel qui, pour relaxer les prévenues, sociétés d’aviation civile, énonce que l’enquête n’ a pas permis de constater les éléments de fraude et s’abstient, en conséquence, d’opérer une vérification relative aux certificats E101 produits par elles (arrêt n° 4, pourvoi n° 1581316).\r\n\"\"\"\r\n\r\n\r\nrepeat = 1000\r\n\r\n# use_fast\r\ntokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=\"camembert/camembert-base-ccnet\", use_fast=True)\r\n\r\n_ = tokenizer(text)\r\nstart = time.time()\r\nfor _ in range(repeat):\r\n _ = tokenizer([text] * 10)\r\nprint(\"fast\", time.time() - start)\r\n\r\ntokenizer = AutoTokenizer.from_pretrained(pretrained_model_name_or_path=\"camembert/camembert-base-ccnet\", use_fast=False)\r\n\r\n_ = tokenizer(text)\r\nstart = time.time()\r\nfor _ in range(repeat):\r\n _ = tokenizer([text] * 10)\r\nprint(\"slow\", time.time() - start)\r\n\r\n```\r\n\r\nProduces\r\n```\r\nwandb: WARNING W&B installed but not logged in. Run `wandb login` or set the WANDB_API_KEY env variable.\r\nfast 16.272130966186523\r\nslow 22.52426290512085\r\n```\r\n... as expected!", "@pommedeterresautee \r\n\r\nHi, I am not sure it's a `fast` tokenizers bug but maybe more a property that was (maybe unintentionnally) dropped from Tokenizers.\r\n\r\n```python\r\nfrom transformers import AutoTokenizer\r\ntokenizer = AutoTokenizer.from_pretrained(\"camembert/camembert-base-ccnet\", use_fast=False)\r\ntokenizer.model_max_length\r\n# Out[4]: 1000000000000000019884624838656\r\n```\r\n\r\nCan you tell us what's the actual bug for you in the end ? Just to make sure the fix I am working on will actually work as generally as possible", "I think you are right, it's more about a dropped property in the config file or a change in the source code than a bug specific to the fast tokenizer.\r\nI discovered the issue because I was comparing the model + tokenizer as exported few months ago with the fast tokenizer of today and thought it was because of the fast tokenizer. My \"old\" export returns me 512 when I call `max_len`.\r\n\r\nStill it's not returning the correct value, fast tokenizer or not.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "Hi, this issue is still not resolved; i have the same problem with `camembert/camembert-large`" ]

[]

[]

[ "Hi there! Thanks for your PR! I see a few problems with this approach.\r\n1. Not all labels need to be padded. If you are doing classification (with one or multiple labels) you don't want to pad them\r\n2. I imagine you are in a token classification problem, and in those, the number of labels is not necessarily the same as the number of tokens, as the labels are for words and tokens can be parts of words.\r\n\r\nI think the proper fix is to create an option in `DataCollatorWithPadding` to activate label padding (so a flag `pad_labels_too` or something like that) that then pads the labels to the maximum length of the labels (so `difference` that you use here might be a different number for the labels).", "Thanks for the reply!\r\n\r\nConsidering that different problem may pad labels differently, I think maybe it's better to leave it as is and use this:\r\n```python\r\nclass MyDataCollatorWithPadding(DataCollatorWithPadding):\r\n def __call__(self, features: List[Dict[str, Union[List[int], torch.Tensor]]]) -> Dict[str, torch.Tensor]:\r\n batch = super().__call__(features)\r\n # add custom label padding here\r\n return batch\r\n```\r\nJust came up with this. 😃 Not sure if it works.", "Just tried it, the above code does not work, because the error is in `self.tokenizer.pad()`.\r\nHere is the truncated trace:\r\n```\r\nsrc/transformers/data/data_collator.py\", line 103, in __call__\r\n batch = self.tokenizer.pad(\r\nsrc/transformers/tokenization_utils_base.py\", line 2408, in pad\r\n return BatchEncoding(batch_outputs, tensor_type=return_tensors)\r\nsrc/transformers/tokenization_utils_base.py\", line 186, in __init__\r\n self.convert_to_tensors(tensor_type=tensor_type, prepend_batch_axis=prepend_batch_axis)\r\nsrc/transformers/tokenization_utils_base.py\", line 571, in convert_to_tensors\r\n raise ValueError(\r\nValueError: Unable to create tensor, you should probably activate truncation and/or padding with 'padding=True' 'truncation=True' to have batched tensors with the same length.\r\n```\r\n\r\nTherefore `pad_labels_too` needs to be in `tokenizer.pad()`.\r\n@sgugger \r\n> the number of labels is not necessarily the same as the number of tokens, as the labels are for words and tokens can be parts of words.\r\n\r\nMaybe we will need a `LabelPaddingStrategy` similar to `PaddingStrategy`. But I don't know what kinds of other label padding strategies needs to be added.", "I think you should use the newly pushed DataCollatorForTokenClassification from #8274.", "Very nice! I guess I will close this PR." ]

[]

[]

[ "Before we merge you'll need to run `make fixup` or `make style` at the root of your transformers clone to pass the code quality test.", "I had not realized that LTP has pinned master to 3.2.0. We can't have a script in examples that doesn't run on master, so I suggest copying the current version and moving it in the examples/contrib folder (or hosting it on your GitHub if you prefer) while still linking to it from the README.\r\n\r\nWe are in the process of rewriting all examples (and this script as it is will change in the next few days) to match the current version of transformers/datasets so this master requirement is really important. ", "> I had not realized that LTP has pinned master to 3.2.0. We can't have a script in examples that doesn't run on master, so I suggest copying the current version and moving it in the examples/contrib folder (or hosting it on your GitHub if you prefer) while still linking to it from the README.\r\n> \r\n> We are in the process of rewriting all examples (and this script as it is will change in the next few days) to match the current version of transformers/datasets so this master requirement is really important.\r\n\r\nSeem the requirements of LTP is fixed. \r\nSo I move to `run_chinese_ref.py` to `examples/contrib` and update readme.", "Great! If you can just merge Lysandre's suggestion, this should be good to merge then.", "Applied the suggestion, merging!" ]

[]

[ { "login": "sshleifer", "id": 6045025, "node_id": "MDQ6VXNlcjYwNDUwMjU=", "avatar_url": "https://avatars.githubusercontent.com/u/6045025?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sshleifer", "html_url": "https://github.com/sshleifer", "followers_url": "https://api.github.com/users/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "(This issue has nothing to do with pegasus or input dimension).\r\n\r\n+ `max_length` should not be set like that, it refers to the maximum length to generate.\r\n+ You probably don't want to initialize from random. \r\ne+ You should be using Seq2SeqTrainer, and this script https://github.com/huggingface/transformers/blob/master/examples/seq2seq/finetune_trainer.py#L2\r\n\r\nYou may need to modify line 188 to pass `max_position_embeddings=2048` and line 189 to pass `model_max_length=2048`." ]

[]

[]

[ "I think you might need a rebase on the latest master as your PR seems to have taken master from a while ago (all the modifications in the README should not be there for instance). If it messes the diff, we can always close this PR and open a new one, the branch will be safe :-)", "EDIT: I've always used `git fetch upstream` and `git rebase upstream/master`, before pushing my local `tapas_v3` branch to my fork on Github. I didn't know that all models starts with 1. now in the README (therefore I manually changed the numbering), should be fixed now, branch is up-to-date.", "@sgugger should I keep rebasing this branch everyday, to keep up with master (as long as the code is not being reviewed)?\r\n\r\nAlso, is it normal that I have to do a force push everytime I perform a rebase and want to push to Github? Because when I want to do simply `git push -u origin tapas_v3`, I always get \r\n```\r\n(env) PS C:\\Users\\niels.rogge\\Documents\\Python projecten\\transformers> git push -u origin tapas_v3\r\nTo https://github.com/NielsRogge/transformers.git\r\n ! [rejected] tapas_v3 -> tapas_v3 (non-fast-forward)\r\nerror: failed to push some refs to 'https://github.com/NielsRogge/transformers.git'\r\nhint: Updates were rejected because the tip of your current branch is behind\r\nhint: its remote counterpart. Integrate the remote changes (e.g.\r\nhint: 'git pull ...') before pushing again.\r\nhint: See the 'Note about fast-forwards' in 'git push --help' for details.\r\n```\r\n\r\nafter a local rebase. ", "I'm far from being an expert on git and I don't use the command line anyway, so can't really help you with that.", "> Hi! Thanks a lot @NielsRogge for implementing this model! It's a great model that definitely deserves its place in the huggingface library, and you did a great job implementing it! I'm reviewing the code below.\r\n\r\nThank you! Great to hear :)\r\n\r\n@LysandreJik I addressed all of the comments. To summarize:\r\n* `TapasConfig` and `TapasTokenizer` now inherit from `PretrainedConfig` and `PreTrainedTokenizer` respectively. For `TapasTokenizer`, a lot of the code of `tokenization_bert.py` was copied (such as the `BasicTokenizer` and `WordPieceTokenizer` classes), since the tokenization logic of text itself is the same. However, some things are different (see `tokenization_tapas.py`). \r\n* `Modeling_tapas_utilities.py` and `tokenization_tapas_utilities.py` are also gone now, they are added to the bottom of `modeling_tapas.py` and `tokenization_tapas.py` respectively. \r\n* `pandas` is not a dependency in the code (`pandas` is not imported in any of the files), but I assume you want to switch to `datasets` so that people can use Tapas using only the `Transformers` library? However, currently, in `tokenization_tapas.py` some Pandas logic is used, for example in the `_tokenize_table` function `.iterrows()` is used, so this will require some changes.\r\n* concerning git, I assume I should stop rebasing at some point? I can do it as long as I'm the only one committing?", "Hey @LysandreJik, thank you for your feedback.\r\n\r\nI've fixed all comments that you had, apart from the tokenizer itself. \r\n\r\n> This is a complicated part, so please let us know if you would like some help along the way/if you want us to take over from here, in which case we would open PRs against your branch with proposals.\r\n\r\nI'm happy to accept PRs against my branch, because it's not that clear to me how the tokenizer should be implemented in the best possible way. Does that mean I should stop rebasing my branch with `upstream/master`? Since I read about the \"golden rule of rebasing\", which states to \"never use it on public branches\" 😄 \r\n\r\n", "Alright, I'll take a look and open a PR on your fork with the proposed changes. Yes, please don't rebase on `master` anymore as it would mess up my history as I start working on your branch.\r\n\r\nWe can keep it as it is until we merge now, and fix the merge conflict as the last step.", "Hi @NielsRogge, I'm nearly done with the tokenizer changes, but we're focusing on getting version v3.5.0 out today and tomorrow. I'll try to open a PR on your repository then. Please hold off from adding commits now, as rebasing/merging would be very painful now! :smile: ", "@LysandreJik no worries, I'm not working on the `tapas_v3` branch.\r\n\r\nHowever, there's still some important work left to do in terms of preparing the data for the model. In the original implementation, this is done in 3 steps:\r\n\r\n## 1. TSV in the SQA format\r\nAny dataset (SQA, WTQ, WikiSQL) is first transformed into a TSV with the same columns as the SQA format:\r\n* id: id of the table-question pair, if applicable.\r\n* annotator: id of the annotator, if applicable.\r\n* position: integer indicating if the question is the first, second, third,... related to the table. Only required in case of conversational setup (such as SQA)\r\n* question: string\r\n* table_file: string, name of a csv file containing the tabular data\r\n* answer_text: list of strings (each string being a cell value that is part of the answer)\r\n* answer_coordinates: list of string tuples (each tuple being a cell coordinate, i.e. row, column pair that is part of the answer)\r\n* aggregation_label: only required in case of strong supervision for aggregation (such as WikiSQL-supervised)\r\n* answer_float: float answer to the question. Only required in case of weak supervision for aggregation (such as WTQ and WikiSQL)\r\n\r\nIf people want to fine-tune `TapasForQuestionAnswering` on their own dataset, they must prepare it in this TSV format, and associated csv files containing the tabular data. It would be great if we can upload all datasets (SQA, WTQ, WikiSQL and WikiSQL-supervised) in SQA format to the HuggingFace datasets hub (they can easily be obtained from the official Tapas Github repo).\r\n\r\n## 2. Intermediate format: Interaction\r\nNext, each table-question pair is transformed into an intermediate protocol buffer message which the authors call **Interaction**. Its properties are defined [here](https://github.com/google-research/tapas/blob/master/tapas/protos/interaction.proto), and include things like Table, Question, Answer, AnswerCoordinate, Cell, NumericValue, NumericValueSpan, etc. Populating all the fields of an Interaction based on the TSV is defined in [interaction_utils.py](https://github.com/google-research/tapas/blob/master/tapas/utils/interaction_utils.py), [interaction_utils_parser.py](https://github.com/google-research/tapas/blob/master/tapas/utils/interaction_utils_parser.py), [number_annotation_utils.py](https://github.com/google-research/tapas/blob/master/tapas/utils/number_annotation_utils.py), [number_utils.py](https://github.com/google-research/tapas/blob/master/tapas/utils/number_utils.py) and [text_utils.py](https://github.com/google-research/tapas/blob/master/tapas/utils/text_utils.py).\r\n\r\n## 3. tf.train.Example\r\nFinally, each interaction is transformed into an actual training example (`tf.train.Example`), containing the input_ids, mask, etc. as `tf.train.Feature` objects. This is defined in [tf_example_utils.py](https://github.com/google-research/tapas/blob/master/tapas/utils/tf_example_utils.py).\r\n\r\n_________________\r\n`TapasTokenizer` must be able to directly convert a row (or multiple rows, i.e. a batch) from a TSV file into a dictionary with PyTorch tensors as values (in other words, combine steps 2 and 3). The remaining work is basically step 2. As I worked with `Pandas` as standard format for tables in my implementation, my idea was to define regular Python classes for each property of the Interaction proto. That is why I have defined a `NumericValue` class, `NumericValueSpan` class, `Cell` class, `Date` class, etc. in `tokenization_tapas.py`. Instances of these classes are then created each time `TapasTokenizer` is called.\r\n\r\nI've noticed that the creation of the numeric values is not entirely correct in the `tapas_v3` branch. I'm now working on a correct implementation of this in a branch called `tapas_v3_up_to_date_with_master` (in which I also regularly rebase with upstream/master). It only involves changes to `tokenization_tapas.py`. The changes can eventually be added to `tapas_v3`. I'll wait until your PR is merged before I add those changes.\r\n\r\nSo my questions are:\r\n- for each of the `xxx_utils.py` files which are used in step 2, there are corresponding `xxx_utils_test.py` files. Could you help in setting up tests in `test_tokenization_tapas.py`, to make sure we're following the original implementation? \r\n- I'm still assuming that tables are `pandas` dataframes in `tokenization_tapas.py`. Is this OK? Or do you want to change to `dataset`? Wouldn't it be more logical to have SQA/WTQ/WikiSQL as `dataset` objects, and the actual tables as `pandas` dataframes? Pandas is not a dependency of `tokenization_tapas.py`, but tables must be provided as a Pandas dataframe to `TapasTokenizer`.\r\n\r\n\r\n\r\n", "Hi @NielsRogge! I just finished the tokenizer and its tests. The tests were kind of painful, as the API is a bit different (accepting a dataframe instead of a string), so I had to override most of the tests.\r\n\r\nHere's how you can review:\r\n\r\n- I did a `make style` on your branch, as I find the code easier to navigate once it's on par with our style.\r\n- However, this introduces a bunch of changes that would make the PR hard to review.\r\n- In order to circumvent this, I've pushed two branches: \r\n - `tapas-style`, which is the exact branch you have, but with `make style` run on it and a few cosmetic adjustments\r\n - `tapas-final`, which builds on top of `tapas-style` to implement all the tokenizer API and tests\r\n- For you to review, the easiest would be to review the PR I opened [here](https://github.com/huggingface/transformers/pull/8482), which aims to merge `tapas-final` into `tapas-style`. This way you can see the actual changes and only these.\r\n- I described my changes in that PR's description so as not to clog this one.\r\n- I setup a todo list if items remaining on the tokenizer, which are not blocking for the merge.\r\n\r\nPlease review https://github.com/huggingface/transformers/pull/8482, and tell me if you're okay with the changes. If you're okay, I'll merge `tapas-final` into `tapas-style`, and open a PR on your fork with the branch `tapas-style`, which will have all the changes.\r\n\r\nThis is one of the best hands-on introduction to git you could ask for :smile: \r\n\r\n\r\nRegarding your questions about data processing:\r\n\r\n> for each of the xxx_utils.py files which are used in step 2, there are corresponding xxx_utils_test.py files. Could you help in setting up tests in test_tokenization_tapas.py, to make sure we're following the original implementation?\r\n\r\nYes, I can help you with that.\r\n\r\n> I'm still assuming that tables are pandas dataframes in tokenization_tapas.py. Is this OK? Or do you want to change to dataset? Wouldn't it be more logical to have SQA/WTQ/WikiSQL as dataset objects, and the actual tables as pandas dataframes? Pandas is not a dependency of tokenization_tapas.py, but tables must be provided as a Pandas dataframe to TapasTokenizer.\r\n\r\nActually, `datasets.Dataset` behave very similarly to `pd.DataFrame`s. Nevertheless, we can start with Pandas DataFrames for now, and change to dataset's Datasets once progress is made", "@LysandreJik I have improved the parsing of numeric values of both the question and table in `prepare_for_model` of `tokenization_tapas.py` to reflect the original implementation. What it does is turn the cells of a table into `Cell` objects (with potentially associated `NumericValue` objects) and the question into a `Question` object (with potentially a list of associated `NumericValueSpan` objects), before adding numeric-related features.\r\n\r\nBesides this, I have fixed some of the comments I had on [my review of your PR](https://github.com/huggingface/transformers/pull/8482), and commented \"done\" on the ones that are fixed.\r\n\r\n## To do:\r\n\r\n- [x] Add correct implementation of `prev_label_ids` in case of a batch of table-question pairs (in case of a batch, all questions should refer to the same table). The implementation should reflect the [original implementation](https://github.com/google-research/tapas/blob/d8638f0909b3de32a85fe7491769d47d645d8e22/tapas/utils/tf_example_utils.py#L1155) as follows: for a given table-question pair in a batch, \r\n\r\n```\r\nprev_label_ids = self.get_answer_ids(\r\n column_ids, row_ids, table_data, answer_text, answer_coordinates\r\n )\r\n```\r\n\r\nHere, it's important that the `get_answer_ids` function is called with the `column_ids` and `row_ids` of the **current** table-question pair in the batch, but the `answer_text` and `answer_coordinates` of the **previous** table-question pair in the batch. \r\n\r\n- [x] Fix the error that I'm currently having in the colab notebooks above (see first message of this PR), when `answer_coordinates` and `answer_text` are provided to the tokenizer. However, what's weird is that when calling `TapasTokenizer` on the real SQA dev set, everything works fine. Might be that I'm doing something wrong with the coordinates and text I provide?\r\n- [x] Add support for the `drop_rows_to_fit` and `cell_trim_length` attributes of `TapasTokenizer`, which should reflect the original API (see also my [suggestion](https://github.com/huggingface/transformers/pull/8482#discussion_r522131827) on how this could be done for `cell_trim_length`). Also, setting `truncation=True` in `TapasTokenizer` doesn't do anything currently. \r\n- [x] I've added support for the special [EMPTY] token for empty cells in a table (based on the `format_text` method, see [here](https://github.com/google-research/tapas/blob/4908213eb4df7aa988573350278b44c4dbe3f71b/tapas/utils/tf_example_utils.py#L330)). Does this have implications for the `add_special_tokens` method? I assume not? What about `get_special_tokens_mask`? To be verified.\r\n- [x] **Testing TapasTokenizer:** make sure that the PyTorch tensors that `TapasTokenizer` creates are exactly the same as those of the original implementation on the same input data. I've created a [notebook](https://colab.research.google.com/drive/1MzyO-QSA5PZNCNoWa2EIrSA8UEqyUZVp) that tests this. Currently there's a misalignment due to the fact that the original implementation tokenizes a cell value like \"1.0\" into [\"1\", \".\"], whereas my implementation tokenizes this into [\"1\", \".\", \"0\"]. Filed a Github issue to resolve this.\r\n- [x] **Testing forward pass:** make sure that `TapasForQuestionAnswering`/`TapasForSequenceClassification` return the same `sequence_output`, `pooled_output`, `logits`, and `loss` tensors as the original implementation on the same input data. I've created notebooks that test this:\r\n- SQA (`tapas_sqa_inter_masklm_base_reset`): [PyTorch](https://colab.research.google.com/drive/14bdSwdzvCF2gDF3L0z58IT1fSNzOXKey#scrollTo=6fvJbFF-xKfh) vs [Tensorflow](https://colab.research.google.com/drive/1KWD187cWDP-lOOwKjwzGtGfVR9UWzZVL#scrollTo=KTlX8ZEuRTBa) is giving me the same output (inference only). UPDATE: also loss calculation is OK, see [PyTorch](https://colab.research.google.com/drive/1z2ZRIBXOTk3Aqh6OYpRak6iJhs7e1l2S#scrollTo=2kakMASqmrG5) vs [Tensorflow](https://colab.research.google.com/drive/1Ba3jARJcAqRTd0uuPxOTjOrruIcXIV54#scrollTo=2ZIdZEJGw5RK).\r\n- WTQ (`tapas_wtq_wikisql_sqa_inter_masklm_base_reset`): [PyTorch](https://colab.research.google.com/drive/1Z4T9ZzMvg3vGZ3dSNWbiA4FVbgMMkq_9#scrollTo=EXS4MmCy8Dti) vs [Tensorflow](https://colab.research.google.com/drive/1klaSP99q2aicwpVV9GrmL5nvrPGqrSPH#scrollTo=SIE7bTJMVuSh). I'm getting the same `sequence_output` and `logits_aggregation` on the same input data :) UPDATE: also loss calculation is OK, see [PyTorch](https://colab.research.google.com/drive/19Uq6k1f1178okv80Julfa0Zg41fvFN9x#scrollTo=LEOCtWmWt2IH) vs [Tensorflow](https://colab.research.google.com/drive/1ScF4R7Au8gbC5lN1ehTQFdDknmr2mMRz#scrollTo=GLgez6jJx9Xc) notebooks.\r\n- Tabfact (`tapas_tabfact_inter_masklm_base_reset`): [PyTorch](https://colab.research.google.com/drive/1JDwWrwHSt8KhGBQ57BDlCFZEe0xMGTQA?usp=sharing#scrollTo=z6esPfPMFH1p) vs [Tensorflow](https://colab.research.google.com/drive/14-6VFjvrIiXsYPQEtv8MN-a8Mpo1UNH7#scrollTo=LYBkiqo38e7l) is giving me the same classification logits, confirming that the relative position embeddings implementation is OK. \r\n- [x] **Testing backward pass:** I've created a [notebook](https://colab.research.google.com/drive/17L97m7cq7J_pnUHmQW6-ksGGVpbbGryP#scrollTo=y0YzoGY24I0C) that fine-tunes `TapasForQuestionAnswering` on 10 examples of the WTQ test set, just to see if it's able to overfit them. I've tried both with randomly initialized classification heads, as well as with the already-finetuned WTQ model. This seems to work well for the former (it can overfit the cell selection, however for aggregation this seems more difficult - probably due to the weak supervision). However, for the already fine-tuned one, the loss stays zero after the third example already. Is this a bug, or is this possible? Update: confirmed by the author.\r\n\r\nActually, reviewing the code of `modeling_tapas.py`(loss calculation + backward pass) is the most important. ", "> @NielsRogge if you want I can take care of the remaining steps for the tokenizer:\r\n> \r\n> > * Add support for the drop_rows_to_fit and cell_trim_length attributes of TapasTokenizer, which should reflect the original API (see also my suggestion on how this could be done for cell_trim_length). Also, setting truncation=True in TapasTokenizer doesn't do anything currently.\r\n> > * I've added support for the special [EMPTY] token for empty cells in a table (based on the format_text method, see here). Does this have implications for the add_special_tokens method? I assume not? What about get_special_tokens_mask? To be verified.\r\n\r\n@LysandreJik ok that would be great!\r\n\r\n\r\n", "Thanks for the reviews, I've updated the requested changes and marked the ones I did as resolved. \r\n\r\n@LysandreJik, could you maybe fix the remaining comments? In short:\r\n\r\n* remove the encoder-decoder logic of `TapasModel` (only remove this from the API of `TapasModel`, but leave them in the code that's copied from BERT (that the user won't see and won't use). I'll let you do this since I don't want to mess up anything.\r\n* remove some tests and add a slow test as requested above\r\n\r\n... then I'll mark these as resolved. Besides these, there's also the truncation of `TapasTokenizer` which should still be implemented. I copied what was left here:\r\n* Add support for the `drop_rows_to_fit` and `cell_trim_length` attributes of `TapasTokenizer`, which should reflect the original API (see also [my suggestion](https://github.com/huggingface/transformers/pull/8482#discussion_r522131827) on how this could be done for `cell_trim_length`). The original implementation can be found [here](https://github.com/google-research/tapas/blob/4908213eb4df7aa988573350278b44c4dbe3f71b/tapas/utils/tf_example_utils.py#L999).\r\n* Add support for the special `[EMPTY]` token for empty cells in a table (see the `_tokenize` method of `TapasTokenizer`, which now uses the `format_text` method as in the [original implementation](https://github.com/google-research/tapas/blob/4908213eb4df7aa988573350278b44c4dbe3f71b/tapas/utils/tf_example_utils.py#L330)). Does this have implications for the `add_special_tokens` method? I assume not? What about `get_special_tokens_mask`? To be verified.\r\n* There was also a small discrepancy between the tokenization of TAPAS and the original implementation, see this [Github issue](https://github.com/google-research/tapas/issues/90#issuecomment-735723963). I don't expect this too big of an issue, but maybe you know more about this.\r\n\r\nAnd then I assume we're done 👍 (finally)", "Sure, will do so. Probably tomorrow morning/afternoon!", "Closing this one as the most up-to-date is now #9117 ." ]

[]

[]

[ "Hello, this has been fixed on: https://github.com/huggingface/transformers/pull/8082 and is now available in the `master` documentation and will be updated in the next version. Thanks for opening an issue!" ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1843244711, "node_id": "MDU6TGFiZWwxODQzMjQ0NzEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model", "name": "New model", "color": "fbca04", "default": false, "description": "" } ]

[]

[ "Will be a part of #6285", "Hey, @sumanthd17 any update on this? ", "@julien-c thanks for your amazing nlp lib. \r\nWhen do you plan to support mT5 ?\r\nWhen #6285 will be release ?\r\nCheers\r\nPhilippe ", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "Please check if it worked :)", "Hmm, I guess it'd be have been more practical to experiment with the push job and not scheduled - but ok, let's wait - if it's not working, I will attack it on all fronts.", "One more thing: \r\n\r\n[run_all_tests_torch_and_tf_gpu](https://github.com/stas00/transformers/blob/e248a114e7f13a970bd8b5e52c0a032c014f4a57/.github/workflows/self-scheduled.yml#L57) has 3 independent test suite runs and currently if one fails the others don't run! Which is not what is wanted I believe. I suggest that we add `if: always() ` to the last 2 test suites as they are independent from the first one. ", "I'm fine with that as long as the workflow run can still be correctly marked as a failure.\r\n\r\nTo rephrase the requirement at the risk of redundancy, we want a red x next to the job when any test fails:\r\n\r\n\r\nScreenshot shows that we are meeting this requirement at the moment.\r\n", "Yes, that's the idea. I think the intention of the `if` condition is to define only whether a job is to be run, and not impact the total outcome. But we will see that already with the results of this PR - as artifact upload job will surely succeed. If the total outcome is [x] and artifacts have run, then we can replicate that condition to the other test suites. ", "OK, It did the trick, see: https://github.com/huggingface/transformers/actions/runs/334754818 \r\nSpecifically, as requested: the final result is [x] and the artifact job did run regardless.\r\n\r\nSo we can apply this `if: always` condition to other `pytest` jobs on the same workflow. There is a nuance of a possibility of pre-pytest jobs failing and the `pytest` jobs running anyway with this condition, but if that situation arrives, it makes no difference - those jobs will just immediately fail. \r\n\r\nNotes:\r\n\r\n* It puts the artifact files in the same place from different jobs, so we need to call that artifact upload job differently for each job\r\n* The so-so part is that the artifacts on github actions are provided as a single zipped file, so you have to first download the file, unpack it and only then you can see the results. \r\n* Moreover it doesn't show the artifact file until **all** jobs have completed, despite saying that the file was successfully uploaded.\r\n\r\n**A Possible workaround:**\r\n\r\nOne possible optimization here could be to `cat reports/report_tests_failures.txt` right after `pytest`, in a separate mini-step, so that you can immediately see just the failures and not wait for everything else to finish and go through the multiple steps to get to this file. (It has to be a separate step (name+run) not to affect the success/failure exit status of the `pytest` step.\r\n\r\nPlease review the outcome/my notes and let me know whether we proceed with this to other jobs.\r\n\r\nSpecifically to moving forward, we probably need to wait for this to be merged: https://github.com/huggingface/transformers/pull/8007 as it has multiple changes to the CI files.\r\n\r\n", "i have read your report. It is very clear, thank you. let's try a careful cat solution where we keep the size of the results as small as reasonably possible. one or two screens of text that show which tests failed (and short tracebacks (pytest --tb=short) ). Thanks for the help this is going to be so much easier to use than the status quo. Let me know if further clarifications/decisions would be helpful, and feel free to push back if implementation is difficult. ", "wrt/ proposed workaround:\r\n\r\nSince the proposed quick `cat` is going to be in its own collapsible \"tab\" and will have only failures, let's start with just `cat reports/report_tests_failures.txt` and we can create other types of reports should it prove too verbose, and just cut those instead.\r\n\r\nBut also I could probably create `reports/report_tests_failures_short.txt` report which will emulate `pytest --tb=short`, so that we will have both long and short reports.\r\n\r\nwrt/ the rest:\r\n\r\nit still stands, correct? i.e. we still want the full artifacts in github actions", "> we still want the full artifacts in github actions\r\n\r\nYes, don't see any downside.", "It looks like the errors are generated with either `--tb=long` or `--tb=short` at run time, so when the reports time comes they are already saved as one or the other, but not both.\r\n\r\nSo if we want the short and the long reports, one possibility is to generate the long report and then to try to make it shorter with some regex or some simple truncation - resulting in a short report.\r\n\r\nAnother approach that might work is collecting the failures as they happen - I need to investigate whether I can control the format in that hook or not without impacting the global reporting, as I'm sure that ideally we do want the full long report too. Please correct me if I'm wrong and `--tb=short` is sufficient for CIs (i.e. there will be no full long failures report anywhere - neither terminal logs nor report files), then it's easy.\r\n", "I trust you to make those choices as you see fit. feel free to ignore tb=short.", "I nailed it, got the cake and ate it too." ]

[]

[]

[ "T5: @patrickvonplaten I think you need to set `_additional_special_tokens`.", "@jsrozner want to try to fix?", "This is a duplicate of #5142 and will be fixed with the PR linked below. Thanks for reporting it - seems like multiple people were running into this issue!!!" ]

[]

[]

[ "Hey @zhiheng-huang,\r\n\r\nThanks for the PR!\r\n\r\nIn general I'm fine with this PR - think adding more types of position embeddings to a model is OK. \r\nAlso can you rebase your PR to the most current version of `master` - I think you are working on a rather old version.\r\n\r\n", "> Hey @zhiheng-huang,\r\n> \r\n> Thanks for the PR!\r\n> \r\n> In general I'm fine with this PR - think adding more types of position embeddings to a model is OK.\r\n> Also can you rebase your PR to the most current version of `master` - I think you are working on a rather old version.\r\n\r\nIt was rebased to master on 10/27, will rebase again for the new revision.", "@patrickvonplaten, I did a rebase and I am not sure this is the correct way to review commit \"Address review comment\". Please let me know if there is a better way to upload the new diff.", "Hey @zhiheng-huang,\r\n\r\nI think there was a problem with the rebase it seems like you added all commits on master on top of your PR. \r\nThis happens from time to time sadly :-/ \r\n\r\nThe way I'd fix it is to first save your changes of the last commit: https://github.com/huggingface/transformers/pull/8108/commits/ffe2e64c64f03c141cc085c8f3f509ae2e0992e2 somewhere (maybe a new branch).\r\n\r\nThen correctly reset the head of your branch before all other commitns were falsely added:\r\n```\r\ngit reset --hard 36729ee\r\n```\r\n\r\nThen add the single commit you saved in another branch\r\n\r\n\r\n```\r\ngit cherry-pick ffe2e64\r\n```\r\n\r\nand finally either you correctly rebase OR the safer option here would probably be to merge the master into your branch\r\n\r\n```\r\ngit fetch upstream/master\r\ngit merge upstream master\r\n```\r\n\r\nHope this helps!", "> Hey @zhiheng-huang,\r\n> \r\n> I think there was a problem with the rebase it seems like you added all commits on master on top of your PR.\r\n> This happens from time to time sadly :-/\r\n> \r\n> The way I'd fix it is to first save your changes of the last commit: [ffe2e64](https://github.com/huggingface/transformers/commit/ffe2e64c64f03c141cc085c8f3f509ae2e0992e2) somewhere (maybe a new branch).\r\n> \r\n> Then correctly reset the head of your branch before all other commitns were falsely added:\r\n> \r\n> ```\r\n> git reset --hard 36729ee\r\n> ```\r\n> \r\n> Then add the single commit you saved in another branch\r\n> \r\n> ```\r\n> git cherry-pick ffe2e64\r\n> ```\r\n> \r\n> and finally either you correctly rebase OR the safer option here would probably be to merge the master into your branch\r\n> \r\n> ```\r\n> git fetch upstream/master\r\n> git merge upstream master\r\n> ```\r\n> \r\n> Hope this helps!\r\n\r\nThanks. @patrickvonplaten. this helps but I may have to revert the commits merged to zhiheng-huang:transformers-relative-embedding. I created a new PR at https://github.com/huggingface/transformers/pull/8276 to continue the review. Thanks!" ]

[]

[]

[ "LGTM! cc @patrickvonplaten for awareness!" ]

[]

[]

[]

[]

[]

[]

[]

[]

[ "retriever.index.get_doc_dicts(docs_dict[\"doc_ids\"])[0]['text'] gets me the text of the retrieved documents." ]

[]

[]

[ "As explained in the README of the examples, you need an [installation from source](https://huggingface.co/transformers/installation.html#installing-from-source) to run the examples, which you don't have, otherwise you would have this object.\r\n\r\nAlternatively, you can run the examples associated to your current version by using the files on the [last release tag](https://github.com/huggingface/transformers/releases/tag/v3.4.0)." ]

[]

[]

[]

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

[ { "login": "sshleifer", "id": 6045025, "node_id": "MDQ6VXNlcjYwNDUwMjU=", "avatar_url": "https://avatars.githubusercontent.com/u/6045025?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sshleifer", "html_url": "https://github.com/sshleifer", "followers_url": "https://api.github.com/users/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "It is worth mentioning that PyTorch's SiLU op is an optimized implementation: https://github.com/pytorch/pytorch/pull/42976", "You are correct, we should update and deprecate. Do you want to take a stab at a PR?" ]

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

[]

[ "This would be cool! I don't believe it's on our roadmap currently (cc @patrickvonplaten), but should be part of a general TF overhaul we'll be doing in the coming months.\r\n\r\nNo date for that yet, but I think you can expect it in the future.", "Could be a \"Good First Issue\" :D Yeah, I think this is not really on our roadmap because `Reformer` is a pretty complicated model (It tweaks the backprop pass) and does not really have pretrained weights :-/ ", "Sounds like a good first project to deep dive in TensorFlow and rip a few hair out in the process :) ", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "I'm experiencing the same issue. Setting `vocab_size=tokenizer.vocab_size` does not help.\r\n\r\nI've noticed if I artificially inflate the vocab_size (i.e. `vocab_size=tokenizer.vocab_size+1`), the negative dimension in the error will also be greater by 1 (ie `-199743` instead of `-199744`.", "Hey! From your description it sounds like you haven't changed the cutoff points for adaptive embeddings. (the different sizes of the clusters for the hierarchical softmax generation). This causes an issue as the last cluster of embeddings, the one for the least frequent words, has size `vocab_size - cutoffs[-1]` so if the last cutoff is bigger than the vocab size, that's negative.\r\n\r\nNow for only 256 vocab words, adaptive embeddings don't really matter anyway, so I'd recommend running\r\n\r\n```\r\nfrom transformers import TransfoXLConfig, TransfoXLModel\r\nconfiguration = TransfoXLConfig(vocab_size=256, cutoffs=[])\r\nmodel = TransfoXLModel(configuration)\r\n```", "This worked for me, thanks a lot @TevenLeScao ! If you had a larger vocab size would you just recommend setting the last cutoff to be `0 < cutoff < vocab_size`?\r\n", "Ah actually I re-read the code and docs and the two ends of the cutoffs are already provided; they're appended later, so what you want is actually `cutoffs=[]`, even if `cutoffs=[0, 256]` seems to work anyway (I've edited my previous answer).\r\n\r\nIn any case, to answer your question, yes, for a larger vocab size it is actually quite helpful to have `0 < cutoff < vocab_size`! Personally I start considering doing that for a vocabulary on the order of a few tens of thousands - so like 40000 for example, but your mileage may vary, I recommend experimenting yourself and checking whether it makes a difference :) it should mostly help with memory use.", "Awesome, thanks for the helpful advice! I'd posted about the same issue on https://discuss.huggingface.co/t/transfoxllmheadmodel-trying-to-create-tensor-with-negative-dimension-199500/1768/2 but it remained unanswered, so I linked your comment in that thread as a solution.", "@TevenLeScao Thanks very much, it works great for me, close the issue now." ]

[]

[]

[]

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

[]

[]

[]

[]

[]

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

[]

[ "You are correct! Do you want to open a PR fixing the docs?", "Sure, I'll do that. Closing this issue for now, I'll refer to this in the PR.", "There is one more thing, sorry I didn't realize this earlier, and please let me know if I am wrong, The [span_to_answer](https://huggingface.co/transformers/main_classes/pipelines.html#transformers.QuestionAnsweringPipeline.span_to_answer) method of this pipeline object doesn't add much value, because the object itself when called with `question` and `context` will return the start and end indexes of the answer in the context. Moreover, it will give the wrong results because `span_to_answer` is expecting token index and not string indexes.\r\nIn continuation to above code;\r\n```python\r\nprint(len(tokenizer.tokenize(text)))\r\n# output: 96\r\n\r\n print(qa_pipeline.span_to_answer(text=text,start=int(result[0][\"start\"]), end=int(result[0][\"end\"])))\r\n# this will print {'answer': '', 'start': 0, 'end': 0} because the start index of the answer is 256 and end index is 264 while the \r\n# tokenized length is 96, this line in the function will stop the loop\r\n# if token_idx > end:\r\n# break\r\n\r\n```\r\nThis part in `__call__` [method](https://huggingface.co/transformers/main_classes/pipelines.html#transformers.QuestionAnsweringPipeline.__call__) is already taking care of remapping to string indexes:\r\n```python\r\n # Convert the answer (tokens) back to the original text\r\n answers += [\r\n {\r\n \"score\": score.item(),\r\n \"start\": np.where(char_to_word == feature.token_to_orig_map[s])[0][0].item(),\r\n \"end\": np.where(char_to_word == feature.token_to_orig_map[e])[0][-1].item(),\r\n \"answer\": \" \".join(\r\n example.doc_tokens[feature.token_to_orig_map[s] : feature.token_to_orig_map[e] + 1]\r\n ),\r\n }\r\n```\r\nUnless we have some method to get the token index, this method I think will not work.\r\n", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "If we try to go back to using it, perhaps there is a way to merge reports from 2 half-full tests - otherwise it creates an unnecessary slowdown on CI to run both together if we don't need to." ]

[]

[]

[]

[]

[]

[]

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

[]

[]

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

[]

[]

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

[]

[]

[]

[]

[ "I'm sorry, but you could explain to me why an `IterableDataset` with a `__len__` is not a regular `Dataset`?", "In my case, I wrap a `Dataset` using a class that inherits `IterableDataset`, and defines a `__len__()`.\r\nThe purpose is to implement smart batching[1]. I use `IterableDataset` so I can control how to iterate the data.\r\n\r\nI don't know if it's possible/easier to `Dataset`+`Sampler`, if so please let me know.\r\nAlso note that (after the change) if I drop `__len__()` to suppress the bug, I would then need to specify `max_iter` (or something like that), which is inconvenient.\r\n[1] (https://wandb.ai/pommedeterresautee/speed_training/reports/Train-HuggingFace-Models-Twice-As-Fast--VmlldzoxMDgzOTI)\r\n\r\n", "It does seem a bit hacky but I guess we can add that test. Do you want to suggest a PR with the change?" ]

[]

[]

[ "We will probably need at least the full error trace", "```\r\nCalling BertTokenizer.from_pretrained() with the path to a single file or url is deprecated\r\nSpecial tokens have been added in the vocabulary, make sure the associated word embedding are fine-tuned or trained.\r\nTraceback (most recent call last):\r\n File \"/Users/gt/Desktop/transformers-3.4.0/my_code/test.py\", line 4, in <module>\r\n model = TFAutoModel.from_pretrained(\"/Users/gt/Desktop/transformers-3.4.0/my_code/bert-base-uncased-pytorch_model.bin\")\r\n File \"/Users/gt/Desktop/transformers-3.4.0/src/transformers/modeling_tf_auto.py\", line 493, in from_pretrained\r\n pretrained_model_name_or_path, return_unused_kwargs=True, **kwargs\r\n File \"/Users/gt/Desktop/transformers-3.4.0/src/transformers/configuration_auto.py\", line 330, in from_pretrained\r\n config_dict, _ = PretrainedConfig.get_config_dict(pretrained_model_name_or_path, **kwargs)\r\n File \"/Users/gt/Desktop/transformers-3.4.0/src/transformers/configuration_utils.py\", line 374, in get_config_dict\r\n config_dict = cls._dict_from_json_file(resolved_config_file)\r\n File \"/Users/gt/Desktop/transformers-3.4.0/src/transformers/configuration_utils.py\", line 456, in _dict_from_json_file\r\n text = reader.read()\r\n File \"/Users/gt/Py36-tf1.4/lib/python3.6/codecs.py\", line 321, in decode\r\n (result, consumed) = self._buffer_decode(data, self.errors, final)\r\nUnicodeDecodeError: 'utf-8' codec can't decode byte 0x80 in position 0: invalid start byte\r\n```", "This line: `model = TFAutoModel.from_pretrained(\"/Users/gt/Desktop/transformers-3.4.0/my_code/bert-base-uncased-pytorch_model.bin\")`\r\n\r\nshould point to a directory contining both the model file and the configuration. Also, you're loading a `pytorch_model.bin` in a `TFAutoModel`, whereas this is a TensorFlow automodel.\r\n\r\n- You should make sure that you're loading from a directory containing either `(pytorch_model.bin, config.json)` for PyTorch, or `(tf_model.h5, config.json)` for TensorFlow\r\n- You can load a PyTorch model in TensorFlow, but you should specify `from_pt=True`, and you can load a TensorFlow model in PyTorch but you should specify the `from_tf=True` option.\r\n\r\nYou can find more information about this in the [quick tour](https://huggingface.co/transformers/quicktour.html#under-the-hood-pretrained-models).", "@LysandreJik Thank you but\r\n\r\n```\r\nfrom transformers import AutoTokenizer, TFAutoModel\r\n\r\ntokenizer = AutoTokenizer.from_pretrained(\"/Users/gt/Desktop/transformers-3.4.0/my_code/\")\r\nmodel = TFAutoModel.from_pretrained(\"/Users/gt/Desktop/transformers-3.4.0/my_code/\")\r\n\r\ninputs = tokenizer(\"Hello world!\", return_tensors=\"tf\")\r\noutputs = model(**inputs)\r\nprint(outputs)\r\n```\r\n\r\n\r\n\r\n\r\n```\r\nTraceback (most recent call last):\r\n File \"/Users/gt/Desktop/transformers-3.4.0/my_code/test.py\", line 3, in <module>\r\n tokenizer = AutoTokenizer.from_pretrained(\"/Users/gt/Desktop/transformers-3.4.0/my_code/\")\r\n File \"/Users/gt/Desktop/transformers-3.4.0/src/transformers/tokenization_auto.py\", line 333, in from_pretrained\r\n return tokenizer_class_py.from_pretrained(pretrained_model_name_or_path, *inputs, **kwargs)\r\n File \"/Users/gt/Desktop/transformers-3.4.0/src/transformers/tokenization_utils_base.py\", line 1591, in from_pretrained\r\n list(cls.vocab_files_names.values()),\r\nOSError: Model name '/Users/gt/Desktop/transformers-3.4.0/my_code/' was not found in tokenizers model name list (bert-base-uncased, bert-large-uncased, bert-base-cased, bert-large-cased, bert-base-multilingual-uncased, bert-base-multilingual-cased, bert-base-chinese, bert-base-german-cased, bert-large-uncased-whole-word-masking, bert-large-cased-whole-word-masking, bert-large-uncased-whole-word-masking-finetuned-squad, bert-large-cased-whole-word-masking-finetuned-squad, bert-base-cased-finetuned-mrpc, bert-base-german-dbmdz-cased, bert-base-german-dbmdz-uncased, TurkuNLP/bert-base-finnish-cased-v1, TurkuNLP/bert-base-finnish-uncased-v1, wietsedv/bert-base-dutch-cased). We assumed '/Users/gt/Desktop/transformers-3.4.0/my_code/' was a path, a model identifier, or url to a directory containing vocabulary files named ['vocab.txt'] but couldn't find such vocabulary files at this path or url.\r\n```", "Hello, as said before you need your files to be correctly named. Your model should be `pytorch_model.bin` or `tf_model.h5`, your configuration `config.json`, and your tokenizer should also be pointing to a file that has an appropriate name. You seem to be loading a `bert-base-cased` model, which should be used with a `BertTokenizer` that uses `vocab.txt` files, as it is shown in the error.", "Thank you." ]

[]

[]

[ "Closing as I believe this is an error :)" ]

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

[]

[ "This issue has been automatically marked as stale and been closed because it has not had recent activity. Thank you for your contributions.\n\nIf you think this still needs to be addressed please comment on this thread." ]

[ { "id": 1314768611, "node_id": "MDU6TGFiZWwxMzE0NzY4NjEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/wontfix", "name": "wontfix", "color": "ffffff", "default": true, "description": null }, { "id": 1843244711, "node_id": "MDU6TGFiZWwxODQzMjQ0NzEx", "url": "https://api.github.com/repos/huggingface/transformers/labels/New%20model", "name": "New model", "color": "fbca04", "default": false, "description": "" } ]

[]

[ "Hi! This is great, thanks for offering to contribute it! From what I understand, `FastFormers` contains several scripts that can be applied to `transformers` models out of the box, that is, training, distillation, pruning, using quantization alongside the onnx runtime and fp16 optimizations. \r\n\r\nIs that correct? If that is so, the easiest way would be to add the corresponding scripts to the `examples/` directory, probably under `examples/fastformers`. If there are modifications made to the model themselves, we can take a look together at how we can integrate those in the library.", "Hi, thanks for your interest! From what I understand, I think your model falls in the category of dynamic acceleration. For these types of paper, I recommend you to integrate it to `examples/`, just like [PABEE](https://github.com/huggingface/transformers/tree/master/examples/bert-loses-patience) and [DeeBERT](https://github.com/huggingface/transformers/tree/master/examples/deebert). I've emailed you an invitation to our Slack channel if it works for you. cc @LysandreJik ", " @LysandreJik yes, that is correct. Thanks @JetRunner, let's discuss more on the slack.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n", "I am sorry, but I have been fully loaded with some other stuffs. I won't be able to make a progress. I'd like to close this to avoid any confusion." ]

[]

[]

[]

[]

[]

[ "Ideally there should be a modified_py_and_rst_files to speed up this check (and only apply it to modified files), but this works in the meantime. @stas00 if you want to do that last bit of optimization, let me know, otherwise I'll do that later.", "Go for it, @sgugger. The modified files var is there, so it should be easy to apply it anywhere.\r\n\r\nIf you get stuck I'm here to help." ]

[]

[]

[]

[]

[]

[ "download all the source code" ]

[]

[]

[ "Have you checked out the filtered list on the model hub? https://huggingface.co/models?filter=zh ", "Thank you!" ]

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

[]

[ "Hi, have you tried running this code without using CUDA (i.e., on CPU)? The errors are usually more intelligible that way.\r\n\r\nCould this be due to an OOM error and cuda not recovering from it?", "> Hi, have you tried running this code without using CUDA (i.e., on CPU)? The errors are usually more intelligible that way.\r\n> \r\n> Could this be due to an OOM error and cuda not recovering from it?\r\n\r\nHi @LysandreJik , thanks for your response! I made some exploration following your suggestion: I move the job to CPU only and tried again. I think this time it has something to do with the position embeddings indexing. Error message:\r\n\r\n```\r\nTraceback (most recent call last):\r\n File \"finetune-marco.py\", line 94, in <module>\r\n marco.run()\r\n File \"/mnt/nfs/work1/user/user/LF-for-IR/Marco.py\", line 177, in run\r\n self.train()\r\n File \"/mnt/nfs/work1/user/user/LF-for-IR/Marco.py\", line 258, in train\r\n outputs = self.model(\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/torch/nn/modules/module.py\", line 550, in __call__\r\n result = self.forward(*input, **kwargs)\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/transformers/modeling_longformer.py\", line 1445, in forward\r\n outputs = self.longformer(\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/torch/nn/modules/module.py\", line 550, in __call__\r\n result = self.forward(*input, **kwargs)\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/transformers/modeling_longformer.py\", line 1261, in forward\r\n embedding_output = self.embeddings(\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/torch/nn/modules/module.py\", line 550, in __call__\r\n result = self.forward(*input, **kwargs)\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/transformers/modeling_longformer.py\", line 170, in forward\r\n position_embeddings = self.position_embeddings(position_ids)\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/torch/nn/modules/module.py\", line 550, in __call__\r\n result = self.forward(*input, **kwargs)\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/torch/nn/modules/sparse.py\", line 112, in forward\r\n return F.embedding(\r\n File \"/home/user/miniconda3/envs/marco/lib/python3.8/site-packages/torch/nn/functional.py\", line 1724, in embedding\r\n return torch.embedding(weight, input, padding_idx, scale_grad_by_freq, sparse)\r\nIndexError: index out of range in self\r\n\r\n```\r\n\r\n So if I go to ```class LongformerEmbeddings(nn.Module)``` in ```modeling_longformer.py```and print the ```self.position_ids``` and ```position_ids``` inside the ```forward()``` function, I get:\r\n\r\n```\r\ntensor([[ 0, 1, 2, 3, 4, 5, 6, 7, 8, 9, 10, 11,\r\n 12, 13, 14, 15, 16, 17, 18, 19, 20, 21, 22, 23,\r\n 24, 25, 26, 27, 28, 29, 30, 31, 32, 33, 34, 35,\r\n 36, 37, 38, 39, 40, 41, 42, 43, 44, 45, 46, 47,\r\n 48, 49, 50, 51, 52, 53, 54, 55, 56, 57, 58, 59,\r\n 60, 61, 62, 63, 64, 65, 66, 67, 68, 69, 70, 71,\r\n 72, 73, 74, 75, 76, 77, 78, 79, 80, 81, 82, 83,\r\n 84, 85, 86, 87, 88, 89, 90, 91, 92, 93, 94, 95,\r\n 96, 97, 98, 99, 100, 101, 102, 103, 104, 105, 106, 107,\r\n 108, 109, 110, 111, 112, 113, 114, 115, 116, 117, 118, 119,\r\n 120, 121, 122, 123, 124, 125, 126, 127, 128, 129, 130, 131,\r\n 132, 133, 134, 135, 136, 137, 138, 139, 140, 141, 142, 143,\r\n 144, 145, 146, 147, 148, 149, 150, 151, 152, 153, 154, 155,\r\n 156, 157, 158, 159, 160, 161, 162, 163, 164, 165, 166, 167,\r\n 168, 169, 170, 171, 172, 173, 174, 175, 176, 177, 178, 179,\r\n 180, 181, 182, 183, 184, 185, 186, 187, 188, 189, 190, 191,\r\n 192, 193, 194, 195, 196, 197, 198, 199, 200, 201, 202, 203,\r\n 204, 205, 206, 207, 208, 209, 210, 211, 212, 213, 214, 215,\r\n 216, 217, 218, 219, 220, 221, 222, 223, 224, 225, 226, 227,\r\n 228, 229, 230, 231, 232, 233, 234, 235, 236, 237, 238, 239,\r\n 240, 241, 242, 243, 244, 245, 246, 247, 248, 249, 250, 251,\r\n 252, 253, 254, 255, 256, 257, 258, 259, 260, 261, 262, 263,\r\n 264, 265, 266, 267, 268, 269, 270, 271, 272, 273, 274, 275,\r\n 276, 277, 278, 279, 280, 281, 282, 283, 284, 285, 286, 287,\r\n 288, 289, 290, 291, 292, 293, 294, 295, 296, 297, 298, 299,\r\n 300, 301, 302, 303, 304, 305, 306, 307, 308, 309, 310, 311,\r\n 312, 313, 314, 315, 316, 317, 318, 319, 320, 321, 322, 323,\r\n 324, 325, 326, 327, 328, 329, 330, 331, 332, 333, 334, 335,\r\n 336, 337, 338, 339, 340, 341, 342, 343, 344, 345, 346, 347,\r\n 348, 349, 350, 351, 352, 353, 354, 355, 356, 357, 358, 359,\r\n 360, 361, 362, 363, 364, 365, 366, 367, 368, 369, 370, 371,\r\n 372, 373, 374, 375, 376, 377, 378, 379, 380, 381, 382, 383,\r\n 384, 385, 386, 387, 388, 389, 390, 391, 392, 393, 394, 395,\r\n 396, 397, 398, 399, 400, 401, 402, 403, 404, 405, 406, 407,\r\n 408, 409, 410, 411, 412, 413, 414, 415, 416, 417, 418, 419,\r\n 420, 421, 422, 423, 424, 425, 426, 427, 428, 429, 430, 431,\r\n 432, 433, 434, 435, 436, 437, 438, 439, 440, 441, 442, 443,\r\n 444, 445, 446, 447, 448, 449, 450, 451, 452, 453, 454, 455,\r\n 456, 457, 458, 459, 460, 461, 462, 463, 464, 465, 466, 467,\r\n 468, 469, 470, 471, 472, 473, 474, 475, 476, 477, 478, 479,\r\n 480, 481, 482, 483, 484, 485, 486, 487, 488, 489, 490, 491,\r\n 492, 493, 494, 495, 496, 497, 498, 499, 500, 501, 502, 503,\r\n 504, 505, 506, 507, 508, 509, 510, 511, 512, 513, 514, 515,\r\n 516, 517, 518, 519, 520, 521, 522, 523, 524, 525, 526, 527,\r\n 528, 529, 530, 531, 532, 533, 534, 535, 536, 537, 538, 539,\r\n 540, 541, 542, 543, 544, 545, 546, 547, 548, 549, 550, 551,\r\n 552, 553, 554, 555, 556, 557, 558, 559, 560, 561, 562, 563,\r\n 564, 565, 566, 567, 568, 569, 570, 571, 572, 573, 574, 575,\r\n 576, 577, 578, 579, 580, 581, 582, 583, 584, 585, 586, 587,\r\n 588, 589, 590, 591, 592, 593, 594, 595, 596, 597, 598, 599,\r\n 600, 601, 602, 603, 604, 605, 606, 607, 608, 609, 610, 611,\r\n 612, 613, 614, 615, 616, 617, 618, 619, 620, 621, 622, 623,\r\n 624, 625, 626, 627, 628, 629, 630, 631, 632, 633, 634, 635,\r\n 636, 637, 638, 639, 640, 641, 642, 643, 644, 645, 646, 647,\r\n 648, 649, 650, 651, 652, 653, 654, 655, 656, 657, 658, 659,\r\n 660, 661, 662, 663, 664, 665, 666, 667, 668, 669, 670, 671,\r\n 672, 673, 674, 675, 676, 677, 678, 679, 680, 681, 682, 683,\r\n 684, 685, 686, 687, 688, 689, 690, 691, 692, 693, 694, 695,\r\n 696, 697, 698, 699, 700, 701, 702, 703, 704, 705, 706, 707,\r\n 708, 709, 710, 711, 712, 713, 714, 715, 716, 717, 718, 719,\r\n 720, 721, 722, 723, 724, 725, 726, 727, 728, 729, 730, 731,\r\n 732, 733, 734, 735, 736, 737, 738, 739, 740, 741, 742, 743,\r\n 744, 745, 746, 747, 748, 749, 750, 751, 752, 753, 754, 755,\r\n 756, 757, 758, 759, 760, 761, 762, 763, 764, 765, 766, 767,\r\n 768, 769, 770, 771, 772, 773, 774, 775, 776, 777, 778, 779,\r\n 780, 781, 782, 783, 784, 785, 786, 787, 788, 789, 790, 791,\r\n 792, 793, 794, 795, 796, 797, 798, 799, 800, 801, 802, 803,\r\n 804, 805, 806, 807, 808, 809, 810, 811, 812, 813, 814, 815,\r\n 816, 817, 818, 819, 820, 821, 822, 823, 824, 825, 826, 827,\r\n 828, 829, 830, 831, 832, 833, 834, 835, 836, 837, 838, 839,\r\n 840, 841, 842, 843, 844, 845, 846, 847, 848, 849, 850, 851,\r\n 852, 853, 854, 855, 856, 857, 858, 859, 860, 861, 862, 863,\r\n 864, 865, 866, 867, 868, 869, 870, 871, 872, 873, 874, 875,\r\n 876, 877, 878, 879, 880, 881, 882, 883, 884, 885, 886, 887,\r\n 888, 889, 890, 891, 892, 893, 894, 895, 896, 897, 898, 899,\r\n 900, 901, 902, 903, 904, 905, 906, 907, 908, 909, 910, 911,\r\n 912, 913, 914, 915, 916, 917, 918, 919, 920, 921, 922, 923,\r\n 924, 925, 926, 927, 928, 929, 930, 931, 932, 933, 934, 935,\r\n 936, 937, 938, 939, 940, 941, 942, 943, 944, 945, 946, 947,\r\n 948, 949, 950, 951, 952, 953, 954, 955, 956, 957, 958, 959,\r\n 960, 961, 962, 963, 964, 965, 966, 967, 968, 969, 970, 971,\r\n 972, 973, 974, 975, 976, 977, 978, 979, 980, 981, 982, 983,\r\n 984, 985, 986, 987, 988, 989, 990, 991, 992, 993, 994, 995,\r\n 996, 997, 998, 999, 1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007,\r\n 1008, 1009, 1010, 1011, 1012, 1013, 1014, 1015, 1016, 1017, 1018, 1019,\r\n 1020, 1021, 1022, 1023, 1024, 1025, 1026, 1027, 1028, 1029, 1030, 1031,\r\n 1032, 1033, 1034, 1035, 1036, 1037, 1038, 1039, 1040, 1041, 1042, 1043,\r\n 1044, 1045, 1046, 1047, 1048, 1049, 1050, 1051, 1052, 1053, 1054, 1055,\r\n 1056, 1057, 1058, 1059, 1060, 1061, 1062, 1063, 1064, 1065, 1066, 1067,\r\n 1068, 1069, 1070, 1071, 1072, 1073, 1074, 1075, 1076, 1077, 1078, 1079,\r\n 1080, 1081, 1082, 1083, 1084, 1085, 1086, 1087, 1088, 1089, 1090, 1091,\r\n 1092, 1093, 1094, 1095, 1096, 1097, 1098, 1099, 1100, 1101, 1102, 1103,\r\n 1104, 1105, 1106, 1107, 1108, 1109, 1110, 1111, 1112, 1113, 1114, 1115,\r\n 1116, 1117, 1118, 1119, 1120, 1121, 1122, 1123, 1124, 1125, 1126, 1127,\r\n 1128, 1129, 1130, 1131, 1132, 1133, 1134, 1135, 1136, 1137, 1138, 1139,\r\n 1140, 1141, 1142, 1143, 1144, 1145, 1146, 1147, 1148, 1149, 1150, 1151,\r\n 1152, 1153, 1154, 1155, 1156, 1157, 1158, 1159, 1160, 1161, 1162, 1163,\r\n 1164, 1165, 1166, 1167, 1168, 1169, 1170, 1171, 1172, 1173, 1174, 1175,\r\n 1176, 1177, 1178, 1179, 1180, 1181, 1182, 1183, 1184, 1185, 1186, 1187,\r\n 1188, 1189, 1190, 1191, 1192, 1193, 1194, 1195, 1196, 1197, 1198, 1199,\r\n 1200, 1201, 1202, 1203, 1204, 1205, 1206, 1207, 1208, 1209, 1210, 1211,\r\n 1212, 1213, 1214, 1215, 1216, 1217, 1218, 1219, 1220, 1221, 1222, 1223,\r\n 1224, 1225, 1226, 1227, 1228, 1229, 1230, 1231, 1232, 1233, 1234, 1235,\r\n 1236, 1237, 1238, 1239, 1240, 1241, 1242, 1243, 1244, 1245, 1246, 1247,\r\n 1248, 1249, 1250, 1251, 1252, 1253, 1254, 1255, 1256, 1257, 1258, 1259,\r\n 1260, 1261, 1262, 1263, 1264, 1265, 1266, 1267, 1268, 1269, 1270, 1271,\r\n 1272, 1273, 1274, 1275, 1276, 1277, 1278, 1279, 1280, 1281, 1282, 1283,\r\n 1284, 1285, 1286, 1287, 1288, 1289, 1290, 1291, 1292, 1293, 1294, 1295,\r\n 1296, 1297, 1298, 1299, 1300, 1301, 1302, 1303, 1304, 1305, 1306, 1307,\r\n 1308, 1309, 1310, 1311, 1312, 1313, 1314, 1315, 1316, 1317, 1318, 1319,\r\n 1320, 1321, 1322, 1323, 1324, 1325, 1326, 1327, 1328, 1329, 1330, 1331,\r\n 1332, 1333, 1334, 1335, 1336, 1337, 1338, 1339, 1340, 1341, 1342, 1343,\r\n 1344, 1345, 1346, 1347, 1348, 1349, 1350, 1351, 1352, 1353, 1354, 1355,\r\n 1356, 1357, 1358, 1359, 1360, 1361, 1362, 1363, 1364, 1365, 1366, 1367,\r\n 1368, 1369, 1370, 1371, 1372, 1373, 1374, 1375, 1376, 1377, 1378, 1379,\r\n 1380, 1381, 1382, 1383, 1384, 1385, 1386, 1387, 1388, 1389, 1390, 1391,\r\n 1392, 1393, 1394, 1395, 1396, 1397, 1398, 1399, 1400, 1401, 1402, 1403,\r\n 1404, 1405, 1406, 1407, 1408, 1409, 1410, 1411, 1412, 1413, 1414, 1415,\r\n 1416, 1417, 1418, 1419, 1420, 1421, 1422, 1423, 1424, 1425, 1426, 1427,\r\n 1428, 1429, 1430, 1431, 1432, 1433, 1434, 1435, 1436, 1437, 1438, 1439,\r\n 1440, 1441, 1442, 1443, 1444, 1445, 1446, 1447, 1448, 1449, 1450, 1451,\r\n 1452, 1453, 1454, 1455, 1456, 1457, 1458, 1459, 1460, 1461, 1462, 1463,\r\n 1464, 1465, 1466, 1467, 1468, 1469, 1470, 1471, 1472, 1473, 1474, 1475,\r\n 1476, 1477, 1478, 1479, 1480, 1481, 1482, 1483, 1484, 1485, 1486, 1487,\r\n 1488, 1489, 1490, 1491, 1492, 1493, 1494, 1495, 1496, 1497, 1498, 1499,\r\n 1500, 1501, 1502, 1503, 1504, 1505, 1506, 1507, 1508, 1509, 1510, 1511,\r\n 1512, 1513, 1514, 1515, 1516, 1517, 1518, 1519, 1520, 1521, 1522, 1523,\r\n 1524, 1525, 1526, 1527, 1528, 1529, 1530, 1531, 1532, 1533, 1534, 1535,\r\n 1536, 1537, 1538, 1539, 1540, 1541, 1542, 1543, 1544, 1545, 1546, 1547,\r\n 1548, 1549, 1550, 1551, 1552, 1553, 1554, 1555, 1556, 1557, 1558, 1559,\r\n 1560, 1561, 1562, 1563, 1564, 1565, 1566, 1567, 1568, 1569, 1570, 1571,\r\n 1572, 1573, 1574, 1575, 1576, 1577, 1578, 1579, 1580, 1581, 1582, 1583,\r\n 1584, 1585, 1586, 1587, 1588, 1589, 1590, 1591, 1592, 1593, 1594, 1595,\r\n 1596, 1597, 1598, 1599, 1600, 1601, 1602, 1603, 1604, 1605, 1606, 1607,\r\n 1608, 1609, 1610, 1611, 1612, 1613, 1614, 1615, 1616, 1617, 1618, 1619,\r\n 1620, 1621, 1622, 1623, 1624, 1625, 1626, 1627, 1628, 1629, 1630, 1631,\r\n 1632, 1633, 1634, 1635, 1636, 1637, 1638, 1639, 1640, 1641, 1642, 1643,\r\n 1644, 1645, 1646, 1647, 1648, 1649, 1650, 1651, 1652, 1653, 1654, 1655,\r\n 1656, 1657, 1658, 1659, 1660, 1661, 1662, 1663, 1664, 1665, 1666, 1667,\r\n 1668, 1669, 1670, 1671, 1672, 1673, 1674, 1675, 1676, 1677, 1678, 1679,\r\n 1680, 1681, 1682, 1683, 1684, 1685, 1686, 1687, 1688, 1689, 1690, 1691,\r\n 1692, 1693, 1694, 1695, 1696, 1697, 1698, 1699, 1700, 1701, 1702, 1703,\r\n 1704, 1705, 1706, 1707, 1708, 1709, 1710, 1711, 1712, 1713, 1714, 1715,\r\n 1716, 1717, 1718, 1719, 1720, 1721, 1722, 1723, 1724, 1725, 1726, 1727,\r\n 1728, 1729, 1730, 1731, 1732, 1733, 1734, 1735, 1736, 1737, 1738, 1739,\r\n 1740, 1741, 1742, 1743, 1744, 1745, 1746, 1747, 1748, 1749, 1750, 1751,\r\n 1752, 1753, 1754, 1755, 1756, 1757, 1758, 1759, 1760, 1761, 1762, 1763,\r\n 1764, 1765, 1766, 1767, 1768, 1769, 1770, 1771, 1772, 1773, 1774, 1775,\r\n 1776, 1777, 1778, 1779, 1780, 1781, 1782, 1783, 1784, 1785, 1786, 1787,\r\n 1788, 1789, 1790, 1791, 1792, 1793, 1794, 1795, 1796, 1797, 1798, 1799,\r\n 1800, 1801, 1802, 1803, 1804, 1805, 1806, 1807, 1808, 1809, 1810, 1811,\r\n 1812, 1813, 1814, 1815, 1816, 1817, 1818, 1819, 1820, 1821, 1822, 1823,\r\n 1824, 1825, 1826, 1827, 1828, 1829, 1830, 1831, 1832, 1833, 1834, 1835,\r\n 1836, 1837, 1838, 1839, 1840, 1841, 1842, 1843, 1844, 1845, 1846, 1847,\r\n 1848, 1849, 1850, 1851, 1852, 1853, 1854, 1855, 1856, 1857, 1858, 1859,\r\n 1860, 1861, 1862, 1863, 1864, 1865, 1866, 1867, 1868, 1869, 1870, 1871,\r\n 1872, 1873, 1874, 1875, 1876, 1877, 1878, 1879, 1880, 1881, 1882, 1883,\r\n 1884, 1885, 1886, 1887, 1888, 1889, 1890, 1891, 1892, 1893, 1894, 1895,\r\n 1896, 1897, 1898, 1899, 1900, 1901, 1902, 1903, 1904, 1905, 1906, 1907,\r\n 1908, 1909, 1910, 1911, 1912, 1913, 1914, 1915, 1916, 1917, 1918, 1919,\r\n 1920, 1921, 1922, 1923, 1924, 1925, 1926, 1927, 1928, 1929, 1930, 1931,\r\n 1932, 1933, 1934, 1935, 1936, 1937, 1938, 1939, 1940, 1941, 1942, 1943,\r\n 1944, 1945, 1946, 1947, 1948, 1949, 1950, 1951, 1952, 1953, 1954, 1955,\r\n 1956, 1957, 1958, 1959, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967,\r\n 1968, 1969, 1970, 1971, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979,\r\n 1980, 1981, 1982, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991,\r\n 1992, 1993, 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003,\r\n 2004, 2005, 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015,\r\n 2016, 2017, 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026, 2027,\r\n 2028, 2029, 2030, 2031, 2032, 2033, 2034, 2035, 2036, 2037, 2038, 2039,\r\n 2040, 2041, 2042, 2043, 2044, 2045, 2046, 2047, 2048, 2049, 2050, 2051,\r\n 2052, 2053, 2054, 2055, 2056, 2057, 2058, 2059, 2060, 2061, 2062, 2063,\r\n 2064, 2065, 2066, 2067, 2068, 2069, 2070, 2071, 2072, 2073, 2074, 2075,\r\n 2076, 2077, 2078, 2079, 2080, 2081, 2082, 2083, 2084, 2085, 2086, 2087,\r\n 2088, 2089, 2090, 2091, 2092, 2093, 2094, 2095, 2096, 2097, 2098, 2099,\r\n 2100, 2101, 2102, 2103, 2104, 2105, 2106, 2107, 2108, 2109, 2110, 2111,\r\n 2112, 2113, 2114, 2115, 2116, 2117, 2118, 2119, 2120, 2121, 2122, 2123,\r\n 2124, 2125, 2126, 2127, 2128, 2129, 2130, 2131, 2132, 2133, 2134, 2135,\r\n 2136, 2137, 2138, 2139, 2140, 2141, 2142, 2143, 2144, 2145, 2146, 2147,\r\n 2148, 2149, 2150, 2151, 2152, 2153, 2154, 2155, 2156, 2157, 2158, 2159,\r\n 2160, 2161, 2162, 2163, 2164, 2165, 2166, 2167, 2168, 2169, 2170, 2171,\r\n 2172, 2173, 2174, 2175, 2176, 2177, 2178, 2179, 2180, 2181, 2182, 2183,\r\n 2184, 2185, 2186, 2187, 2188, 2189, 2190, 2191, 2192, 2193, 2194, 2195,\r\n 2196, 2197, 2198, 2199, 2200, 2201, 2202, 2203, 2204, 2205, 2206, 2207,\r\n 2208, 2209, 2210, 2211, 2212, 2213, 2214, 2215, 2216, 2217, 2218, 2219,\r\n 2220, 2221, 2222, 2223, 2224, 2225, 2226, 2227, 2228, 2229, 2230, 2231,\r\n 2232, 2233, 2234, 2235, 2236, 2237, 2238, 2239, 2240, 2241, 2242, 2243,\r\n 2244, 2245, 2246, 2247, 2248, 2249, 2250, 2251, 2252, 2253, 2254, 2255,\r\n 2256, 2257, 2258, 2259, 2260, 2261, 2262, 2263, 2264, 2265, 2266, 2267,\r\n 2268, 2269, 2270, 2271, 2272, 2273, 2274, 2275, 2276, 2277, 2278, 2279,\r\n 2280, 2281, 2282, 2283, 2284, 2285, 2286, 2287, 2288, 2289, 2290, 2291,\r\n 2292, 2293, 2294, 2295, 2296, 2297, 2298, 2299, 2300, 2301, 2302, 2303,\r\n 2304, 2305, 2306, 2307, 2308, 2309, 2310, 2311, 2312, 2313, 2314, 2315,\r\n 2316, 2317, 2318, 2319, 2320, 2321, 2322, 2323, 2324, 2325, 2326, 2327,\r\n 2328, 2329, 2330, 2331, 2332, 2333, 2334, 2335, 2336, 2337, 2338, 2339,\r\n 2340, 2341, 2342, 2343, 2344, 2345, 2346, 2347, 2348, 2349, 2350, 2351,\r\n 2352, 2353, 2354, 2355, 2356, 2357, 2358, 2359, 2360, 2361, 2362, 2363,\r\n 2364, 2365, 2366, 2367, 2368, 2369, 2370, 2371, 2372, 2373, 2374, 2375,\r\n 2376, 2377, 2378, 2379, 2380, 2381, 2382, 2383, 2384, 2385, 2386, 2387,\r\n 2388, 2389, 2390, 2391, 2392, 2393, 2394, 2395, 2396, 2397, 2398, 2399,\r\n 2400, 2401, 2402, 2403, 2404, 2405, 2406, 2407, 2408, 2409, 2410, 2411,\r\n 2412, 2413, 2414, 2415, 2416, 2417, 2418, 2419, 2420, 2421, 2422, 2423,\r\n 2424, 2425, 2426, 2427, 2428, 2429, 2430, 2431, 2432, 2433, 2434, 2435,\r\n 2436, 2437, 2438, 2439, 2440, 2441, 2442, 2443, 2444, 2445, 2446, 2447,\r\n 2448, 2449, 2450, 2451, 2452, 2453, 2454, 2455, 2456, 2457, 2458, 2459,\r\n 2460, 2461, 2462, 2463, 2464, 2465, 2466, 2467, 2468, 2469, 2470, 2471,\r\n 2472, 2473, 2474, 2475, 2476, 2477, 2478, 2479, 2480, 2481, 2482, 2483,\r\n 2484, 2485, 2486, 2487, 2488, 2489, 2490, 2491, 2492, 2493, 2494, 2495,\r\n 2496, 2497, 2498, 2499, 2500, 2501, 2502, 2503, 2504, 2505, 2506, 2507,\r\n 2508, 2509, 2510, 2511, 2512, 2513, 2514, 2515, 2516, 2517, 2518, 2519,\r\n 2520, 2521, 2522, 2523, 2524, 2525, 2526, 2527, 2528, 2529, 2530, 2531,\r\n 2532, 2533, 2534, 2535, 2536, 2537, 2538, 2539, 2540, 2541, 2542, 2543,\r\n 2544, 2545, 2546, 2547, 2548, 2549, 2550, 2551, 2552, 2553, 2554, 2555,\r\n 2556, 2557, 2558, 2559, 2560, 2561, 2562, 2563, 2564, 2565, 2566, 2567,\r\n 2568, 2569, 2570, 2571, 2572, 2573, 2574, 2575, 2576, 2577, 2578, 2579,\r\n 2580, 2581, 2582, 2583, 2584, 2585, 2586, 2587, 2588, 2589, 2590, 2591,\r\n 2592, 2593, 2594, 2595, 2596, 2597, 2598, 2599, 2600, 2601, 2602, 2603,\r\n 2604, 2605, 2606, 2607, 2608, 2609, 2610, 2611, 2612, 2613, 2614, 2615,\r\n 2616, 2617, 2618, 2619, 2620, 2621, 2622, 2623, 2624, 2625, 2626, 2627,\r\n 2628, 2629, 2630, 2631, 2632, 2633, 2634, 2635, 2636, 2637, 2638, 2639,\r\n 2640, 2641, 2642, 2643, 2644, 2645, 2646, 2647, 2648, 2649, 2650, 2651,\r\n 2652, 2653, 2654, 2655, 2656, 2657, 2658, 2659, 2660, 2661, 2662, 2663,\r\n 2664, 2665, 2666, 2667, 2668, 2669, 2670, 2671, 2672, 2673, 2674, 2675,\r\n 2676, 2677, 2678, 2679, 2680, 2681, 2682, 2683, 2684, 2685, 2686, 2687,\r\n 2688, 2689, 2690, 2691, 2692, 2693, 2694, 2695, 2696, 2697, 2698, 2699,\r\n 2700, 2701, 2702, 2703, 2704, 2705, 2706, 2707, 2708, 2709, 2710, 2711,\r\n 2712, 2713, 2714, 2715, 2716, 2717, 2718, 2719, 2720, 2721, 2722, 2723,\r\n 2724, 2725, 2726, 2727, 2728, 2729, 2730, 2731, 2732, 2733, 2734, 2735,\r\n 2736, 2737, 2738, 2739, 2740, 2741, 2742, 2743, 2744, 2745, 2746, 2747,\r\n 2748, 2749, 2750, 2751, 2752, 2753, 2754, 2755, 2756, 2757, 2758, 2759,\r\n 2760, 2761, 2762, 2763, 2764, 2765, 2766, 2767, 2768, 2769, 2770, 2771,\r\n 2772, 2773, 2774, 2775, 2776, 2777, 2778, 2779, 2780, 2781, 2782, 2783,\r\n 2784, 2785, 2786, 2787, 2788, 2789, 2790, 2791, 2792, 2793, 2794, 2795,\r\n 2796, 2797, 2798, 2799, 2800, 2801, 2802, 2803, 2804, 2805, 2806, 2807,\r\n 2808, 2809, 2810, 2811, 2812, 2813, 2814, 2815, 2816, 2817, 2818, 2819,\r\n 2820, 2821, 2822, 2823, 2824, 2825, 2826, 2827, 2828, 2829, 2830, 2831,\r\n 2832, 2833, 2834, 2835, 2836, 2837, 2838, 2839, 2840, 2841, 2842, 2843,\r\n 2844, 2845, 2846, 2847, 2848, 2849, 2850, 2851, 2852, 2853, 2854, 2855,\r\n 2856, 2857, 2858, 2859, 2860, 2861, 2862, 2863, 2864, 2865, 2866, 2867,\r\n 2868, 2869, 2870, 2871, 2872, 2873, 2874, 2875, 2876, 2877, 2878, 2879,\r\n 2880, 2881, 2882, 2883, 2884, 2885, 2886, 2887, 2888, 2889, 2890, 2891,\r\n 2892, 2893, 2894, 2895, 2896, 2897, 2898, 2899, 2900, 2901, 2902, 2903,\r\n 2904, 2905, 2906, 2907, 2908, 2909, 2910, 2911, 2912, 2913, 2914, 2915,\r\n 2916, 2917, 2918, 2919, 2920, 2921, 2922, 2923, 2924, 2925, 2926, 2927,\r\n 2928, 2929, 2930, 2931, 2932, 2933, 2934, 2935, 2936, 2937, 2938, 2939,\r\n 2940, 2941, 2942, 2943, 2944, 2945, 2946, 2947, 2948, 2949, 2950, 2951,\r\n 2952, 2953, 2954, 2955, 2956, 2957, 2958, 2959, 2960, 2961, 2962, 2963,\r\n 2964, 2965, 2966, 2967, 2968, 2969, 2970, 2971, 2972, 2973, 2974, 2975,\r\n 2976, 2977, 2978, 2979, 2980, 2981, 2982, 2983, 2984, 2985, 2986, 2987,\r\n 2988, 2989, 2990, 2991, 2992, 2993, 2994, 2995, 2996, 2997, 2998, 2999,\r\n 3000, 3001, 3002, 3003, 3004, 3005, 3006, 3007, 3008, 3009, 3010, 3011,\r\n 3012, 3013, 3014, 3015, 3016, 3017, 3018, 3019, 3020, 3021, 3022, 3023,\r\n 3024, 3025, 3026, 3027, 3028, 3029, 3030, 3031, 3032, 3033, 3034, 3035,\r\n 3036, 3037, 3038, 3039, 3040, 3041, 3042, 3043, 3044, 3045, 3046, 3047,\r\n 3048, 3049, 3050, 3051, 3052, 3053, 3054, 3055, 3056, 3057, 3058, 3059,\r\n 3060, 3061, 3062, 3063, 3064, 3065, 3066, 3067, 3068, 3069, 3070, 3071,\r\n 3072, 3073, 3074, 3075, 3076, 3077, 3078, 3079, 3080, 3081, 3082, 3083,\r\n 3084, 3085, 3086, 3087, 3088, 3089, 3090, 3091, 3092, 3093, 3094, 3095,\r\n 3096, 3097, 3098, 3099, 3100, 3101, 3102, 3103, 3104, 3105, 3106, 3107,\r\n 3108, 3109, 3110, 3111, 3112, 3113, 3114, 3115, 3116, 3117, 3118, 3119,\r\n 3120, 3121, 3122, 3123, 3124, 3125, 3126, 3127, 3128, 3129, 3130, 3131,\r\n 3132, 3133, 3134, 3135, 3136, 3137, 3138, 3139, 3140, 3141, 3142, 3143,\r\n 3144, 3145, 3146, 3147, 3148, 3149, 3150, 3151, 3152, 3153, 3154, 3155,\r\n 3156, 3157, 3158, 3159, 3160, 3161, 3162, 3163, 3164, 3165, 3166, 3167,\r\n 3168, 3169, 3170, 3171, 3172, 3173, 3174, 3175, 3176, 3177, 3178, 3179,\r\n 3180, 3181, 3182, 3183, 3184, 3185, 3186, 3187, 3188, 3189, 3190, 3191,\r\n 3192, 3193, 3194, 3195, 3196, 3197, 3198, 3199, 3200, 3201, 3202, 3203,\r\n 3204, 3205, 3206, 3207, 3208, 3209, 3210, 3211, 3212, 3213, 3214, 3215,\r\n 3216, 3217, 3218, 3219, 3220, 3221, 3222, 3223, 3224, 3225, 3226, 3227,\r\n 3228, 3229, 3230, 3231, 3232, 3233, 3234, 3235, 3236, 3237, 3238, 3239,\r\n 3240, 3241, 3242, 3243, 3244, 3245, 3246, 3247, 3248, 3249, 3250, 3251,\r\n 3252, 3253, 3254, 3255, 3256, 3257, 3258, 3259, 3260, 3261, 3262, 3263,\r\n 3264, 3265, 3266, 3267, 3268, 3269, 3270, 3271, 3272, 3273, 3274, 3275,\r\n 3276, 3277, 3278, 3279, 3280, 3281, 3282, 3283, 3284, 3285, 3286, 3287,\r\n 3288, 3289, 3290, 3291, 3292, 3293, 3294, 3295, 3296, 3297, 3298, 3299,\r\n 3300, 3301, 3302, 3303, 3304, 3305, 3306, 3307, 3308, 3309, 3310, 3311,\r\n 3312, 3313, 3314, 3315, 3316, 3317, 3318, 3319, 3320, 3321, 3322, 3323,\r\n 3324, 3325, 3326, 3327, 3328, 3329, 3330, 3331, 3332, 3333, 3334, 3335,\r\n 3336, 3337, 3338, 3339, 3340, 3341, 3342, 3343, 3344, 3345, 3346, 3347,\r\n 3348, 3349, 3350, 3351, 3352, 3353, 3354, 3355, 3356, 3357, 3358, 3359,\r\n 3360, 3361, 3362, 3363, 3364, 3365, 3366, 3367, 3368, 3369, 3370, 3371,\r\n 3372, 3373, 3374, 3375, 3376, 3377, 3378, 3379, 3380, 3381, 3382, 3383,\r\n 3384, 3385, 3386, 3387, 3388, 3389, 3390, 3391, 3392, 3393, 3394, 3395,\r\n 3396, 3397, 3398, 3399, 3400, 3401, 3402, 3403, 3404, 3405, 3406, 3407,\r\n 3408, 3409, 3410, 3411, 3412, 3413, 3414, 3415, 3416, 3417, 3418, 3419,\r\n 3420, 3421, 3422, 3423, 3424, 3425, 3426, 3427, 3428, 3429, 3430, 3431,\r\n 3432, 3433, 3434, 3435, 3436, 3437, 3438, 3439, 3440, 3441, 3442, 3443,\r\n 3444, 3445, 3446, 3447, 3448, 3449, 3450, 3451, 3452, 3453, 3454, 3455,\r\n 3456, 3457, 3458, 3459, 3460, 3461, 3462, 3463, 3464, 3465, 3466, 3467,\r\n 3468, 3469, 3470, 3471, 3472, 3473, 3474, 3475, 3476, 3477, 3478, 3479,\r\n 3480, 3481, 3482, 3483, 3484, 3485, 3486, 3487, 3488, 3489, 3490, 3491,\r\n 3492, 3493, 3494, 3495, 3496, 3497, 3498, 3499, 3500, 3501, 3502, 3503,\r\n 3504, 3505, 3506, 3507, 3508, 3509, 3510, 3511, 3512, 3513, 3514, 3515,\r\n 3516, 3517, 3518, 3519, 3520, 3521, 3522, 3523, 3524, 3525, 3526, 3527,\r\n 3528, 3529, 3530, 3531, 3532, 3533, 3534, 3535, 3536, 3537, 3538, 3539,\r\n 3540, 3541, 3542, 3543, 3544, 3545, 3546, 3547, 3548, 3549, 3550, 3551,\r\n 3552, 3553, 3554, 3555, 3556, 3557, 3558, 3559, 3560, 3561, 3562, 3563,\r\n 3564, 3565, 3566, 3567, 3568, 3569, 3570, 3571, 3572, 3573, 3574, 3575,\r\n 3576, 3577, 3578, 3579, 3580, 3581, 3582, 3583, 3584, 3585, 3586, 3587,\r\n 3588, 3589, 3590, 3591, 3592, 3593, 3594, 3595, 3596, 3597, 3598, 3599,\r\n 3600, 3601, 3602, 3603, 3604, 3605, 3606, 3607, 3608, 3609, 3610, 3611,\r\n 3612, 3613, 3614, 3615, 3616, 3617, 3618, 3619, 3620, 3621, 3622, 3623,\r\n 3624, 3625, 3626, 3627, 3628, 3629, 3630, 3631, 3632, 3633, 3634, 3635,\r\n 3636, 3637, 3638, 3639, 3640, 3641, 3642, 3643, 3644, 3645, 3646, 3647,\r\n 3648, 3649, 3650, 3651, 3652, 3653, 3654, 3655, 3656, 3657, 3658, 3659,\r\n 3660, 3661, 3662, 3663, 3664, 3665, 3666, 3667, 3668, 3669, 3670, 3671,\r\n 3672, 3673, 3674, 3675, 3676, 3677, 3678, 3679, 3680, 3681, 3682, 3683,\r\n 3684, 3685, 3686, 3687, 3688, 3689, 3690, 3691, 3692, 3693, 3694, 3695,\r\n 3696, 3697, 3698, 3699, 3700, 3701, 3702, 3703, 3704, 3705, 3706, 3707,\r\n 3708, 3709, 3710, 3711, 3712, 3713, 3714, 3715, 3716, 3717, 3718, 3719,\r\n 3720, 3721, 3722, 3723, 3724, 3725, 3726, 3727, 3728, 3729, 3730, 3731,\r\n 3732, 3733, 3734, 3735, 3736, 3737, 3738, 3739, 3740, 3741, 3742, 3743,\r\n 3744, 3745, 3746, 3747, 3748, 3749, 3750, 3751, 3752, 3753, 3754, 3755,\r\n 3756, 3757, 3758, 3759, 3760, 3761, 3762, 3763, 3764, 3765, 3766, 3767,\r\n 3768, 3769, 3770, 3771, 3772, 3773, 3774, 3775, 3776, 3777, 3778, 3779,\r\n 3780, 3781, 3782, 3783, 3784, 3785, 3786, 3787, 3788, 3789, 3790, 3791,\r\n 3792, 3793, 3794, 3795, 3796, 3797, 3798, 3799, 3800, 3801, 3802, 3803,\r\n 3804, 3805, 3806, 3807, 3808, 3809, 3810, 3811, 3812, 3813, 3814, 3815,\r\n 3816, 3817, 3818, 3819, 3820, 3821, 3822, 3823, 3824, 3825, 3826, 3827,\r\n 3828, 3829, 3830, 3831, 3832, 3833, 3834, 3835, 3836, 3837, 3838, 3839,\r\n 3840, 3841, 3842, 3843, 3844, 3845, 3846, 3847, 3848, 3849, 3850, 3851,\r\n 3852, 3853, 3854, 3855, 3856, 3857, 3858, 3859, 3860, 3861, 3862, 3863,\r\n 3864, 3865, 3866, 3867, 3868, 3869, 3870, 3871, 3872, 3873, 3874, 3875,\r\n 3876, 3877, 3878, 3879, 3880, 3881, 3882, 3883, 3884, 3885, 3886, 3887,\r\n 3888, 3889, 3890, 3891, 3892, 3893, 3894, 3895, 3896, 3897, 3898, 3899,\r\n 3900, 3901, 3902, 3903, 3904, 3905, 3906, 3907, 3908, 3909, 3910, 3911,\r\n 3912, 3913, 3914, 3915, 3916, 3917, 3918, 3919, 3920, 3921, 3922, 3923,\r\n 3924, 3925, 3926, 3927, 3928, 3929, 3930, 3931, 3932, 3933, 3934, 3935,\r\n 3936, 3937, 3938, 3939, 3940, 3941, 3942, 3943, 3944, 3945, 3946, 3947,\r\n 3948, 3949, 3950, 3951, 3952, 3953, 3954, 3955, 3956, 3957, 3958, 3959,\r\n 3960, 3961, 3962, 3963, 3964, 3965, 3966, 3967, 3968, 3969, 3970, 3971,\r\n 3972, 3973, 3974, 3975, 3976, 3977, 3978, 3979, 3980, 3981, 3982, 3983,\r\n 3984, 3985, 3986, 3987, 3988, 3989, 3990, 3991, 3992, 3993, 3994, 3995,\r\n 3996, 3997, 3998, 3999, 4000, 4001, 4002, 4003, 4004, 4005, 4006, 4007,\r\n 4008, 4009, 4010, 4011, 4012, 4013, 4014, 4015, 4016, 4017, 4018, 4019,\r\n 4020, 4021, 4022, 4023, 4024, 4025, 4026, 4027, 4028, 4029, 4030, 4031,\r\n 4032, 4033, 4034, 4035, 4036, 4037, 4038, 4039, 4040, 4041, 4042, 4043,\r\n 4044, 4045, 4046, 4047, 4048, 4049, 4050, 4051, 4052, 4053, 4054, 4055,\r\n 4056, 4057, 4058, 4059, 4060, 4061, 4062, 4063, 4064, 4065, 4066, 4067,\r\n 4068, 4069, 4070, 4071, 4072, 4073, 4074, 4075, 4076, 4077, 4078, 4079,\r\n 4080, 4081, 4082, 4083, 4084, 4085, 4086, 4087, 4088, 4089, 4090, 4091,\r\n 4092, 4093, 4094, 4095, 4096, 4097]]) torch.Size([1, 4098])\r\ntensor([[ 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,\r\n 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,\r\n 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37,\r\n 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49,\r\n 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61,\r\n 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73,\r\n 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85,\r\n 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97,\r\n 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109,\r\n 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121,\r\n 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133,\r\n 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145,\r\n 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157,\r\n 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169,\r\n 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181,\r\n 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193,\r\n 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205,\r\n 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217,\r\n 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229,\r\n 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241,\r\n 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253,\r\n 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265,\r\n 266, 267, 268, 269, 270, 271, 272, 273, 274, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1],\r\n [ 2, 3, 4, 5, 6, 7, 8, 9, 10, 11, 12, 13,\r\n 14, 15, 16, 17, 18, 19, 20, 21, 22, 23, 24, 25,\r\n 26, 27, 28, 29, 30, 31, 32, 33, 34, 35, 36, 37,\r\n 38, 39, 40, 41, 42, 43, 44, 45, 46, 47, 48, 49,\r\n 50, 51, 52, 53, 54, 55, 56, 57, 58, 59, 60, 61,\r\n 62, 63, 64, 65, 66, 67, 68, 69, 70, 71, 72, 73,\r\n 74, 75, 76, 77, 78, 79, 80, 81, 82, 83, 84, 85,\r\n 86, 87, 88, 89, 90, 91, 92, 93, 94, 95, 96, 97,\r\n 98, 99, 100, 101, 102, 103, 104, 105, 106, 107, 108, 109,\r\n 110, 111, 112, 113, 114, 115, 116, 117, 118, 119, 120, 121,\r\n 122, 123, 124, 125, 126, 127, 128, 129, 130, 131, 132, 133,\r\n 134, 135, 136, 137, 138, 139, 140, 141, 142, 143, 144, 145,\r\n 146, 147, 148, 149, 150, 151, 152, 153, 154, 155, 156, 157,\r\n 158, 159, 160, 161, 162, 163, 164, 165, 166, 167, 168, 169,\r\n 170, 171, 172, 173, 174, 175, 176, 177, 178, 179, 180, 181,\r\n 182, 183, 184, 185, 186, 187, 188, 189, 190, 191, 192, 193,\r\n 194, 195, 196, 197, 198, 199, 200, 201, 202, 203, 204, 205,\r\n 206, 207, 208, 209, 210, 211, 212, 213, 214, 215, 216, 217,\r\n 218, 219, 220, 221, 222, 223, 224, 225, 226, 227, 228, 229,\r\n 230, 231, 232, 233, 234, 235, 236, 237, 238, 239, 240, 241,\r\n 242, 243, 244, 245, 246, 247, 248, 249, 250, 251, 252, 253,\r\n 254, 255, 256, 257, 258, 259, 260, 261, 262, 263, 264, 265,\r\n 266, 267, 268, 269, 270, 271, 272, 273, 274, 275, 276, 277,\r\n 278, 279, 280, 281, 282, 283, 284, 285, 286, 287, 288, 289,\r\n 290, 291, 292, 293, 294, 295, 296, 297, 298, 299, 300, 301,\r\n 302, 303, 304, 305, 306, 307, 308, 309, 310, 311, 312, 313,\r\n 314, 315, 316, 317, 318, 319, 320, 321, 322, 323, 324, 325,\r\n 326, 327, 328, 329, 330, 331, 332, 333, 334, 335, 336, 337,\r\n 338, 339, 340, 341, 342, 343, 344, 345, 346, 347, 348, 349,\r\n 350, 351, 352, 353, 354, 355, 356, 357, 358, 359, 360, 361,\r\n 362, 363, 364, 365, 366, 367, 368, 369, 370, 371, 372, 373,\r\n 374, 375, 376, 377, 378, 379, 380, 381, 382, 383, 384, 385,\r\n 386, 387, 388, 389, 390, 391, 392, 393, 394, 395, 396, 397,\r\n 398, 399, 400, 401, 402, 403, 404, 405, 406, 407, 408, 409,\r\n 410, 411, 412, 413, 414, 415, 416, 417, 418, 419, 420, 421,\r\n 422, 423, 424, 425, 426, 427, 428, 429, 430, 431, 432, 433,\r\n 434, 435, 436, 437, 438, 439, 440, 441, 442, 443, 444, 445,\r\n 446, 447, 448, 449, 450, 451, 452, 453, 454, 455, 456, 457,\r\n 458, 459, 460, 461, 462, 463, 464, 465, 466, 467, 468, 469,\r\n 470, 471, 472, 473, 474, 475, 476, 477, 478, 479, 480, 481,\r\n 482, 483, 484, 485, 486, 487, 488, 489, 490, 491, 492, 493,\r\n 494, 495, 496, 497, 498, 499, 500, 501, 502, 503, 504, 505,\r\n 506, 507, 508, 509, 510, 511, 512, 513, 514, 515, 516, 517,\r\n 518, 519, 520, 521, 522, 523, 524, 525, 526, 527, 528, 529,\r\n 530, 531, 532, 533, 534, 535, 536, 537, 538, 539, 540, 541,\r\n 542, 543, 544, 545, 546, 547, 548, 549, 550, 551, 552, 553,\r\n 554, 555, 556, 557, 558, 559, 560, 561, 562, 563, 564, 565,\r\n 566, 567, 568, 569, 570, 571, 572, 573, 574, 575, 576, 577,\r\n 578, 579, 580, 581, 582, 583, 584, 585, 586, 587, 588, 589,\r\n 590, 591, 592, 593, 594, 595, 596, 597, 598, 599, 600, 601,\r\n 602, 603, 604, 605, 606, 607, 608, 609, 610, 611, 612, 613,\r\n 614, 615, 616, 617, 618, 619, 620, 621, 622, 623, 624, 625,\r\n 626, 627, 628, 629, 630, 631, 632, 633, 634, 635, 636, 637,\r\n 638, 639, 640, 641, 642, 643, 644, 645, 646, 647, 648, 649,\r\n 650, 651, 652, 653, 654, 655, 656, 657, 658, 659, 660, 661,\r\n 662, 663, 664, 665, 666, 667, 668, 669, 670, 671, 672, 673,\r\n 674, 675, 676, 677, 678, 679, 680, 681, 682, 683, 684, 685,\r\n 686, 687, 688, 689, 690, 691, 692, 693, 694, 695, 696, 697,\r\n 698, 699, 700, 701, 702, 703, 704, 705, 706, 707, 708, 709,\r\n 710, 711, 712, 713, 714, 715, 716, 717, 718, 719, 720, 721,\r\n 722, 723, 724, 725, 726, 727, 728, 729, 730, 731, 732, 733,\r\n 734, 735, 736, 737, 738, 739, 740, 741, 742, 743, 744, 745,\r\n 746, 747, 748, 749, 750, 751, 752, 753, 754, 755, 756, 757,\r\n 758, 759, 760, 761, 762, 763, 764, 765, 766, 767, 768, 769,\r\n 770, 771, 772, 773, 774, 775, 776, 777, 778, 779, 780, 781,\r\n 782, 783, 784, 785, 786, 787, 788, 789, 790, 791, 792, 793,\r\n 794, 795, 796, 797, 798, 799, 800, 801, 802, 803, 804, 805,\r\n 806, 807, 808, 809, 810, 811, 812, 813, 814, 815, 816, 817,\r\n 818, 819, 820, 821, 822, 823, 824, 825, 826, 827, 828, 829,\r\n 830, 831, 832, 833, 834, 835, 836, 837, 838, 839, 840, 841,\r\n 842, 843, 844, 845, 846, 847, 848, 849, 850, 851, 852, 853,\r\n 854, 855, 856, 857, 858, 859, 860, 861, 862, 863, 864, 865,\r\n 866, 867, 868, 869, 870, 871, 872, 873, 874, 875, 876, 877,\r\n 878, 879, 880, 881, 882, 883, 884, 885, 886, 887, 888, 889,\r\n 890, 891, 892, 893, 894, 895, 896, 897, 898, 899, 900, 901,\r\n 902, 903, 904, 905, 906, 907, 908, 909, 910, 911, 912, 913,\r\n 914, 915, 916, 917, 918, 919, 920, 921, 922, 923, 924, 925,\r\n 926, 927, 928, 929, 930, 931, 932, 933, 934, 935, 936, 937,\r\n 938, 939, 940, 941, 942, 943, 944, 945, 946, 947, 948, 949,\r\n 950, 951, 952, 953, 954, 955, 956, 957, 958, 959, 960, 961,\r\n 962, 963, 964, 965, 966, 967, 968, 969, 970, 971, 972, 973,\r\n 974, 975, 976, 977, 978, 979, 980, 981, 982, 983, 984, 985,\r\n 986, 987, 988, 989, 990, 991, 992, 993, 994, 995, 996, 997,\r\n 998, 999, 1000, 1001, 1002, 1003, 1004, 1005, 1006, 1007, 1008, 1009,\r\n 1010, 1011, 1012, 1013, 1014, 1015, 1016, 1017, 1018, 1019, 1020, 1021,\r\n 1022, 1023, 1024, 1025, 1026, 1027, 1028, 1029, 1030, 1031, 1032, 1033,\r\n 1034, 1035, 1036, 1037, 1038, 1039, 1040, 1041, 1042, 1043, 1044, 1045,\r\n 1046, 1047, 1048, 1049, 1050, 1051, 1052, 1053, 1054, 1055, 1056, 1057,\r\n 1058, 1059, 1060, 1061, 1062, 1063, 1064, 1065, 1066, 1067, 1068, 1069,\r\n 1070, 1071, 1072, 1073, 1074, 1075, 1076, 1077, 1078, 1079, 1080, 1081,\r\n 1082, 1083, 1084, 1085, 1086, 1087, 1088, 1089, 1090, 1091, 1092, 1093,\r\n 1094, 1095, 1096, 1097, 1098, 1099, 1100, 1101, 1102, 1103, 1104, 1105,\r\n 1106, 1107, 1108, 1109, 1110, 1111, 1112, 1113, 1114, 1115, 1116, 1117,\r\n 1118, 1119, 1120, 1121, 1122, 1123, 1124, 1125, 1126, 1127, 1128, 1129,\r\n 1130, 1131, 1132, 1133, 1134, 1135, 1136, 1137, 1138, 1139, 1140, 1141,\r\n 1142, 1143, 1144, 1145, 1146, 1147, 1148, 1149, 1150, 1151, 1152, 1153,\r\n 1154, 1155, 1156, 1157, 1158, 1159, 1160, 1161, 1162, 1163, 1164, 1165,\r\n 1166, 1167, 1168, 1169, 1170, 1171, 1172, 1173, 1174, 1175, 1176, 1177,\r\n 1178, 1179, 1180, 1181, 1182, 1183, 1184, 1185, 1186, 1187, 1188, 1189,\r\n 1190, 1191, 1192, 1193, 1194, 1195, 1196, 1197, 1198, 1199, 1200, 1201,\r\n 1202, 1203, 1204, 1205, 1206, 1207, 1208, 1209, 1210, 1211, 1212, 1213,\r\n 1214, 1215, 1216, 1217, 1218, 1219, 1220, 1221, 1222, 1223, 1224, 1225,\r\n 1226, 1227, 1228, 1229, 1230, 1231, 1232, 1233, 1234, 1235, 1236, 1237,\r\n 1238, 1239, 1240, 1241, 1242, 1243, 1244, 1245, 1246, 1247, 1248, 1249,\r\n 1250, 1251, 1252, 1253, 1254, 1255, 1256, 1257, 1258, 1259, 1260, 1261,\r\n 1262, 1263, 1264, 1265, 1266, 1267, 1268, 1269, 1270, 1271, 1272, 1273,\r\n 1274, 1275, 1276, 1277, 1278, 1279, 1280, 1281, 1282, 1283, 1284, 1285,\r\n 1286, 1287, 1288, 1289, 1290, 1291, 1292, 1293, 1294, 1295, 1296, 1297,\r\n 1298, 1299, 1300, 1301, 1302, 1303, 1304, 1305, 1306, 1307, 1308, 1309,\r\n 1310, 1311, 1312, 1313, 1314, 1315, 1316, 1317, 1318, 1319, 1320, 1321,\r\n 1322, 1323, 1324, 1325, 1326, 1327, 1328, 1329, 1330, 1331, 1332, 1333,\r\n 1334, 1335, 1336, 1337, 1338, 1339, 1340, 1341, 1342, 1343, 1344, 1345,\r\n 1346, 1347, 1348, 1349, 1350, 1351, 1352, 1353, 1354, 1355, 1356, 1357,\r\n 1358, 1359, 1360, 1361, 1362, 1363, 1364, 1365, 1366, 1367, 1368, 1369,\r\n 1370, 1371, 1372, 1373, 1374, 1375, 1376, 1377, 1378, 1379, 1380, 1381,\r\n 1382, 1383, 1384, 1385, 1386, 1387, 1388, 1389, 1390, 1391, 1392, 1393,\r\n 1394, 1395, 1396, 1397, 1398, 1399, 1400, 1401, 1402, 1403, 1404, 1405,\r\n 1406, 1407, 1408, 1409, 1410, 1411, 1412, 1413, 1414, 1415, 1416, 1417,\r\n 1418, 1419, 1420, 1421, 1422, 1423, 1424, 1425, 1426, 1427, 1428, 1429,\r\n 1430, 1431, 1432, 1433, 1434, 1435, 1436, 1437, 1438, 1439, 1440, 1441,\r\n 1442, 1443, 1444, 1445, 1446, 1447, 1448, 1449, 1450, 1451, 1452, 1453,\r\n 1454, 1455, 1456, 1457, 1458, 1459, 1460, 1461, 1462, 1463, 1464, 1465,\r\n 1466, 1467, 1468, 1469, 1470, 1471, 1472, 1473, 1474, 1475, 1476, 1477,\r\n 1478, 1479, 1480, 1481, 1482, 1483, 1484, 1485, 1486, 1487, 1488, 1489,\r\n 1490, 1491, 1492, 1493, 1494, 1495, 1496, 1497, 1498, 1499, 1500, 1501,\r\n 1502, 1503, 1504, 1505, 1506, 1507, 1508, 1509, 1510, 1511, 1512, 1513,\r\n 1514, 1515, 1516, 1517, 1518, 1519, 1520, 1521, 1522, 1523, 1524, 1525,\r\n 1526, 1527, 1528, 1529, 1530, 1531, 1532, 1533, 1534, 1535, 1536, 1537,\r\n 1538, 1539, 1540, 1541, 1542, 1543, 1544, 1545, 1546, 1547, 1548, 1549,\r\n 1550, 1551, 1552, 1553, 1554, 1555, 1556, 1557, 1558, 1559, 1560, 1561,\r\n 1562, 1563, 1564, 1565, 1566, 1567, 1568, 1569, 1570, 1571, 1572, 1573,\r\n 1574, 1575, 1576, 1577, 1578, 1579, 1580, 1581, 1582, 1583, 1584, 1585,\r\n 1586, 1587, 1588, 1589, 1590, 1591, 1592, 1593, 1594, 1595, 1596, 1597,\r\n 1598, 1599, 1600, 1601, 1602, 1603, 1604, 1605, 1606, 1607, 1608, 1609,\r\n 1610, 1611, 1612, 1613, 1614, 1615, 1616, 1617, 1618, 1619, 1620, 1621,\r\n 1622, 1623, 1624, 1625, 1626, 1627, 1628, 1629, 1630, 1631, 1632, 1633,\r\n 1634, 1635, 1636, 1637, 1638, 1639, 1640, 1641, 1642, 1643, 1644, 1645,\r\n 1646, 1647, 1648, 1649, 1650, 1651, 1652, 1653, 1654, 1655, 1656, 1657,\r\n 1658, 1659, 1660, 1661, 1662, 1663, 1664, 1665, 1666, 1667, 1668, 1669,\r\n 1670, 1671, 1672, 1673, 1674, 1675, 1676, 1677, 1678, 1679, 1680, 1681,\r\n 1682, 1683, 1684, 1685, 1686, 1687, 1688, 1689, 1690, 1691, 1692, 1693,\r\n 1694, 1695, 1696, 1697, 1698, 1699, 1700, 1701, 1702, 1703, 1704, 1705,\r\n 1706, 1707, 1708, 1709, 1710, 1711, 1712, 1713, 1714, 1715, 1716, 1717,\r\n 1718, 1719, 1720, 1721, 1722, 1723, 1724, 1725, 1726, 1727, 1728, 1729,\r\n 1730, 1731, 1732, 1733, 1734, 1735, 1736, 1737, 1738, 1739, 1740, 1741,\r\n 1742, 1743, 1744, 1745, 1746, 1747, 1748, 1749, 1750, 1751, 1752, 1753,\r\n 1754, 1755, 1756, 1757, 1758, 1759, 1760, 1761, 1762, 1763, 1764, 1765,\r\n 1766, 1767, 1768, 1769, 1770, 1771, 1772, 1773, 1774, 1775, 1776, 1777,\r\n 1778, 1779, 1780, 1781, 1782, 1783, 1784, 1785, 1786, 1787, 1788, 1789,\r\n 1790, 1791, 1792, 1793, 1794, 1795, 1796, 1797, 1798, 1799, 1800, 1801,\r\n 1802, 1803, 1804, 1805, 1806, 1807, 1808, 1809, 1810, 1811, 1812, 1813,\r\n 1814, 1815, 1816, 1817, 1818, 1819, 1820, 1821, 1822, 1823, 1824, 1825,\r\n 1826, 1827, 1828, 1829, 1830, 1831, 1832, 1833, 1834, 1835, 1836, 1837,\r\n 1838, 1839, 1840, 1841, 1842, 1843, 1844, 1845, 1846, 1847, 1848, 1849,\r\n 1850, 1851, 1852, 1853, 1854, 1855, 1856, 1857, 1858, 1859, 1860, 1861,\r\n 1862, 1863, 1864, 1865, 1866, 1867, 1868, 1869, 1870, 1871, 1872, 1873,\r\n 1874, 1875, 1876, 1877, 1878, 1879, 1880, 1881, 1882, 1883, 1884, 1885,\r\n 1886, 1887, 1888, 1889, 1890, 1891, 1892, 1893, 1894, 1895, 1896, 1897,\r\n 1898, 1899, 1900, 1901, 1902, 1903, 1904, 1905, 1906, 1907, 1908, 1909,\r\n 1910, 1911, 1912, 1913, 1914, 1915, 1916, 1917, 1918, 1919, 1920, 1921,\r\n 1922, 1923, 1924, 1925, 1926, 1927, 1928, 1929, 1930, 1931, 1932, 1933,\r\n 1934, 1935, 1936, 1937, 1938, 1939, 1940, 1941, 1942, 1943, 1944, 1945,\r\n 1946, 1947, 1948, 1949, 1950, 1951, 1952, 1953, 1954, 1955, 1956, 1957,\r\n 1958, 1959, 1960, 1961, 1962, 1963, 1964, 1965, 1966, 1967, 1968, 1969,\r\n 1970, 1971, 1972, 1973, 1974, 1975, 1976, 1977, 1978, 1979, 1980, 1981,\r\n 1982, 1983, 1984, 1985, 1986, 1987, 1988, 1989, 1990, 1991, 1992, 1993,\r\n 1994, 1995, 1996, 1997, 1998, 1999, 2000, 2001, 2002, 2003, 2004, 2005,\r\n 2006, 2007, 2008, 2009, 2010, 2011, 2012, 2013, 2014, 2015, 2016, 2017,\r\n 2018, 2019, 2020, 2021, 2022, 2023, 2024, 2025, 2026, 2027, 2028, 2029,\r\n 2030, 2031, 2032, 2033, 2034, 2035, 2036, 2037, 2038, 2039, 2040, 2041,\r\n 2042, 2043, 2044, 2045, 2046, 2047, 2048, 2049, 2050, 2051, 2052, 2053,\r\n 2054, 2055, 2056, 2057, 2058, 2059, 2060, 2061, 2062, 2063, 2064, 2065,\r\n 2066, 2067, 2068, 2069, 2070, 2071, 2072, 2073, 2074, 2075, 2076, 2077,\r\n 2078, 2079, 2080, 2081, 2082, 2083, 2084, 2085, 2086, 2087, 2088, 2089,\r\n 2090, 2091, 2092, 2093, 2094, 2095, 2096, 2097, 2098, 2099, 2100, 2101,\r\n 2102, 2103, 2104, 2105, 2106, 2107, 2108, 2109, 2110, 2111, 2112, 2113,\r\n 2114, 2115, 2116, 2117, 2118, 2119, 2120, 2121, 2122, 2123, 2124, 2125,\r\n 2126, 2127, 2128, 2129, 2130, 2131, 2132, 2133, 2134, 2135, 2136, 2137,\r\n 2138, 2139, 2140, 2141, 2142, 2143, 2144, 2145, 2146, 2147, 2148, 2149,\r\n 2150, 2151, 2152, 2153, 2154, 2155, 2156, 2157, 2158, 2159, 2160, 2161,\r\n 2162, 2163, 2164, 2165, 2166, 2167, 2168, 2169, 2170, 2171, 2172, 2173,\r\n 2174, 2175, 2176, 2177, 2178, 2179, 2180, 2181, 2182, 2183, 2184, 2185,\r\n 2186, 2187, 2188, 2189, 2190, 2191, 2192, 2193, 2194, 2195, 2196, 2197,\r\n 2198, 2199, 2200, 2201, 2202, 2203, 2204, 2205, 2206, 2207, 2208, 2209,\r\n 2210, 2211, 2212, 2213, 2214, 2215, 2216, 2217, 2218, 2219, 2220, 2221,\r\n 2222, 2223, 2224, 2225, 2226, 2227, 2228, 2229, 2230, 2231, 2232, 2233,\r\n 2234, 2235, 2236, 2237, 2238, 2239, 2240, 2241, 2242, 2243, 2244, 2245,\r\n 2246, 2247, 2248, 2249, 2250, 2251, 2252, 2253, 2254, 2255, 2256, 2257,\r\n 2258, 2259, 2260, 2261, 2262, 2263, 2264, 2265, 2266, 2267, 2268, 2269,\r\n 2270, 2271, 2272, 2273, 2274, 2275, 2276, 2277, 2278, 2279, 2280, 2281,\r\n 2282, 2283, 2284, 2285, 2286, 2287, 2288, 2289, 2290, 2291, 2292, 2293,\r\n 2294, 2295, 2296, 2297, 2298, 2299, 2300, 2301, 2302, 2303, 2304, 2305,\r\n 2306, 2307, 2308, 2309, 2310, 2311, 2312, 2313, 2314, 2315, 2316, 2317,\r\n 2318, 2319, 2320, 2321, 2322, 2323, 2324, 2325, 2326, 2327, 2328, 2329,\r\n 2330, 2331, 2332, 2333, 2334, 2335, 2336, 2337, 2338, 2339, 2340, 2341,\r\n 2342, 2343, 2344, 2345, 2346, 2347, 2348, 2349, 2350, 2351, 2352, 2353,\r\n 2354, 2355, 2356, 2357, 2358, 2359, 2360, 2361, 2362, 2363, 2364, 2365,\r\n 2366, 2367, 2368, 2369, 2370, 2371, 2372, 2373, 2374, 2375, 2376, 2377,\r\n 2378, 2379, 2380, 2381, 2382, 2383, 2384, 2385, 2386, 2387, 2388, 2389,\r\n 2390, 2391, 2392, 2393, 2394, 2395, 2396, 2397, 2398, 2399, 2400, 2401,\r\n 2402, 2403, 2404, 2405, 2406, 2407, 2408, 2409, 2410, 2411, 2412, 2413,\r\n 2414, 2415, 2416, 2417, 2418, 2419, 2420, 2421, 2422, 2423, 2424, 2425,\r\n 2426, 2427, 2428, 2429, 2430, 2431, 2432, 2433, 2434, 2435, 2436, 2437,\r\n 2438, 2439, 2440, 2441, 2442, 2443, 2444, 2445, 2446, 2447, 2448, 2449,\r\n 2450, 2451, 2452, 2453, 2454, 2455, 2456, 2457, 2458, 2459, 2460, 2461,\r\n 2462, 2463, 2464, 2465, 2466, 2467, 2468, 2469, 2470, 2471, 2472, 2473,\r\n 2474, 2475, 2476, 2477, 2478, 2479, 2480, 2481, 2482, 2483, 2484, 2485,\r\n 2486, 2487, 2488, 2489, 2490, 2491, 2492, 2493, 2494, 2495, 2496, 2497,\r\n 2498, 2499, 2500, 2501, 2502, 2503, 2504, 2505, 2506, 2507, 2508, 2509,\r\n 2510, 2511, 2512, 2513, 2514, 2515, 2516, 2517, 2518, 2519, 2520, 2521,\r\n 2522, 2523, 2524, 2525, 2526, 2527, 2528, 2529, 2530, 2531, 2532, 2533,\r\n 2534, 2535, 2536, 2537, 2538, 2539, 2540, 2541, 2542, 2543, 2544, 2545,\r\n 2546, 2547, 2548, 2549, 2550, 2551, 2552, 2553, 2554, 2555, 2556, 2557,\r\n 2558, 2559, 2560, 2561, 2562, 2563, 2564, 2565, 2566, 2567, 2568, 2569,\r\n 2570, 2571, 2572, 2573, 2574, 2575, 2576, 2577, 2578, 2579, 2580, 2581,\r\n 2582, 2583, 2584, 2585, 2586, 2587, 2588, 2589, 2590, 2591, 2592, 2593,\r\n 2594, 2595, 2596, 2597, 2598, 2599, 2600, 2601, 2602, 2603, 2604, 2605,\r\n 2606, 2607, 2608, 2609, 2610, 2611, 2612, 2613, 2614, 2615, 2616, 2617,\r\n 2618, 2619, 2620, 2621, 2622, 2623, 2624, 2625, 2626, 2627, 2628, 2629,\r\n 2630, 2631, 2632, 2633, 2634, 2635, 2636, 2637, 2638, 2639, 2640, 2641,\r\n 2642, 2643, 2644, 2645, 2646, 2647, 2648, 2649, 2650, 2651, 2652, 2653,\r\n 2654, 2655, 2656, 2657, 2658, 2659, 2660, 2661, 2662, 2663, 2664, 2665,\r\n 2666, 2667, 2668, 2669, 2670, 2671, 2672, 2673, 2674, 2675, 2676, 2677,\r\n 2678, 2679, 2680, 2681, 2682, 2683, 2684, 2685, 2686, 2687, 2688, 2689,\r\n 2690, 2691, 2692, 2693, 2694, 2695, 2696, 2697, 2698, 2699, 2700, 2701,\r\n 2702, 2703, 2704, 2705, 2706, 2707, 2708, 2709, 2710, 2711, 2712, 2713,\r\n 2714, 2715, 2716, 2717, 2718, 2719, 2720, 2721, 2722, 2723, 2724, 2725,\r\n 2726, 2727, 2728, 2729, 2730, 2731, 2732, 2733, 2734, 2735, 2736, 2737,\r\n 2738, 2739, 2740, 2741, 2742, 2743, 2744, 2745, 2746, 2747, 2748, 2749,\r\n 2750, 2751, 2752, 2753, 2754, 2755, 2756, 2757, 2758, 2759, 2760, 2761,\r\n 2762, 2763, 2764, 2765, 2766, 2767, 2768, 2769, 2770, 2771, 2772, 2773,\r\n 2774, 2775, 2776, 2777, 2778, 2779, 2780, 2781, 2782, 2783, 2784, 2785,\r\n 2786, 2787, 2788, 2789, 2790, 2791, 2792, 2793, 2794, 2795, 2796, 2797,\r\n 2798, 2799, 2800, 2801, 2802, 2803, 2804, 2805, 2806, 2807, 2808, 2809,\r\n 2810, 2811, 2812, 2813, 2814, 2815, 2816, 2817, 2818, 2819, 2820, 2821,\r\n 2822, 2823, 2824, 2825, 2826, 2827, 2828, 2829, 2830, 2831, 2832, 2833,\r\n 2834, 2835, 2836, 2837, 2838, 2839, 2840, 2841, 2842, 2843, 2844, 2845,\r\n 2846, 2847, 2848, 2849, 2850, 2851, 2852, 2853, 2854, 2855, 2856, 2857,\r\n 2858, 2859, 2860, 2861, 2862, 2863, 2864, 2865, 2866, 2867, 2868, 2869,\r\n 2870, 2871, 2872, 2873, 2874, 2875, 2876, 2877, 2878, 2879, 2880, 2881,\r\n 2882, 2883, 2884, 2885, 2886, 2887, 2888, 2889, 2890, 2891, 2892, 2893,\r\n 2894, 2895, 2896, 2897, 2898, 2899, 2900, 2901, 2902, 2903, 2904, 2905,\r\n 2906, 2907, 2908, 2909, 2910, 2911, 2912, 2913, 2914, 2915, 2916, 2917,\r\n 2918, 2919, 2920, 2921, 2922, 2923, 2924, 2925, 2926, 2927, 2928, 2929,\r\n 2930, 2931, 2932, 2933, 2934, 2935, 2936, 2937, 2938, 2939, 2940, 2941,\r\n 2942, 2943, 2944, 2945, 2946, 2947, 2948, 2949, 2950, 2951, 2952, 2953,\r\n 2954, 2955, 2956, 2957, 2958, 2959, 2960, 2961, 2962, 2963, 2964, 2965,\r\n 2966, 2967, 2968, 2969, 2970, 2971, 2972, 2973, 2974, 2975, 2976, 2977,\r\n 2978, 2979, 2980, 2981, 2982, 2983, 2984, 2985, 2986, 2987, 2988, 2989,\r\n 2990, 2991, 2992, 2993, 2994, 2995, 2996, 2997, 2998, 2999, 3000, 3001,\r\n 3002, 3003, 3004, 3005, 3006, 3007, 3008, 3009, 3010, 3011, 3012, 3013,\r\n 3014, 3015, 3016, 3017, 3018, 3019, 3020, 3021, 3022, 3023, 3024, 3025,\r\n 3026, 3027, 3028, 3029, 3030, 3031, 3032, 3033, 3034, 3035, 3036, 3037,\r\n 3038, 3039, 3040, 3041, 3042, 3043, 3044, 3045, 3046, 3047, 3048, 3049,\r\n 3050, 3051, 3052, 3053, 3054, 3055, 3056, 3057, 3058, 3059, 3060, 3061,\r\n 3062, 3063, 3064, 3065, 3066, 3067, 3068, 3069, 3070, 3071, 3072, 3073,\r\n 3074, 3075, 3076, 3077, 3078, 3079, 3080, 3081, 3082, 3083, 3084, 3085,\r\n 3086, 3087, 3088, 3089, 3090, 3091, 3092, 3093, 3094, 3095, 3096, 3097,\r\n 3098, 3099, 3100, 3101, 3102, 3103, 3104, 3105, 3106, 3107, 3108, 3109,\r\n 3110, 3111, 3112, 3113, 3114, 3115, 3116, 3117, 3118, 3119, 3120, 3121,\r\n 3122, 3123, 3124, 3125, 3126, 3127, 3128, 3129, 3130, 3131, 3132, 3133,\r\n 3134, 3135, 3136, 3137, 3138, 3139, 3140, 3141, 3142, 3143, 3144, 3145,\r\n 3146, 3147, 3148, 3149, 3150, 3151, 3152, 3153, 3154, 3155, 3156, 3157,\r\n 3158, 3159, 3160, 3161, 3162, 3163, 3164, 3165, 3166, 3167, 3168, 3169,\r\n 3170, 3171, 3172, 3173, 3174, 3175, 3176, 3177, 3178, 3179, 3180, 3181,\r\n 3182, 3183, 3184, 3185, 3186, 3187, 3188, 3189, 3190, 3191, 3192, 3193,\r\n 3194, 3195, 3196, 3197, 3198, 3199, 3200, 3201, 3202, 3203, 3204, 3205,\r\n 3206, 3207, 3208, 3209, 3210, 3211, 3212, 3213, 3214, 3215, 3216, 3217,\r\n 3218, 3219, 3220, 3221, 3222, 3223, 3224, 3225, 3226, 3227, 3228, 3229,\r\n 3230, 3231, 3232, 3233, 3234, 3235, 3236, 3237, 3238, 3239, 3240, 3241,\r\n 3242, 3243, 3244, 3245, 3246, 3247, 3248, 3249, 3250, 3251, 3252, 3253,\r\n 3254, 3255, 3256, 3257, 3258, 3259, 3260, 3261, 3262, 3263, 3264, 3265,\r\n 3266, 3267, 3268, 3269, 3270, 3271, 3272, 3273, 3274, 3275, 3276, 3277,\r\n 3278, 3279, 3280, 3281, 3282, 3283, 3284, 3285, 3286, 3287, 3288, 3289,\r\n 3290, 3291, 3292, 3293, 3294, 3295, 3296, 3297, 3298, 3299, 3300, 3301,\r\n 3302, 3303, 3304, 3305, 3306, 3307, 3308, 3309, 3310, 3311, 3312, 3313,\r\n 3314, 3315, 3316, 3317, 3318, 3319, 3320, 3321, 3322, 3323, 3324, 3325,\r\n 3326, 3327, 3328, 3329, 3330, 3331, 3332, 3333, 3334, 3335, 3336, 3337,\r\n 3338, 3339, 3340, 3341, 3342, 3343, 3344, 3345, 3346, 3347, 3348, 3349,\r\n 3350, 3351, 3352, 3353, 3354, 3355, 3356, 3357, 3358, 3359, 3360, 3361,\r\n 3362, 3363, 3364, 3365, 3366, 3367, 3368, 3369, 3370, 3371, 3372, 3373,\r\n 3374, 3375, 3376, 3377, 3378, 3379, 3380, 3381, 3382, 3383, 3384, 3385,\r\n 3386, 3387, 3388, 3389, 3390, 3391, 3392, 3393, 3394, 3395, 3396, 3397,\r\n 3398, 3399, 3400, 3401, 3402, 3403, 3404, 3405, 3406, 3407, 3408, 3409,\r\n 3410, 3411, 3412, 3413, 3414, 3415, 3416, 3417, 3418, 3419, 3420, 3421,\r\n 3422, 3423, 3424, 3425, 3426, 3427, 3428, 3429, 3430, 3431, 3432, 3433,\r\n 3434, 3435, 3436, 3437, 3438, 3439, 3440, 3441, 3442, 3443, 3444, 3445,\r\n 3446, 3447, 3448, 3449, 3450, 3451, 3452, 3453, 3454, 3455, 3456, 3457,\r\n 3458, 3459, 3460, 3461, 3462, 3463, 3464, 3465, 3466, 3467, 3468, 3469,\r\n 3470, 3471, 3472, 3473, 3474, 3475, 3476, 3477, 3478, 3479, 3480, 3481,\r\n 3482, 3483, 3484, 3485, 3486, 3487, 3488, 3489, 3490, 3491, 3492, 3493,\r\n 3494, 3495, 3496, 3497, 3498, 3499, 3500, 3501, 3502, 3503, 3504, 3505,\r\n 3506, 3507, 3508, 3509, 3510, 3511, 3512, 3513, 3514, 3515, 3516, 3517,\r\n 3518, 3519, 3520, 3521, 3522, 3523, 3524, 3525, 3526, 3527, 3528, 3529,\r\n 3530, 3531, 3532, 3533, 3534, 3535, 3536, 3537, 3538, 3539, 3540, 3541,\r\n 3542, 3543, 3544, 3545, 3546, 3547, 3548, 3549, 3550, 3551, 3552, 3553,\r\n 3554, 3555, 3556, 3557, 3558, 3559, 3560, 3561, 3562, 3563, 3564, 3565,\r\n 3566, 3567, 3568, 3569, 3570, 3571, 3572, 3573, 3574, 3575, 3576, 3577,\r\n 3578, 3579, 3580, 3581, 3582, 3583, 3584, 3585, 3586, 3587, 3588, 3589,\r\n 3590, 3591, 3592, 3593, 3594, 3595, 3596, 3597, 3598, 3599, 3600, 3601,\r\n 3602, 3603, 3604, 3605, 3606, 3607, 3608, 3609, 3610, 3611, 3612, 3613,\r\n 3614, 3615, 3616, 3617, 3618, 3619, 3620, 3621, 3622, 3623, 3624, 3625,\r\n 3626, 3627, 3628, 3629, 3630, 3631, 3632, 3633, 3634, 3635, 3636, 3637,\r\n 3638, 3639, 3640, 3641, 3642, 3643, 3644, 3645, 3646, 3647, 3648, 3649,\r\n 3650, 3651, 3652, 3653, 3654, 3655, 3656, 3657, 3658, 3659, 3660, 3661,\r\n 3662, 3663, 3664, 3665, 3666, 3667, 3668, 3669, 3670, 3671, 3672, 3673,\r\n 3674, 3675, 3676, 3677, 3678, 3679, 3680, 3681, 3682, 3683, 3684, 3685,\r\n 3686, 3687, 3688, 3689, 3690, 3691, 3692, 3693, 3694, 3695, 3696, 3697,\r\n 3698, 3699, 3700, 3701, 3702, 3703, 3704, 3705, 3706, 3707, 3708, 3709,\r\n 3710, 3711, 3712, 3713, 3714, 3715, 3716, 3717, 3718, 3719, 3720, 3721,\r\n 3722, 3723, 3724, 3725, 3726, 3727, 3728, 3729, 3730, 3731, 3732, 3733,\r\n 3734, 3735, 3736, 3737, 3738, 3739, 3740, 3741, 3742, 3743, 3744, 3745,\r\n 3746, 3747, 3748, 3749, 3750, 3751, 3752, 3753, 3754, 3755, 3756, 3757,\r\n 3758, 3759, 3760, 3761, 3762, 3763, 3764, 3765, 3766, 3767, 3768, 3769,\r\n 3770, 3771, 3772, 3773, 3774, 3775, 3776, 3777, 3778, 3779, 3780, 3781,\r\n 3782, 3783, 3784, 3785, 3786, 3787, 3788, 3789, 3790, 3791, 3792, 3793,\r\n 3794, 3795, 3796, 3797, 3798, 3799, 3800, 3801, 3802, 3803, 3804, 3805,\r\n 3806, 3807, 3808, 3809, 3810, 3811, 3812, 3813, 3814, 3815, 3816, 3817,\r\n 3818, 3819, 3820, 3821, 3822, 3823, 3824, 3825, 3826, 3827, 3828, 3829,\r\n 3830, 3831, 3832, 3833, 3834, 3835, 3836, 3837, 3838, 3839, 3840, 3841,\r\n 3842, 3843, 3844, 3845, 3846, 3847, 3848, 3849, 3850, 3851, 3852, 3853,\r\n 3854, 3855, 3856, 3857, 3858, 3859, 3860, 3861, 3862, 3863, 3864, 3865,\r\n 3866, 3867, 3868, 3869, 3870, 3871, 3872, 3873, 3874, 3875, 3876, 3877,\r\n 3878, 3879, 3880, 3881, 3882, 3883, 3884, 3885, 3886, 3887, 3888, 3889,\r\n 3890, 3891, 3892, 3893, 3894, 3895, 3896, 3897, 3898, 3899, 3900, 3901,\r\n 3902, 3903, 3904, 3905, 3906, 3907, 3908, 3909, 3910, 3911, 3912, 3913,\r\n 3914, 3915, 3916, 3917, 3918, 3919, 3920, 3921, 3922, 3923, 3924, 3925,\r\n 3926, 3927, 3928, 3929, 3930, 3931, 3932, 3933, 3934, 3935, 3936, 3937,\r\n 3938, 3939, 3940, 3941, 3942, 3943, 3944, 3945, 3946, 3947, 3948, 3949,\r\n 3950, 3951, 3952, 3953, 3954, 3955, 3956, 3957, 3958, 3959, 3960, 3961,\r\n 3962, 3963, 3964, 3965, 3966, 3967, 3968, 3969, 3970, 3971, 3972, 3973,\r\n 3974, 3975, 3976, 3977, 3978, 3979, 3980, 3981, 3982, 3983, 3984, 3985,\r\n 3986, 3987, 3988, 3989, 3990, 3991, 3992, 3993, 3994, 3995, 3996, 3997,\r\n 3998, 3999, 4000, 4001, 4002, 4003, 4004, 4005, 4006, 4007, 4008, 4009,\r\n 4010, 4011, 4012, 4013, 4014, 4015, 4016, 4017, 4018, 4019, 4020, 4021,\r\n 4022, 4023, 4024, 4025, 4026, 4027, 4028, 4029, 4030, 4031, 4032, 4033,\r\n 4034, 4035, 4036, 4037, 4038, 4039, 4040, 4041, 4042, 4043, 4044, 4045,\r\n 4046, 4047, 4048, 4049, 4050, 4051, 4052, 4053, 4054, 4055, 4056, 4057,\r\n 4058, 4059, 4060, 4061, 4062, 4063, 4064, 4065, 4066, 4067, 4068, 4069,\r\n 4070, 4071, 4072, 4073, 4074, 4075, 4076, 4077, 4078, 4079, 4080, 4081,\r\n 4082, 4083, 4084, 4085, 4086, 4087, 4088, 4089, 4090, 4091, 4092, 4093,\r\n 4094, 4095, 4096, 4097, 4098, 4099, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1,\r\n 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1, 1]]) torch.Size([2, 4608])\r\n```\r\n\r\nWell, the model pads the sequence from 4098 to 4608 (because it's a multiple of 512, though I did not tell it to do so), and then the ```position_ids``` it ended up with starts with 1 instead of 0 and then there are also 4098 and 4099 in it, which I believe should be out of index? Needs confirmation from developers! Thx!", "@LysandreJik I am pretty sure that ```create_position_ids_from_input_ids()``` in modeling_longformer.py can generate position_ids that are larger than 4097 and lead to out-of-index problem.", "Hey @PxYu - yeah the reason is that even though `Longformer` has a `max_position_embeddings` of 4098 in its official config: https://huggingface.co/allenai/longformer-base-4096, the maximum input it can handle is only `4096` or otherwise the `create_position_ids` function will crash. \r\n\r\nThis problem is analogues for `Roberta` as discussed here: https://github.com/huggingface/transformers/pull/8044#issuecomment-716513140 .\r\n\r\n@LysandreJik - I think it's wrong that Roberta and Longformer (as it was built on Roberta) have `max_position_embeddings` of 514 and 4098 respectively. I think we should definitely change the default `max_position_embeddings` in `RobertaConfig` and `LongformerConfig` and even if it breaks backward compatibility, I would advocate for changing the parameter in the configs of the main models as well. Someone using a `input_ids` of length `max_position_embeddings = 514 or = 4098` would probably have led to errors anyways IMO. These issues will probably happen more often otherwise. What do you think? ", "As seen offline, having the `max_position_embeddings` set to 514 is an unfortunate decision we made when implementing the RoBERTa model and inheriting from the BERT model, a behavior we've since changed.\r\n\r\nUnfortunately, changing the `max_position_embeddings` would be impossible now, as this would imply modifying the model code so that it handles the embeddings to be of size `max_position_embeddings + 2`, which would break all current existing models. We could reach for an if/else statement analyzing `transformers` version, but this would be very error prone, and would needlessly complicate the code.\r\n\r\nWe should document two things:\r\n- The `max_position_embeddings` should not be used to create sequences, the tokenizer's `max_len` should be used instead. Since models and tokenizers work in pairs, it is not incoherent to have to rely on the tokenizer attribute to create model sequences.\r\n- We should document that models should try to respect, as much as possible, the convention that `model.max_position_embeddings == tokenizer.max_len` to prevent such confusion from happening in newer models.", "Agree! I think we could however also change the default value of `max_position_embeddings` of `LongformerConfig` and `RobertaConfig` to 4096 and 512 - this should not break anything as people load their saved config via `from_pretrained(...)`.\r\n\r\nThe big advantage of doing so would be to somewhat prevent future models that built on top of Roberta/Longformer from having these wrong numbers.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "Guys, since this is not code but a bunch of definitions please make direct suggestions that can be merged. Thank you. \r\n\r\nI originally just wanted to add `docs` and `flax` to `dev` and had no idea about all the other needs.", "I can't make a suggestion that moves stuff around in the file. I can push a commit to your branch if you want, but that's the best I can do.", "I understand. I just don't know to work in this fashion. If you give me a spec I can work to implement it. Otherwise let's just merge this and subsequent PRs can do further improvements. ", "Sure, I'll add my comments in a separate PR. Thanks for doing this one!\r\n" ]

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

[]

[ "README.md file added for gurkan08/bert-turkish-text-classification model" ]

[]

[]

[]

[]

[]

[ "On hold for now until we update the input/output of the alignement methods in `tokenizers` to handle pairs of input sentences following internal discussion with @n1t0 and @Narsil.", "Remaining error on the CI (NER pipeline not working for slow tokenizers) should be solved by #8364 \r\n\r\nEdit: ok solved now that #8364 is merged", "examples/seq2seq/finetune.py is failing after this PR:\r\n\r\n```\r\nCUDA_VISIBLE_DEVICES=0 pytest -sv examples/seq2seq/test_seq2seq_examples.py::TestTheRest::test_finetune_0_patrickvonplaten_t5_tiny_random\r\n```\r\n\r\n\r\n```\r\nTraceback (most recent call last):\r\n File \"finetune.py\", line 442, in <module>\r\n main(args)\r\n File \"finetune.py\", line 409, in main\r\n trainer: pl.Trainer = generic_train(\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/examples/lightning_base.py\", line 398, in generic_train\r\n trainer.fit(model)\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py\", line 444, in fit\r\n results = self.accelerator_backend.train()\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/accelerators/gpu_accelerator.py\", line 63, in train\r\n results = self.train_or_test()\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/accelerators/accelerator.py\", line 74, in train_or_test\r\n results = self.trainer.train()\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/trainer/trainer.py\", line 493, in train\r\n self.train_loop.run_training_epoch()\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/trainer/training_loop.py\", line 554, in run_training_epoch\r\n for batch_idx, (batch, is_last_batch) in train_dataloader:\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/profiler/profilers.py\", line 80, in profile_iterable\r\n value = next(iterator)\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/pytorch_lightning/trainer/connectors/data_connector.py\", line 46, in _with_is_last\r\n last = next(it)\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/utils/data/dataloader.py\", line 519, in __next__\r\n data = self._next_data()\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/utils/data/dataloader.py\", line 1169, in _next_data\r\n return self._process_data(data)\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/utils/data/dataloader.py\", line 1195, in _process_data\r\n data.reraise()\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/_utils.py\", line 428, in reraise\r\n raise self.exc_type(msg)\r\nAttributeError: Caught AttributeError in DataLoader worker process 0.\r\nOriginal Traceback (most recent call last):\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/utils/data/_utils/worker.py\", line 202, in _worker_loop\r\n data = fetcher.fetch(index)\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/torch/utils/data/_utils/fetch.py\", line 47, in fetch\r\n return self.collate_fn(data)\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/examples/seq2seq/utils.py\", line 251, in collate_fn\r\n batch_encoding: Dict[str, torch.Tensor] = self.tokenizer.prepare_seq2seq_batch(\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/models/bart/tokenization_bart_fast.py\", line 127, in prepare_seq2seq_batch\r\n model_inputs: BatchEncoding = self(\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/tokenization_utils_base.py\", line 2319, in __call__\r\n return self.batch_encode_plus(\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/tokenization_utils_base.py\", line 2504, in batch_encode_plus\r\n return self._batch_encode_plus(\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/models/gpt2/tokenization_gpt2_fast.py\", line 167, in _batch_encode_plus\r\n return super()._batch_encode_plus(*args, **kwargs)\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/tokenization_utils_fast.py\", line 433, in _batch_encode_plus\r\n return BatchEncoding(sanitized_tokens, sanitized_encodings, tensor_type=return_tensors)\r\n File \"/mnt/nvme1/code/huggingface/transformers-master/src/transformers/tokenization_utils_base.py\", line 242, in __init__\r\n n_sequences = encoding[0].n_sequences\r\nAttributeError: 'tokenizers.Encoding' object has no attribute 'n_sequences'\r\n\r\nException ignored in: <function tqdm.__del__ at 0x7f60b55d7b80>\r\nTraceback (most recent call last):\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/tqdm/std.py\", line 1128, in __del__\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/tqdm/std.py\", line 1341, in close\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/tqdm/std.py\", line 1520, in display\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/tqdm/std.py\", line 1131, in __repr__\r\n File \"/home/stas/anaconda3/envs/main-38/lib/python3.8/site-packages/tqdm/std.py\", line 1481, in format_dict\r\nTypeError: cannot unpack non-iterable NoneType object\r\n```", "Hi @stas00, it seems you don't have the latest tokenizer version!", "Bummer! Thank you for identifying that, @LysandreJik - I confirm updating `tokenizers` fixes it.\r\n\r\nSince this happened more than once now, would it be possible to add and maintain run-time checks as we have done here:\r\n\r\nhttps://github.com/huggingface/transformers/blob/dd52804f5fce0a568ffbb3dc7fd088d2de0a0e56/examples/lightning_base.py#L38-L47\r\n\r\nHow a developer is to know that they need to update a dependency of a project otherwise? `git pull` doesn't trigger `pip install -e \".[dev]\"`\r\n\r\nExcept the code above should emit an error on failed check. It should be an error in `lightning_base.py` too, but I couldn't convince @sshleifer to make it so. Warnings are plentiful and this approach of using a warning just doesn't serve its purpose, IMHO.\r\n", "I like the function (though we can expand its name and replace `ver` by `version` inside to make it more readable ;-) ) and I think we could have a dynamic check at init. I don't see any problem with throwing an error if the minimum version of tokenizers isn't installed, but maybe @LysandreJik or @patrickvonplaten have a different opinion.", "Oh, any name works, it was just code first, then I thought that it'll eventually end up being a reusable function, so the half-baked function emerged.\r\n\r\nI'm glad you agree that it should assert if the minimal requirements aren't met.\r\n\r\nTechnically we should then add `packaging` and `pkg_resources` to the project dependencies, but since they are prerequisites for `setuptools` - every user should already have them.\r\n\r\nAnd we will need 2 versions of errors: \r\n* `Try: pip install -r examples/requirements.txt` for examples, \r\n* `Try: pip install transformers -U\"` for the core.", "Agree! I actually also already ran into the same issue @stas00 -> I'm definitely in favor of such a function! Should we run the function at `import transformers` already ? ", "So did I, and I thought everything was broken until @sgugger showed me the way. Your function would definitely help in that regard @stas00 :)", "> Should we run the function at `import transformers` already ?\r\n\r\nTwo ways I can see:\r\n\r\n1. right where the corresponding import is done - so it's easier to see the specific requirement - but it could be scattered - but it could be more difficult to maintain. Ideally, python would have `import module 1.2.3`, like some other languages have, but it doesn't at the moment.\r\n2. all in one place, at the very top of `__init__.py`, all requirements next to each other so it's easy to maintain. This check relies on packaging tools - i.e. derives it from the `site-packages` dir, so it shouldn't even load the module. i.e. we don't try to look for `xxx.__version__` here, since not all packages have it.\r\n\r\nI'd say (2) is the easier way.\r\n\r\nLast night I was dreaming of a trigger feature, where if git sees `setup.py` modified it'd alert someone to update `__init__.py` requirements - but it was a dream.\r\n", "I think it can be in any file imported by the `__init__` (as long as the function is executed), so we could also have this in `file_utils.py`. Though the `__init__` is fine by me too if you think it's better.", "Sure, then let's add a dedicated module then? It'd be the most simple/intuitive then\r\n\r\n```\r\n$ cat version_requirements.py\r\nrequire(\"tokenizers\", \"1.2.3\")\r\n...\r\n```\r\nand `__init__.py`:\r\n```\r\nimport .version_requirements\r\n```", "hear, hear! if we have such a file then we can even use it to feed `setup.py`! so we have a single place where we edit all the minimal version requirements and don't need to touch `setup.py` and potentially forget to sync requirements.\r\n\r\nIn which case I'd use a dict and feed it to `require_version` (or whatever we end up calling it).\r\n\r\nClearly setup has a lot of optional things, so perhaps then we load this file at the end of __init__ and only check versions for the things that got loaded?\r\n\r\nor we just test only the package names that we know we need to check, but use that dict for setup.py's needs.\r\n\r\nLet me know if these ideas are an overkill.", "Here is a quick prototype to what I'm thinking:\r\n```\r\n$ cat src/transformers/version_requirements.py\r\nmin_vers = dict(\r\n tokenizers: \"==0.9.4\",\r\n tqdm: \">=4.27\",\r\n jaxlib: \"==0.1.55\",\r\n)\r\nrun_time_keys = \"tokenizers tqdm\".split()\r\nfor k in run_time_keys:\r\n require_min_ver(k, min_vers[k])\r\n\r\n$ cat setup.py\r\nfrom version_requirements import min_vers\r\n# of course we won't hardcode each entry - this is a just to demonstrate\r\nextras[\"flax\"] = [f\"jaxlib{min_vers{'jax_lib']}\", ...\r\n\r\n```\r\n\r\nso you can see the dictionary has all the versions, but we actively check only the versions that are non-optional.\r\n", "One downside of this is that it would move dependencies out of the setup.py (which is where people would expect to see them). Do you think there is a way to structure this so the one place we look at minimum version is the setup? It would be less surprising I think. ", "I agree.\r\n\r\nWe could have all the version requirements defined in `setup.py` and when it's run it'd update `src/transformers/version_requirements.py` instead. Then we would actually want 2 files under transformers - one that `setup.py` will maintain - it will be just a dict dump - so that it could overwrite the file completely and another for the selective run-time checks that would refer to the first file generated by setup, since we will only check a handful of these many dependencies at run time.\r\n\r\n```\r\n$ cat setup.py\r\nmin_vers = dict(\r\n tokenizers: \"==0.9.4\",\r\n tqdm: \">=4.27\",\r\n jaxlib: \"==0.1.55\",\r\n)\r\n\r\n# add code to dump min_vers dict into `src/transformers/version_requirements.py`\r\n\r\n# of course we won't hardcode each entry - this is a just to demonstrate\r\nextras[\"flax\"] = [f\"jaxlib{min_vers{'jax_lib']}\", ...\r\n\r\n\r\n$ cat src/transformers/version_requirements.py\r\n# AUTOGENERATED - MODIFY setup.py INSTEAD! #\r\nmin_vers = dict(\r\n tokenizers: \"==0.9.4\",\r\n tqdm: \">=4.27\",\r\n jaxlib: \"==0.1.55\",\r\n)\r\n\r\n$ cat src/transformers/version_run_time_check.py\r\nfrom .version_requirements import min_vers\r\n\r\n# define which module versions we always want to check (only a few)\r\nrun_time_keys = \"tokenizers tqdm\".split()\r\nfor k in run_time_keys:\r\n require_min_ver(k, min_vers[k])\r\n\r\n$ cat src/transformers/__init__.py\r\nimport .version_run_time_check\r\n```\r\n\r\nThis is of course all just a visual prototype.\r\n", "If you want to tackle this, please go ahead with something along these lines. We can refine more on an actual PR.", "OK, I will make a partial sub-set of modules and when you like how it looks expand it to all modules.", "Is there any way to workaround the version check?\r\nI want to use some features from `tokenizers 0.10`, but Transformers raise `VersionConflict`.\r\n\r\nSurely, this can cause some very non-obvious bugs, but at least I'll be able to work with my code before the new version of Transformers is released.", "I think this is a very reasonable need, @Guitaricet. But it's probably best to discuss it in a dedicated issue. Could you please file a [feature request](https://github.com/huggingface/transformers/issues/new/choose) and let's see what others would think?\r\n\r\nI'd say an env var to override the checks should do the trick. Should be easy to add if the others agree with having it." ]

[]

[ { "login": "sshleifer", "id": 6045025, "node_id": "MDQ6VXNlcjYwNDUwMjU=", "avatar_url": "https://avatars.githubusercontent.com/u/6045025?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sshleifer", "html_url": "https://github.com/sshleifer", "followers_url": "https://api.github.com/users/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "I am pretty sure my guess at what happened was correct.\r\n\r\nOn one of my machines, I have the bart-large/config.json from October 12 with etag \"40bd49bcec9d93d8b0bfbd020088e2e1b6e6bb03e8e80aa5144638f90ca6bd61\" and it is 1.26kb. It contains an entry `\"output_past\": false`. Today, there is a file with a new etag \"8b65d3b9a47e96c1909d807f7e7f41dd1ed95092b139965be7b914aa4fb5fd08\" and it is 1.52kb. It does not contain any `output_past` entry.\r\n\r\nWhat could I have done to prevent this from happening? Is there a way to specific a specific version of the models?", "I can confirm your diagnostic is correct.\r\n\r\nYou can check the diff between the two versions with:\r\n```\r\ncurl -i https://s3.amazonaws.com/models.huggingface.co/bert/facebook/bart-large/config.json?versionId=PFecmBwmg83YUwpv_kkc3kBzoCGebvu7\r\ncurl -i https://s3.amazonaws.com/models.huggingface.co/bert/facebook/bart-large/config.json?versionId=JhIFsOvvLtrLn0vJjGNN6ZhJGUlbXBEP\r\n```\r\n\r\nI'll let @sshleifer and @patrickvonplaten chime in about the actual file change, but to answer your last question:\r\n\r\n> What could I have done to prevent this from happening? Is there a way to specific a specific version of the models?\r\n\r\nWe will roll out a way to specify specific versions of models in the near future.", "I mistakenly changed the `config`, my fault.\r\nCan you pass `output_past` to `__init__` or do you need me to add back the `output_past` key?", "I was able to just remove the flag and it works with the updated config. To be honest, I don't know what the flag did -- I modified someone else's model which used it.\r\n\r\nThanks for looking into this. This issue can be closed" ]

[]

[]

[ "Hey @ysgit - I like the idea! However, I think we should actually create a new output tuple for the cross-attention (called `cross_attention` when `return_dict`=True). For this we will have to create new ModelOutputs for Bert, RoBERTa, etc...\r\n", "@patrickvonplaten that makes sense, I guess I suspected that might be the reaction, I'll see if that's something I can manage although I'm a little hampered by not being able to get the test suite to successfully run locally", "@patrickvonplaten I have done as you suggested and separated cross attentions into a new variable in the output. please take a look and let me know what you think. many thanks!", "In a future PR or a \"Good First Issue\", we could add this functionality to the TFSeq2Seq models as well, but I'm a bit hesitant given the current problems with the `output_attentions` flag in TF.\r\nWhat do you think @sgugger @LysandreJik @jplu ?", "This should be easier to integrate in TF in the next release :)", "Thanks everyone!" ]

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

[]

[ "@dharakotecha, you would need to create a MaskedLM head similar to distilbert. You would also need to either use one of the masekd token <extra_id_0> or add a new token for mask. \r\n\r\nSome experimentation on the t5 encoder shows that it does produce pretty reasonable output, but does not do MLM at the encoder side. The cloze/fill in the blank happens in the decoder side.\r\n\r\nThis is just quick and dirty. You would need to create a module similar to a MLM for distilbert with the loss, feeding in the output from the encoder into the projector:\r\n\r\n```\r\nt5 = AutoModel.from_pretrained('t5-small').cuda()\r\ntokenizer = AutoTokenizer.from_pretrained('t5-small')\r\nencoder = t5.encoder\r\nembeddings = encoder.embed_tokens\r\nconfig = t5.config\r\n#tied the embedding to the projector. \r\nvocab_projector = nn.Linear(config.d_model, config.vocab_size).cuda()\r\nvocab_projector.weight = embeddings.weightvocab_projector = nn.Linear(config.d_model, config.vocab_size).cuda()\r\nvocab_projector.bias.data = torch.nn.functional.pad(vocab_projector.bias.data, (0, vocab_projector.weight.shape[0] - vocab_projector.bias.shape[0],), \"constant\", 0)\r\n\r\ndef pred(predictions):\r\n for pred in predictions:\r\n print (\"**\")\r\n sorted_preds, sorted_idx = pred.sort(dim=-1, descending=True)\r\n ret = []\r\n for k in range(2):\r\n predicted_index = [sorted_idx[i, k].item() for i in range(0,len(predictions[0]))]\r\n predicted_token = ' '.join([tokenizer.convert_ids_to_tokens([predicted_index[x]])[0] for x in range(1,len(predictions[0]))]).replace('Ġ', ' ').replace(' ', ' ').replace('##', '')\r\n ret.append(predicted_token)\r\n return ret\r\n\r\n\r\ninput_txt = [\"</s>Lincoln was an American president and lawyer\"]\r\ninputs = tokenizer(input_txt, return_tensors='pt', add_special_tokens=True, padding=True)\r\npredictions = vocab_projector(encoder(inputs.input_ids.cuda())[0])\r\npred(predictions)\r\n\r\n\r\n```\r\nOutputs:\r\n['\\u2581Lincoln \\u2581was \\u2581psiho American \\u2581President & \\u2581lawyer </s>', '\\u2581senzati tais Mitglied \\u2581American \\u2581president \\u2581and \\u2581Lawyer \\u2581summarize']**\r\n\r\nBut, using the extra mask, will not infer the missing token.\r\n\r\n```\r\ninput_txt = [\"</s>Lincoln <extra_id_0> American president and lawyer\"]\r\ninputs = tokenizer(input_txt, return_tensors='pt', add_special_tokens=True, padding=True)\r\npredictions = vocab_projector(encoder(inputs.input_ids.cuda())[0])\r\npred(predictions)\r\n\r\n```\r\nWil output:\r\n['\\u2581Lincoln <extra_id_0> \\u2581American \\u2581president & \\u2581lawyer </s>', '\\u2581Abraham \\u2581botez American \\u2581President \\u2581and \\u2581Lawyer gasesc']**\r\n\r\nHope this helps,\r\n\r\n", "I just realized that you asking about tensor flow too... I'm assuming you could do the similar thing but creating a MLM head in tesnor flow just for the encoder.", "This issue has been automatically marked as stale and been closed because it has not had recent activity. Thank you for your contributions.\n\nIf you think this still needs to be addressed please comment on this thread." ]

[]

[]

[ "I can check the other 3 boxes if needed. They didn't seem to apply to this particular PR so I left them unchecked.\r\n" ]

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

[ { "login": "sshleifer", "id": 6045025, "node_id": "MDQ6VXNlcjYwNDUwMjU=", "avatar_url": "https://avatars.githubusercontent.com/u/6045025?v=4", "gravatar_id": "", "url": "https://api.github.com/users/sshleifer", "html_url": "https://github.com/sshleifer", "followers_url": "https://api.github.com/users/sshleifer/followers", "following_url": "https://api.github.com/users/sshleifer/following{/other_user}", "gists_url": "https://api.github.com/users/sshleifer/gists{/gist_id}", "starred_url": "https://api.github.com/users/sshleifer/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/sshleifer/subscriptions", "organizations_url": "https://api.github.com/users/sshleifer/orgs", "repos_url": "https://api.github.com/users/sshleifer/repos", "events_url": "https://api.github.com/users/sshleifer/events{/privacy}", "received_events_url": "https://api.github.com/users/sshleifer/received_events", "type": "User", "site_admin": false } ]

[ "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "A few loose-ends to tie, but that will be for tomorrow!" ]

[]

[]

[ "Remember that the examples you pull from master need a [source install](https://huggingface.co/transformers/installation.html#installing-from-source). If you want the version that runs with the last release, you need to use that tag, here are the [examples for the last release](https://github.com/huggingface/transformers/releases/tag/v3.4.0) (v3.4.0).", "You make a compelling argument :)\r\nI did a checkout of the v3.4.0 release, and besides dying from a CUDA out of memory, it appears to be working :)" ]

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

[]

[ "Hello\r\nSame problem for us.\r\nDo you plan to investigate / fix it ?\r\nCheers\r\nPhilippe", "The UniLM model has not been released in the library yet.", "This issue has been automatically marked as stale because it has not had recent activity. It will be closed if no further activity occurs. Thank you for your contributions.\n" ]

[]

[]

[ "Didn't work. Will re-open when I have something better." ]

[]

[]

[]

[]

[]

[ "Code changes pass tests @sgugger ! Thanks @alvarobartt for having a look at those also.\r\n\r\nI guess I need to take care of docs/source/main_classes/callback.rst as well to complete the checklist though?", "Yes, if you could add a line to the rst file, that would be great!", "You might need to rebase to have the latest script for doc formatting, which should do everything to make the CI happy with just `make style`. Let me know if you need help.", "Sorry for the somewhat messy hacktoberfest @sgugger ! \r\nI wasn't so aware of rst idiosyncrasies, and ways to diagnose issues, so I struggled a bit. Should be in order now (bonus, I fixed a typo on the way https://github.com/huggingface/transformers/pull/8062/commits/286f20c0594c3d16c824963c24c8fb1bc1d43bc6)" ]

[]

[]

[]

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

[]

[ "?", "This issue has been automatically marked as stale and been closed because it has not had recent activity. Thank you for your contributions.\n\nIf you think this still needs to be addressed please comment on this thread." ]

[]

[]

[ "Wouldn't it be better to:\r\n- ask the model authors for the relevant models to update their config.json to include a `id2label`\r\n- and/or to modify them automatically for them on the hub?\r\n\r\nI feel this PR is adding a feature that works around an issue when we should be fixing the root issue (also fixes the displayed labels, other future features, etc.). Wdyt?", "@julien-c Well, to be clear this PR ~~does~~ did two things:\r\n\r\n1. Switches from always using the last index to using the index determined by looking in the model config if present\r\n2. Gives the user the option to manually override the index in case the information isn't present in the config\r\n\r\nIf I understand correctly, your issue is just with (2). I think it's a fair point. I don't think we'll be able to ensure that [all NLI models](https://huggingface.co/models?search=nli) have a clearly defined label mapping though. But instead of an override arg, I think it might be better to just add a warning if the entailment label ID can't be found in the config.", "@joeddav Ok yes 1/ is great.\r\n\r\n> I don't think we'll be able to ensure that [all NLI models](https://huggingface.co/models?search=nli) have a clearly defined label mapping though.\r\n\r\nWhy not?", "@julien-c Just because there are almost 100 results for [\"NLI\"](https://huggingface.co/models?search=nli) on the model hub and I'd guess from a quick sampling that the majority don't have a label mapping defined. For each model we'd have to figure out which label is which, which would mean either getting the author to look it up and tell us or else running tests on the correct dataset to figure it ourselves.\r\n\r\nDo you think it'd be worthwhile to warn the user when uploading or creating configs with generic/missing label mappings (or with any other important fields missing) going forward? Defining a label2id seems like a rather obscure property that I would assume is purely cosmetic if I were uploading a model, i.e. I wouldn't expect it to actually impact code behavior for someone using my model." ]

[]

[]

[ "Ooops, closed it by mistake too early. But it has been resolved here:\r\nhttps://github.com/huggingface/transformers/pull/8107" ]

[]

[]

[ "p.s. apparently there has been a deprecation warning with pytorch-1.6, but since this project isn't quite paying attention to the warnings it lead to this crash with pytorch-1.7. It seems to be a weird situation where a deprecation warning hasn't been turned into an error and instead leading to a crash with this particular issue, but perhaps setting an intention to keep the warnings in check would save a possible hassle in the future. https://github.com/huggingface/transformers/issues/8060\r\n\r\nPerhaps what would help is to automatically turn selective types of warnings to errors, and thus not let those slide until they become a problem.\r\n\r\nAlso having a scheduled CI that runs occasionally on pytorch-nightly (and any release candidates) would give an early alert." ]

[]

[]

[]