Datasets:

---

DanielPFlorian

/

Transformers-Github-Issues

like 0

[]

[]

[]

[]

[]

[]

[]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "Hi @7AM7 \r\n\r\nI think this is because there is `_step` in TF checkpoint, which should be ignored when loading the weights.\r\n\r\nfor this you should write your own conversion script. You could take and modify this function\r\nhttps://github.com/huggingface/transformers/blob/30f065890e77f2917895b175b9a1df503b89e202/src/transformers/models/gpt2/modeling_gpt2.py#L68\r\n\r\nadding some check like this here would solve this\r\n```python\r\nfor name, shape in init_vars:\r\n if \"_step\" not in name name:\r\n```\r\n", "i modified this function def load_tf_weights_in_gpt2 and ignored \"_step\" Like if name!=\"_step\" or if \"_step\" not in name And still same error\r\n\r\n", "in that case, you could check what extra variables are there in the `names` and then remove those from `names` and `arrays`. ", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[]

[]

[]

[ "Did you install TensorFlow in your environment? You might need a more recent TensorFlow version if so.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[]

[]

[]

[]

[]

[]

[ "This PR ist ready for merging from my point of view." ]

[]

[]

[ "fix pr at #11430", "Closed by #11430, thanks @PhilipMay!" ]

[]

[]

[ "Can someone look into it? as I am passing the same data structure as mentioned in above error i n trainer.train(). PFB.\r\n\r\n<_AssertCardinalityDataset shapes: ({input_ids: (None,), attention_mask: (None,), feature_index: (), qas_id: ()}, {start_positions: (), end_positions: (), cls_index: (), p_mask: (None,), is_impossible: ()}), types: ({input_ids: tf.int32, attention_mask: tf.int32, feature_index: tf.int64, qas_id: tf.string}, {start_positions: tf.int64, end_positions: tf.int64, cls_index: tf.int64, p_mask: tf.int32, is_impossible: tf.int32})>", "Hi @PremalMatalia, thank you for opening an issue. Pinging @Rocketknight1 as the TensorFlow developer.\r\n\r\nPlease be aware that we're in the process of deprecating the `TFTrainer` and that we will not be maintaining it anymore as it doesn't offer features that cannot be handled by Keras directly. We're in the process of moving examples to Keras, and @Rocketknight1 has already started with the text classification example [here](https://github.com/huggingface/transformers/blob/master/examples/tensorflow/text-classification/run_text_classification.py).\r\n\r\nQA is sure to follow.", "Thanks @LysandreJik for your response. Good to know that TFTrainer is being deprecated so I can focus on some other ways to fine-tune.\r\n\r\nAny other references to follow at this moment for question answering of SQuAD? ", "Hi! I'll take a look, but the error is quite convoluted. Can you link me to any examples you're following for this? If our code examples aren't working we definitely want to fix that.", "@Rocketknight1 ...I was following below run_tf_squad.py file for fine tuning.\r\n\r\nhttps://github.com/huggingface/transformers/blob/d9c62047a8d75e18d2849d345ab3394875a712ef/examples/question-answering/run_tf_squad.py\r\n", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[]

[]

[]

[ "I think I address most comments (the only thing missing is removing the `from_pt` when the flax checkpoint gets upload)! Thanks for the swift feedback", "Hey @CoderPat,\r\n\r\nWe've merged a last design extension for the Flax design today, [here](https://github.com/huggingface/transformers/commit/f748bd424213ca8e76e6ad9ffe2beece2ff2655e) -> could you merge master into your PR one last time and adapt the code to add those extensions (example docstring + model outputs + all_attentions + all_hidden_states) - super sorry for the merge conflict again, but this will be the last one!" ]

[]

[]

[ "The function prepares the dataset for the standard models as well the XLNet model that requires more arguments. You will need to drop those columns for an Albert models.\r\n\r\nWe are in the process of reworking the TensorFlow examples, so there should be one clearer example for QA soon!", "Thanks @sgugger ...\r\nI tried to modify `squad_convert_examples_to_features` as in attached file to return without these columns but encountered anther error as below:\r\n[transformer_fine_tune_pre_trained_model_tftpu (1).zip](https://github.com/huggingface/transformers/files/6379976/transformer_fine_tune_pre_trained_model_tftpu.1.zip)\r\n\r\n\r\n\r\n\r\n\r\n##ERROR:\r\nValueError Traceback (most recent call last)\r\n<ipython-input-59-d85abec8ae26> in <module>()\r\n 1 # Training\r\n 2 if training_args.do_train:\r\n----> 3 trainer.train()\r\n 4 trainer.save_model()\r\n 5 tokenizer.save_pretrained(training_args.output_dir)\r\n\r\n10 frames\r\n/usr/local/lib/python3.7/dist-packages/tensorflow/python/framework/func_graph.py in wrapper(*args, **kwargs)\r\n 975 except Exception as e: # pylint:disable=broad-except\r\n 976 if hasattr(e, \"ag_error_metadata\"):\r\n--> 977 raise e.ag_error_metadata.to_exception(e)\r\n 978 else:\r\n 979 raise\r\n\r\nValueError: in user code:\r\n\r\n /usr/local/lib/python3.7/dist-packages/transformers/trainer_tf.py:697 distributed_training_steps *\r\n self.args.strategy.run(self.apply_gradients, inputs)\r\n /usr/local/lib/python3.7/dist-packages/transformers/trainer_tf.py:641 apply_gradients *\r\n self.optimizer.apply_gradients(list(zip(gradients, self.model.trainable_variables)))\r\n /usr/local/lib/python3.7/dist-packages/transformers/optimization_tf.py:232 apply_gradients *\r\n return super(AdamWeightDecay, self).apply_gradients(zip(grads, tvars), name=name, **kwargs)\r\n /usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/optimizer_v2/optimizer_v2.py:604 apply_gradients **\r\n self._create_all_weights(var_list)\r\n /usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/optimizer_v2/optimizer_v2.py:783 _create_all_weights\r\n self._create_slots(var_list)\r\n /usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/optimizer_v2/adam.py:127 _create_slots\r\n self.add_slot(var, 'm')\r\n /usr/local/lib/python3.7/dist-packages/tensorflow/python/keras/optimizer_v2/optimizer_v2.py:844 add_slot\r\n .format(strategy, var))\r\n\r\n ValueError: Trying to create optimizer slot variable under the scope for tf.distribute.Strategy (<tensorflow.python.distribute.one_device_strategy.OneDeviceStrategy object at 0x7f12026a45d0>), which is different from the scope used for the original variable (<tf.Variable 'tf_albert_for_question_answering/albert/embeddings/word_embeddings/weight:0' shape=(30000, 128) dtype=float32, numpy=\r\n array([[ 0.01270407, 0.05987824, -0.06812993, ..., -0.01226719,\r\n 0.00817283, 0.00785217],\r\n [ 0.00733008, -0.00101211, -0.01069043, ..., -0.00968418,\r\n -0.0400394 , -0.04233308],\r\n [-0.02059615, 0.007892 , 0.02363562, ..., 0.01533034,\r\n -0.00429517, -0.01246009],\r\n ...,\r\n [ 0.0135935 , 0.00349383, 0.01223597, ..., -0.05456466,\r\n 0.09235671, -0.05717891],\r\n [-0.00492554, -0.05208753, -0.00323149, ..., 0.03003517,\r\n 0.0196551 , 0.06015572],\r\n [ 0.03892251, -0.024089 , -0.01364627, ..., 0.04010094,\r\n 0.05124779, -0.03588157]], dtype=float32)>). Make sure the slot variables are created under the same strategy scope. This may happen if you're restoring from a checkpoint outside the scope", "Can someone please help..? I am stuck here", "I have managed to remove extra tokens from the dataset but then TFTrainer.train() got stuck in infinite loop without any errors or logs.\r\n\r\nBelow is the link with latest code. Please suggest.\r\nhttps://colab.research.google.com/drive/17Rx2rkiqag6YAz_FnU9HyHYtqHJgpNs0?usp=sharing", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "Hi @qute012 \r\n\r\nThe `tie_weights` method ties all the weights of the encoder and decoder including the embedding. For this to work, the encoder and decoder need to be the same model (same class) i.e either BERT2BERT or ROBERTA2ROBERTA2 and with same size.\r\n\r\n> If i want to use different tokenizer between encoder and decoder inputs, does tie function ignore sharing embedding such as strict option?\r\n\r\nno, it does not ignore sharing embedding in that case because as I wrote above it expects both encoder and decoder to be the same model so implicitly assumes that the tokenizer will also be the same.\r\n\r\nBut if that's what you want to do you could manually untie the embeddings or just re-initialize both of them so they won't be shared/tied.", "Thanks to reply @patil-suraj \r\n\r\nFor example, is it right that encoder's embedding weights can be adjusted by decoder's input?\r\n\r\nThen if i want to untie, should i remove or comment out below code manually?\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/encoder_decoder/modeling_encoder_decoder.py#L183\r\n\r\ndo you have any plan to add parameter that choosing tie function in EncoderDecoderModel class? I know bert2bert is better performance than random decoder's embedding weights, but it requires to extending for experiment newly when each encoder and decoder use different vocabulary. If you okay, i will PR.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "It is likely that the IntLayerNorm layers are not warmed up. IntLayerNorm layer has to adjust its internal parameter (`self.shift`) during the quantization-aware training process. It is the one that prevents overflow (i.e. keeps the internal activation values to be less than 2**32) and is initialized with zero. Here is the relevant part in the code: https://github.com/huggingface/transformers/blob/master/src/transformers/models/ibert/quant_modules.py#L508\r\n\r\nTherefore, if you skip the quantization-aware training process and immediately use the model for inference, those layers may produce some unexpected outcomes. Could this be your case?", "Yeah I'm not using quantization-aware training so that'll be the reason. Thanks for the answer!" ]

[]

[]

[ "The setup was updated super late: we already required tensorflow >= 2.3 for a while when we finally went to it. I don't know which version of Transformers supports tensorflow 2.0 but I would guess it's 3.0 or even below.", "Okay, hope the document can be clearer. Every time I read `README.md`, it says `This repository is tested on Python 3.6+, PyTorch 1.0.0+ (PyTorch 1.3.1+ for examples) and TensorFlow 2.0` similar words, I really think it is the specific version 2.0 of tensorflow. However, it is not(tensorflow>=2.3 in `setup.py`). It would be better if the version information is clearer I guess.\r\nHope `transformers` get better and better.", "Yes, this part of the README has not been updated in a while (the PyTorch version is also wrong). Will adjust!", "> Yes, this part of the README has not been updated in a while (the PyTorch version is also wrong). Will adjust!\r\n\r\nThanks 🤗." ]

[]

[]

[ "I believe that is a correct workaround. Would you like to make a PR with it?", "Sure, happy to.", "@chitkwan, are you still inspired to make a PR to fix this? Thank you!", "Oh this has been fixed in #11748 I believe. Sorry I did not reference it in this issue.", "ah yes! \r\n\r\n@chitkwan, could you please validate that the `master` branch with the fix works for you and close this issue if it is so? Thank you!", "Sorry -- this fell off my todo list but thank you for the fix. \r\n\r\nThe original race condition I reported may not be easy to reproduce but I'll give it a go and report back. ", "I reran my failure condition and it no longer fails, so I think this can be closed. Thanks!" ]

[]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "@BigBird01 : could you please help with this?", "I believe this parameter additionally splits the text on the punctuation, as it can be seen from the method it's calling:\r\n\r\nhttps://github.com/huggingface/transformers/blob/afe479adb5474250215438fe27db9dc9dbbbde09/src/transformers/models/deberta_v2/tokenization_deberta_v2.py#L446-L464\r\n\r\nI think this docstring should help out:\r\n\r\nhttps://github.com/huggingface/transformers/blob/afe479adb5474250215438fe27db9dc9dbbbde09/src/transformers/models/deberta_v2/tokenization_deberta_v2.py#L500-L512\r\n\r\nTo see it in practice, you can try it with:\r\n\r\n```py\r\n>>> tok = DebertaV2Tokenizer.from_pretrained(\"microsoft/deberta-v2-xlarge\")\r\n>>> tok._tokenizer._run_split_on_punc(\"Hey, how's he doing?\")\r\n['Hey', ',', ' how', \"'\", 's he doing', '?']\r\n```", "Thanks @LysandreJik for the explanation. Yes, it will split input sentences by punctuation then tokenize the segments by SPM tokenizer. We found this can help the performance of **SQuAD** task, but not on other tasks, e.g. MNLI. So we set it to false by default.\r\n", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

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

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

[ "> Great job @stas00! I like the solution you picked to be able to start initializing part of deepspeed inside the training arguments. Does it fully solve the chicken and egg problem you add?\r\n\r\nThank you!\r\n\r\nFor the current needs of `zero.Init()`, yes! As long as the user separates creating `TrainingArguments` from creating the `Trainer` and calling `from_pretrained` in between, which is what all examples do. It took quite a lot of trial and error, but I think it's pretty clean now. \r\n\r\nSplitting the configuration processing in 2 stages helped a lot too.\r\n\r\nI hope that me using a weak ref global object is a good solution, since w/o it we would somehow have to make the framework aware of deepspeed in multiple places and somehow pass the config to it - most likely by sticking the DS object into the model object. The neat thing is that it being a weak ref it goes away automatically as soon as the TrainerArguments are `gc`'ed. If down the road we discover a better way nothing prevents us from switching to it.\r\n\r\nI will clean up all the XXX's, I was originally planning to wait for 0.3.16 and include all the fixes there, but last time it took more than 10 days for them to make a new release, so I decided it'd be better for users to be able to use this code already, and will make another PR with extra changes next for deepspeed==0.3.16.\r\n", "The weakref is okay by me. The only other way to achieve this would be to have some \"singleton\" class where all instances share the same state, but the weakref is actually more adapted in this case." ]

[]

[]

[ "I found this somehow obscure function argument called `sample` at `AlbertTokenizer`:\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/albert/tokenization_albert.py#L189\r\n\r\nIt seems to enable subword regularization but with fixed parameters for `nbest_size` and `alpha`.\r\n\r\nhttps://github.com/google/sentencepiece/blob/351600c2971401f4e849147579aa1b5d42f614e1/python/src/sentencepiece/__init__.py#L110-L111\r\n\r\nI would remove that `sample` parameter and replace that with my solution which is more flexible. But that would mean we have a breaking change. As an alternative I could add my solution but keep the `sample` argument. But that would add more complexity to the code.\r\n\r\nWhat do you think? @sgugger @LysandreJik @stefan-it \r\n\r\nPS: Same here:\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/bert_generation/tokenization_bert_generation.py#L113\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/big_bird/tokenization_big_bird.py#L143\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/pegasus/tokenization_pegasus.py#L169\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/reformer/tokenization_reformer.py#L109\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/t5/tokenization_t5.py#L237\r\n\r\nhttps://github.com/huggingface/transformers/blob/52166f672ed337934d90cc525c226d93209e0923/src/transformers/models/xlnet/tokenization_xlnet.py#L191", "This argument is not called from anywhere so it's only accessible if users somehow rewrote the tokenize method to pass it along to the private method `_tokenize`. Therefore I think it's fine to do the breaking change and clean up the code using `sample=True`, but let's see what @patrickvonplaten and @LysandreJik think before going forward (note that Lysandre is on vacation until this Wednesday so he'll reply at the end of the week :-) ).", "Yes, removing the `sample` and cleaning up the `_tokenize()` method sounds good to me. As @sgugger said, it is private and nowhere is a `sample` or a `**kwargs` passed to that method.", "> Yes, removing the `sample` and cleaning up the `_tokenize()` method sounds good to me. As @sgugger said, it is private and nowhere is a `sample` or a `**kwargs` passed to that method.\r\n\r\nAgree!", "rebase upstream/master done", "> Yes, LGTM! Thanks a lot.\r\n\r\nHey @LysandreJik - this is not done yet. Please do not merge now. ;-)", "Oh, I was misled! There are indeed a few tokenizers remaining. Thank you for letting me know!", "This is ready to be merged from my point of view.", "Can you take care of the merge conflicts? Will review tomorrow :-)", "> Can you take care of the merge conflicts? Will review tomorrow :-)\r\n\r\n@sgugger All conflicts resolved & green CI\r\n\r\n\r\n", "> Great work on the tests, this is great. The tests could indeed be refactored in a common test if you feel like it.\r\n\r\nI will refactor the tests the next days. Shame on me that I criticized the lack of DRY in the tokenizers but did not follow the DRY principle in the tests.", "This is strange:\r\n\r\n`FAILED tests/test_hf_api.py::HfApiEndpointsTest::test_list_repos_objs - reque...`\r\n\r\nSee here: https://app.circleci.com/pipelines/github/huggingface/transformers/23276/workflows/bf1ad505-efdc-4394-8852-a07702b9f5be/jobs/209965/parallel-runs/0/steps/0-108\r\n\r\nWill trigget CI again,,,", "@LysandreJik @sgugger Tests are refactored and DRY now. CI is green again.\r\nIMO ready for merge.\r\n\r\nMaybe you want to investigate the flaky test (see my comment above)." ]

[]

[]

[ "Hi @seregadgl20-oss \r\n\r\nIt would be nice if you use the forum (https://discuss.huggingface.co/) to ask such general questions. Issues are for bugs and feature requests. Thank you!", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "Hi, the tokenizer of RoBERTa is a byte-level BPE tokenizer. It is a subword tokenizer.\r\n\r\nIn your example with `app ` it completes the word, but there also exists full words within it's dictionary, as you can see when the input sequence is appropriate:\r\n\r\n```py\r\n>>> nlp('Hello, how are you <mask> sir?')\r\n[\r\n {'sequence': 'Hello, how are you doing sir?', 'score': 0.6784416437149048, 'token': 608, 'token_str': ' doing'}, \r\n {'sequence': 'Hello, how are you feeling sir?', 'score': 0.08236288279294968, 'token': 2157, 'token_str': ' feeling'}, \r\n {'sequence': 'Hello, how are you, sir?', 'score': 0.06469670683145523, 'token': 6, 'token_str': ','}, \r\n {'sequence': 'Hello, how are you looking sir?', 'score': 0.04527667537331581, 'token': 546, 'token_str': ' looking'}, \r\n {'sequence': 'Hello, how are you going sir?', 'score': 0.02970985323190689, 'token': 164, 'token_str': ' going'}]\r\n```", "@LysandreJik Thanks for your reply! I think this behavior is proper during the pretraining process: roberta completes or predicts the next token when we give the following input `app<mask>`. But I believe this is not expected when we call `fill-mask` pipeline.\r\n\r\nTake `app <mask>` as example, there is a space in `app <mask>`, which means the `app` token is a complete token and roberta should predict the next token. However, after tokenizing, the space vanished. To realize what I said, maybe the only way is to rewrite the predict function to find those token with a space in front of it.\r\n\r\nP.S. add an extra space `app <mask>` do not bring expected results as well.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[]

[]

[ { "login": "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": "patrickvonplaten", "id": 23423619, "node_id": "MDQ6VXNlcjIzNDIzNjE5", "avatar_url": "https://avatars.githubusercontent.com/u/23423619?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patrickvonplaten", "html_url": "https://github.com/patrickvonplaten", "followers_url": "https://api.github.com/users/patrickvonplaten/followers", "following_url": "https://api.github.com/users/patrickvonplaten/following{/other_user}", "gists_url": "https://api.github.com/users/patrickvonplaten/gists{/gist_id}", "starred_url": "https://api.github.com/users/patrickvonplaten/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patrickvonplaten/subscriptions", "organizations_url": "https://api.github.com/users/patrickvonplaten/orgs", "repos_url": "https://api.github.com/users/patrickvonplaten/repos", "events_url": "https://api.github.com/users/patrickvonplaten/events{/privacy}", "received_events_url": "https://api.github.com/users/patrickvonplaten/received_events", "type": "User", "site_admin": false }, { "login": "patil-suraj", "id": 27137566, "node_id": "MDQ6VXNlcjI3MTM3NTY2", "avatar_url": "https://avatars.githubusercontent.com/u/27137566?v=4", "gravatar_id": "", "url": "https://api.github.com/users/patil-suraj", "html_url": "https://github.com/patil-suraj", "followers_url": "https://api.github.com/users/patil-suraj/followers", "following_url": "https://api.github.com/users/patil-suraj/following{/other_user}", "gists_url": "https://api.github.com/users/patil-suraj/gists{/gist_id}", "starred_url": "https://api.github.com/users/patil-suraj/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/patil-suraj/subscriptions", "organizations_url": "https://api.github.com/users/patil-suraj/orgs", "repos_url": "https://api.github.com/users/patil-suraj/repos", "events_url": "https://api.github.com/users/patil-suraj/events{/privacy}", "received_events_url": "https://api.github.com/users/patil-suraj/received_events", "type": "User", "site_admin": false } ]

[ "I think I could submit a pull request to this, if I had \r\n1) feedback on the idea (do you think it makes sense to do that?)\r\n2) a little help changing existing tests and/or implementing new tests to reflect the change.\r\n\r\nAlso, maybe one would need the new argument to be `Optional[LogitsProcessor]` instead of `Optional[LogitsProcessorList]`. Because `LogitsProcessotList` is a subclass of `LogitsProcessor`, this would allow adding both a list of logits processors and a single logits processor.\r\n\r\nWhat do you folks think? Would you accept this pull request (after maybe giving me some tips related to the tests)? ", "Hey @wadimiusz,\r\n\r\nSorry to only come back to you now! I think in general, I'm fine with such an extension. The only problem I see is that a user could add a custom logits processor that already exists (*e.g.* a user would create his own `LengthPenaltyLogitsProcessor`) and also pass `length_penalty=...` . But even in this case I guess we could just apply both processors and there shouldn't be a big problem. \r\n\r\n=> So I'm ok with this extension. Interested in hearing your thoughts about this @patil-suraj @Narsil", "I think it's a very nice idea !.\r\n\r\nThe problem you mention @patrickvonplaten I think will be relevant mostly for power users (that want to add a LogitsProcessor) so they should be careful in terms of how they use this tool. I guess we could emphasis this in the documentation for the `generate` function, that the simpler arguments are preferred for non advanced usage.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "@wadimiusz Is there any update on this? I think it would be a great addition.", "Hi @ScientiaEtVeritas, the feature seems not hard to implement and I think I already have the code somewhere, but it would require nice and thorough tests that I don't have the time to write right now. If you could help me with the tests, we could submit a pull request together :)", "There used to be a PR that might be used as a starting point:\r\n\r\nhttps://github.com/huggingface/transformers/pull/12219\r\n\r\nThanks if you can work on this ! " ]

[]

[]

[ "Hi @harveenchadha \r\n\r\nif you pass `skip_special_tokens=True` to `decode` method, it will skip all the special tokens. ", "Hi Suraj,\r\n\r\nThanks! That works. But in the inference service that is deployed to test in browser, how will this change?", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "unsubscribe<https://github.com/notifications/unsubscribe-auth/ATPBJ2FNGKBXXXSN6ZH6POLTKIDLZANCNFSM43PQRSGA>.\r\n‏/\r\n0Merged #11410<https://github.com/huggingface/transformers/pull/11410> into master.\r\n\r\n—\r\nYou are receiving this because you are subscribed to this thread.\r\nReply to this email directly, view it on GitHub<https://github.com/huggingface/transformers/pull/11410#event-4639275602>, or unsubscribe<https://github.com/notifications/unsubscribe-auth/ATPBJ2CW5M6UJ4RCYT2YIOLTKIFCTANCNFSM43PQRSGA>.\r\n" ]

[]

[]

[ "Hi @liubest \r\n\r\nPlease make sure that your torch installation can detect the GPU. All scripts will run on GPU if it's available.\r\nIf you want to run on multiple GPUs, follow the docs [here](https://github.com/huggingface/transformers/tree/master/examples/pytorch#distributed-training-and-mixed-precision).\r\n\r\nAnd if you just want to use a single GPU, you could select the device by setting `CUDA_VISIBLE_DEVICES=0\" which will select the first GPU.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[ { "id": 2796628563, "node_id": "MDU6TGFiZWwyNzk2NjI4NTYz", "url": "https://api.github.com/repos/huggingface/transformers/labels/WIP", "name": "WIP", "color": "234C99", "default": false, "description": "Label your PR/Issue with WIP for some long outstanding Issues/PRs that are work in progress" }, { "id": 2991663546, "node_id": "MDU6TGFiZWwyOTkxNjYzNTQ2", "url": "https://api.github.com/repos/huggingface/transformers/labels/Testing", "name": "Testing", "color": "19A601", "default": false, "description": "" } ]

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

[ "`/usr/bin/time -v`'s output for `-n3`:\r\n```\r\n Command being timed: \"python -m pytest -n 3 --dist=loadfile -s --make-reports=tests_torch ./tests/\"\r\n User time (seconds): 1507.88\r\n System time (seconds): 56.49\r\n Percent of CPU this job got: 261%\r\n Elapsed (wall clock) time (h:mm:ss or m:ss): 9:59.30\r\n Average shared text size (kbytes): 0\r\n Average unshared data size (kbytes): 0\r\n Average stack size (kbytes): 0\r\n Average total size (kbytes): 0\r\n Maximum resident set size (kbytes): 7038804\r\n Average resident set size (kbytes): 0\r\n Major (requiring I/O) page faults: 1\r\n Minor (reclaiming a frame) page faults: 11559434\r\n Voluntary context switches: 3084008\r\n Involuntary context switches: 171112\r\n Swaps: 0\r\n File system inputs: 16456\r\n File system outputs: 3261440\r\n Socket messages sent: 0\r\n Socket messages received: 0\r\n Signals delivered: 0\r\n Page size (bytes): 4096\r\n Exit status: 0\r\n```\r\nwith `-n 4`:\r\n\r\n```\r\n Command being timed: \"python -m pytest -n 4 --dist=loadfile -s --make-reports=tests_torch ./tests/\"\r\n User time (seconds): 1533.02\r\n System time (seconds): 56.00\r\n Percent of CPU this job got: 306%\r\n Elapsed (wall clock) time (h:mm:ss or m:ss): 8:37.98\r\n Average shared text size (kbytes): 0\r\n Average unshared data size (kbytes): 0\r\n Average stack size (kbytes): 0\r\n Average total size (kbytes): 0\r\n Maximum resident set size (kbytes): 5797344\r\n Average resident set size (kbytes): 0\r\n Major (requiring I/O) page faults: 1231\r\n Minor (reclaiming a frame) page faults: 11090301\r\n Voluntary context switches: 2680563\r\n Involuntary context switches: 433387\r\n Swaps: 0\r\n File system inputs: 277920\r\n File system outputs: 3261200\r\n Socket messages sent: 0\r\n Socket messages received: 0\r\n Signals delivered: 0\r\n Page size (bytes): 4096\r\n Exit status: 0\r\n```\r\n\r\nSo clearly this is not right max rss is smaller for `-n 4` then `-n 3` so it appears not to include `pytest` workers. The online information has very conflicting statements about whether forked processes are accounted for or not.\r\n\r\nSo we can't use this one.\r\n", "Thank you for this very in-depth analysis of the situation. It would probably be helpful to have a visualization of each test and how much memory it takes, it could help in singling out memory outliers; and it could also help to detect whether we actually have a memory leak.", "Yes, this is all a big project. Just little time to do it.\r\n\r\nI think the low-hanging fruit is to use `flake-finder` on some tests and see if the memory grows, to first identify if we have a leak. Normally unit-test refuses running the same test more than once.\r\n\r\nhttps://huggingface.co/transformers/testing.html#repeat-tests\r\n\r\nSo may be even an exhaustive search:\r\n\r\nfor each test record mem usage while:\r\n - run test once\r\n - run test 10 times\r\n \r\nI will see if I find some resources to try that.", "This pytest plugins looked promising: https://github.com/CFMTech/pytest-monitor but I can't get it to work.\r\n\r\nAccording to docs you just:\r\n```\r\npip install pytest-monitor\r\n```\r\nand then run `pytest` normally, and it should create a sqlite db with all the data in it, but when I open it I get no test records in it:\r\n\r\n```\r\npytest tests/test_logging.py\r\nsqlite3 .pymon\r\nsqlite> select * from TEST_METRICS;\r\n```\r\n\r\nIt should print the resource stats here, but it doesn't.\r\n\r\nI do get the sessions recorded, but it's not what we want:\r\n```\r\nsqlite> select * from EXECUTION_CONTEXTS;\r\n266c14dea4f9f8a6dae5e46be30e70b3|12|4367.12125|x86_64|Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz|128696|hope|x86_64|64bit|Linux - 5.4.0-70-generic|3.8.8 (default, Feb 24 2021, 21:46:12)\r\n[GCC 7.3.0]\r\nfaa034f4c783dc951159d07212d3a200|12|4300.12775|x86_64|Intel(R) Core(TM) i7-8700 CPU @ 3.20GHz|128696|hope|x86_64|64bit|Linux - 5.4.0-70-generic|3.8.8 (default, Feb 24 2021, 21:46:12)\r\n[GCC 7.3.0]\r\n```\r\n\r\nPerhaps I'm missing something - read through the whole long docs at https://pytest-monitor.readthedocs.io/en/latest/index.html but I don't see that I'm doing anything wrong.\r\n\r\n**edit**: it doesn't work with unittest - too bad it doesn't mention that fact on their website. I created a normal test and then it records data.\r\n", "A little bit at a time I've been trying to work on this issue. At the moment trying to find a reliable way to take the measurements.\r\n\r\nI have thought of at least 3 main ways the leak could be occurring.\r\n\r\n1. leak in some API \r\n2. leak in badly written test that doesn't clean up after itself - so some object is created in the test and somehow it doesn't get destroyed (see also 3) \r\n3. \"functional leak\" as a side-effect of loading extra libraries - say we have 10 tests each loading 10 different libraries - each test will then make `pytest` grow just because it loaded something new - which is a variation on (2) - but how could a test unload the libraries it loaded. It'd be very inefficient practically.\r\n\r\nDetection:\r\n\r\n1) should be easy to detect by re-running the same test and noticing memory grow. My current algorithm - is to run the test once, ignore the memory usage because it could be loading a new module/lib, and run it second time to notice any difference here.\r\n\r\n2 and 3.) these are difficult to make sense of and thus much harder to catch (2) because by just looking at numbers one doesn't know if it was just a new library loaded, or was some object not cleaned up after the test." ]

[]

[]

[]

[]

[]

[ "Should also be passed to the constructor for `DistributedLengthGroupedSampler`?", "Good point, added it!" ]

[]

[]

[]

[]

[]

[ "thanks alot ", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "[https://huggingface.co/transformers/examples.html](https://huggingface.co/transformers/examples.html) now points towards [https://huggingface.co/docs/transformers/main/en/examples](https://huggingface.co/docs/transformers/main/en/examples), which is a 404." ]

[]

[]

[ "Thanks for flagging. The PR mentioned above will make it default to accuracy. Of course, you're free to change it to whatever you need!", "Glad to be of help!\r\nAh, I did not see that - I only looked in the Issues to see if it had already been reported. I will check the PR's as well the next time.\r\nThanks very much!", "The PR did not exist before you flagged the issue ;-) I opened it to fix it!", "Ah! Excellent! I am using the PR you proposed locally so I am back in business :)" ]

[]

[]

[ "That's correct indeed, and seems like a bug to me. Would you like to open a PR to fix the issue?", "Thanks for the confirmation. Yes, I should be able to do that.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "Not sure how to help here... can you download the files from another machine and sync them to the restricted env?", "No, that's the only problem that all the devices within organization are same network and have same policies(as mentioned above).. :)\r\nI see huggingFace has already provided many methods but the challenge is that none of it is working for me ..\r\n\r\nBut it does support downloading other format(i.e. i recently downloaded models from Easyocr in .pth and Spacy language models) , So just a query if any solution can be provided around that which can help me using Hugginface models in such restricted environment..\r\n\r\n", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[]

[]

[]

[ "I can add a PR to fix this. " ]

[]

[]

[ "You should pass along small batches to the model to avoid this error: you should create a loop that goes over the I in `range(0, len(padded), batch_size)` and passes along the `padded[i: i+batch_size]` to your model, then concatenates the predictions back together.\r\n\r\nAlso note that this is not a bug in Transformers or a feature request so I invite you to continue the discussion on the [forums](https://discuss.huggingface.co/) if you need further assistance.", "I experienced this same error when using the sentiment analysis pipeline on a list of strings. I set the model argument in the pipeline to `\"nlptown/bert-base-multilingual-uncased-sentiment\"`.", "I am experiencing the same error even on a batch size of 2. ", "same problem" ]

[]

[]

[ "It is because the version of `tokenizers` lib and the version of `transformers` lib do not match." ]

[]

[]

[]

[]

[]

[ "Thanks!" ]

[ { "id": 2934977194, "node_id": "MDU6TGFiZWwyOTM0OTc3MTk0", "url": "https://api.github.com/repos/huggingface/transformers/labels/Flax", "name": "Flax", "color": "4862AD", "default": false, "description": "" } ]

[]

[]

[ { "id": 2934977194, "node_id": "MDU6TGFiZWwyOTM0OTc3MTk0", "url": "https://api.github.com/repos/huggingface/transformers/labels/Flax", "name": "Flax", "color": "4862AD", "default": false, "description": "" } ]

[]

[]

[]

[]

[ "Hey @startnew,\r\n\r\nSorry could you add some code that we could copy-paste into a terminal to reproduce the error? :-) I don't quite follow here - is it the official code (in `src/transformes`) that doesn't work or specific adapted code?", "Thank you for your reply, @patrickvonplaten , you can reproduce the error I encountered by opening the colab link below\r\n[https://colab.research.google.com/drive/1DiMkU0lEeZqj2AT9DrafP_X-PDZMzxuK#scrollTo=GeOz-1Ix-5HE](https://colab.research.google.com/drive/1DiMkU0lEeZqj2AT9DrafP_X-PDZMzxuK#scrollTo=GeOz-1Ix-5HE)", "refer from offical config ,the difference is I give an custum \"embedding_size\" and the \"embeddding_size\" is not equal to \"hidden_size\",the official use \"hidden_size” as \"embedding_size\", I guess this is the cause of the error", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "Hey @startnew,\r\n\r\nIt's sadly a bit too time consuming to dive into another repo - could you maybe post a short code snippet that shows your error without using any external github repos or code? \r\n\r\nUsually, you should be able to customize the `\"embedding_size\"` configuration parameter" ]

[]

[]

[]

[]

[]

[ "cc @philschmid ", "Hey @laphang, \r\n\r\nCould you please share your `estimator` configuration? that would help debug and reproduce your problem. Thanks!\r\n", "@laphang I tried to reproduce your error and for me its works using the following `HuggingFace` estimator. \r\n```python\r\n# estimator\r\nhuggingface_estimator = HuggingFace(entry_point='run_glue.py',\r\n source_dir='./scripts',\r\n metrics_definition=metric_definitions,\r\n instance_type=instance_type,\r\n instance_count=instance_count,\r\n volume_size=volume_size,\r\n role=role,\r\n transformers_version='4.4.2',\r\n pytorch_version='1.6.0',\r\n checkpoint_s3_uri=f's3://{sess.default_bucket()}/checkpoints',\r\n py_version='py36',\r\n distribution= distribution,\r\n hyperparameters = hyperparameters,\r\n debugger_hook_config=False)\r\n```\r\nThis estimator just extends the estimator from our [04_distributed_training_model_parallelism](https://github.com/huggingface/notebooks/blob/master/sagemaker/04_distributed_training_model_parallelism/sagemaker-notebook.ipynb) and includes the `checkpoint_s3_uri`. \r\n\r\n\r\n\r\n> ## Environment info\r\n> * `transformers` version: 4.5.0\r\n> * Platform: AWS Sagemaker\r\n> * Python version: 3.6\r\n> * PyTorch version (GPU?): 1.7.1\r\n> * Tensorflow version (GPU?):\r\n> * Using GPU in script?: yes\r\n> * Using distributed or parallel set-up in script?: yes\r\n\r\nReading your **environment** it seems that you are not yet using the new Hugging Face Deep Learning Container for Amazon SageMaker. Is that true? or have you update them? ", "@philschmid \r\nAh yes, I'm still using the PyTorchEstimator and installing transformers via requirements.txt. I'll try again with the HuggingFace Estimator and get back to you guys. Thanks for the quick response.", "@philschmid yeah, made the changes below from using the PyTorch estimator to the HuggingFace one, and now distributed training with s3 checkpoints is working properly now (training job completes successfully, and all the checkpoints are synced to s3). It's working both using Sagemaker distributed model parallel, and also using torch.distributed.launch\r\n\r\nAlso just wanted to say that I was pleasantly surprised with how seamlessly Transformers is working with SageMaker model parallel. Great work guys!\r\n\r\n```\r\nbefore:\r\nestimator = PyTorch(base_job_name=job_name, \r\n entry_point = 'run_clm.py', \r\n source_dir=source_dir,\r\n code_location=output_path,\r\n role=role,\r\n framework_version='1.7.1',\r\n py_version='py3', \r\n hyperparameters=hyperparameters,\r\n tags=tags, \r\n output_path=output_path, \r\n checkpoint_s3_uri=checkpoint_path, \r\n instance_count=1, \r\n instance_type='ml.p4d.24xlarge', \r\n distribution= distribution, \r\n use_spot_instances=train_use_spot_instances,\r\n max_run=train_max_run,\r\n max_wait=train_max_wait, \r\n metric_definitions=metric_definition\r\n )\r\n\r\nafter:\r\nestimator = HuggingFace(base_job_name=job_name, \r\n entry_point = 'run_clm.py', \r\n source_dir=source_dir,\r\n code_location=output_path,\r\n role=role,\r\n transformers_version='4.4.2',\r\n pytorch_version='1.6.0',\r\n py_version='py36', \r\n hyperparameters=hyperparameters,\r\n tags=tags, \r\n output_path=output_path, \r\n checkpoint_s3_uri=checkpoint_s3_uri, \r\n debugger_hook_config=False,\r\n instance_count=1, \r\n instance_type='ml.p4d.24xlarge', \r\n distribution= distribution, \r\n use_spot_instances=train_use_spot_instances,\r\n max_run=train_max_run,\r\n max_wait=train_max_wait, \r\n metric_definitions=metric_definition\r\n )\r\n```\r\n\r\n", "@laphang that are great news and thank you for the kind words! 🤗 \r\nShould you have any questions or problems in the future feel free to tag me directly in the issue.", "@philschmid I am getting the same error as @laphang was getting, even with hugging face estimator. Only, the first checkpoints is getting saved in the checkpoint_uri_location and rest don't appear in s3. After the end of training job, it is taking an hour showing uploading and ends with an error \"InternalServerError: We encountered an internal error. Please try again\".\r\n\r\nIt has started since I added sagemaker distributed data parallel into Hugging face estimator. It has kind of become a blocker for our model training, any help would be really appreciated.\r\n\r\n`distribution = {'smdistributed':{'dataparallel':{ 'enabled': True }}}\r\n\r\nhuggingface_estimator = HuggingFace(\r\n entry_point='train.py',\r\n source_dir='./scripts',\r\n sagemaker_sess=sess,\r\n instance_type='ml.p4d.24xlarge',\r\n instance_count=1,\r\n volume_size=60,\r\n code_location=output_path,\r\n output_path=output_path,\r\n checkpoint_s3_uri=checkpoint_s3_uri,\r\n tensorboard_output_config=tensorboard_output_config,\r\n role=role,\r\n transformers_version='4.6.1',\r\n pytorch_version='1.7.1',\r\n py_version='py36',\r\n hyperparameters = hyperparameters,\r\n distribution=distribution\r\n)`", "Hey @Harshitcmd, \r\n\r\ncould maybe share your training script? which `TrainingArguments` are you using? \r\n\r\nFor \r\n> After the end of training job, it is taking an hour showing uploading and ends with an error \"InternalServerError: We encountered an internal error. Please try again\".\r\n\r\nIt might be possible that you are saving your checkpoint in `/opt/ml/model` (which will be uploaded to s3 after training) and it gets through saving the checkpoints. \r\n", "Hey @philschmid thanks for replying. \r\n\r\nI have been saving my checkpoints into \"check_dir\": \"/opt/ml/checkpoints\". Before integrating data parallelism with p4d.24xlarge I was using p3.2xlarge with the same training arguments and there all the checkpoints were getting saved into s3 on the go itself. \r\n\r\nPlz have a look into my training arguments.\r\n\r\n` \r\n\r\n training_args = TrainingArguments(\r\n output_dir=args.check_dir,\r\n num_train_epochs=args.epochs,\r\n per_device_train_batch_size=args.train_batch_size,\r\n per_device_eval_batch_size=args.eval_batch_size,\r\n eval_accumulation_steps=1,\r\n warmup_ratio=args.warmup_steps,\r\n evaluation_strategy=\"no\",\r\n logging_dir=f\"/opt/ml/output/tensorboard/\",\r\n learning_rate=float(args.learning_rate),\r\n save_total_limit=10,\r\n save_steps = 200,\r\n logging_steps = 20,\r\n )\r\n\r\n # create Trainer instance\r\n trainer = Trainer(\r\n model=model,\r\n args=training_args,\r\n train_dataset=train_dataset,\r\n eval_dataset=test_dataset,\r\n tokenizer=tokenizer,\r\n data_collator=data_collator,\r\n callbacks=[TensorBoardCallback]\r\n )`\r\n\r\n", "You might need to add `overwrite_output_dir` to your `TrainingArguments`\r\n> overwrite_output_dir (bool, optional, defaults to False) – If True, overwrite the content of the output directory. Use this to continue training if output_dir points to a checkpoint directory.\r\nI added it for example like that\r\n```python\r\noverwrite_output_dir=True if get_last_checkpoint(args.output_dir) is not None else False,\r\n```\r\n\r\nand to solve your upload issue you should save the model into `/opt/ml/model`. ", "Hey @philschmid,\r\n\r\nI tried adding overwrite_output_dir=True, it's partially solved my issue. Now, the checkpoints are in sync with s3(all the checkpoints and model artifacts are getting saved at the desired location). Even though all the checkpoints got uploaded to the s3 it has showed the status as **Uploading** for an hour and ended with an internal error(weird).\r\n\r\nPS: When I didn't integrate the data parallelism with the same instance type (p4d.24xlarge) everything worked seamlessly. " ]

[]

[]

[ "Follow-up note (part of @lorr1's team that encountered this). This is particularly insidious for any sort of code that tries training with multiple random seeds; there's an assumption that across seeds, weight initialization (for pre-training, fine-tuning weights), dropout, AND data order are all different (and all do have significant bearing on results). \r\n\r\nConsistent data order (as in the existing code) runs counter to that expectation.", "I'm guessing we want a new argument to control that seed though, not the current `args.seed` that is set at the beginning of training, what do you think? ", "I think the seed set at the beginning of training would be fine -- that would be the expected behavior (weights get randomly initialized, then data order is random _conditioned on a single seed_.\r\n\r\nAdding a separate seed just for data order means it's just one more thing you need to keep track of.\r\n\r\nThere's a backwards compatibility issue here possibly (if folks doing multiple random seeds worth of runs have been relying on/reporting those results), but this feels like the simplest solution?", "It's definitely the easiest solution. For the backward compatibility issue, I hope users save their version of Transformers and PyTorch along the seeds for reproducibility. PyTorch does not guarantee the same results across versions (we had an issue with multinomial that changed behavior for instance). So I think it's fine to change the behavior, especially as it's a bug fix.\r\n\r\nWill make a PR with the change." ]

[]

[]

[ "Hi\r\nI also have observed the same issue with t5-base model and mt5-small model ", "Could you maybe make use of the `group_by_length` training argument? This will put the largest batch first to make sure OOM are detected in the very beginning (best feature ever by @sgugger )", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "@LysandreJik , most of it was easy to figure out by looking at other tokenizers. The setup and testing guidelines were very easy to follow, I was up and running very quickly.\r\n\r\nFor the fast tokenizers, few things that might help someone like me who is new to the transformers library:\r\n1. A top-level difference between the fast and slow tokenizers. At first, I did not know there was a tokenization library and took some time to figure that out.\r\n2. Overview of how we implement tokenizers. Things like what do the vocab files do and what does the merges_file do. (Although this could just be me.)\r\n3. In some fast tokenizers, there are files listed which are not being used. That confused me initially as I thought we needed those files for a fast tokenizer.\r\n\r\nHope this helps. \r\n", "@LysandreJik, \r\n\r\n#10498 mentioned implementing a tokenizer for deberta v2 as well. I have created a new feature request #11529 for that.", "Thank you @ShubhamSanghvi, this is all very helpful. We'll take care of including that in the documentation so that it's clearer from now on.\r\n\r\nWill take a look at #11529!" ]

[]

[]

[ "Hi! I'm the TF maintainer for 🤗 Transformers right now. Thanks for this, the code quality looks really good! There's one issue, though - we're currently trying to move away from `TFTrainer` and use more native, idiomatic TF code based on Keras. You can see an example of the kind of TFTrainer-free approach we're working on [here](https://github.com/huggingface/transformers/blob/master/examples/tensorflow/text-classification/run_text_classification.py).\r\n\r\nAs a result, we probably can't accept `Seq2SeqTFTrainer` in the main library right now, but we're definitely planning on adding a Seq2Seq TF example soon. If you'd like, you can try to convert this PR to a TFTrainer-free Seq2Seq example script and put it in /examples/tensorflow, but I understand if that's a lot of work and you don't want to bother right now! ", "Hi @Rocketknight1! Thanks for your feedback, I understand. This actually emerged as a side product from a project that I'm currently working on so I thought I'd share this. But it's good to know the direction you're heading. If you're planning to add some seq2seq examples for Keras in the next few weeks then it's fine to close this PR I guess. Otherwise I will probably rewrite this code in order to be aligned with the Huggingface library and its overall direction. Do you have any date on your mind for the seq2seq TF examples?", "Hi, sorry for the delay! We are indeed planning to include those models, hopefully in a month or so. I don't have an exact date, but it's on my To Do list.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "@Riroaki I agree with you! It should be:\r\n```python\r\ntokenizer([[prompt, choice0], [prompt, choice1]], return_tensors='pt', padding=True)\r\n```", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "@sgugger @SBrandeis \r\nPlease have a look at the example scripts.🙏", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "I'm running into this issue as well.\r\n\r\nI'm not super familiar working with multiple choice models, but I think that given #6074, and the [run_swag.py](https://github.com/huggingface/transformers/blob/master/examples/pytorch/multiple-choice/run_swag.py) example, these should be passed as two lists, instead of one.\r\n\r\nIn other words, in the example, instead of\r\n\r\n```python\r\nencoding = tokenizer([[prompt, prompt], [choice0, choice1]], return_tensors='pt', padding=True)\r\n```\r\n\r\nit should be \r\n\r\n```python\r\nencoding = tokenizer([prompt, prompt], [choice0, choice1], return_tensors='pt', padding=True)\r\n```\r\n\r\nIt would be awesome to make this two-character-deletion change, as it just tripped me up when starting working on a multiple choice model!", "This has been fixed, but you need to switch to the master documentation to see the change." ]

[]

[]

[ "_I tried to load 3 .txt files as a dataset_ - bad idea, try to merge them in one, then let's see what you'll get (which error message or result).", "Try this one:\r\n```\r\nfrom pathlib import Path\r\npaths = [str(x) for x in Path(\".\").glob(\"**/*.txt\")]\r\n```\r\n\r\nP.S. Next time just copy-paste code, don't screen it", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[]

[]

[]

[]

[]

[]

[ "Thank you so much for flagging! This should be fixed by the PR above." ]

[]

[]

[ "Hi @sgugger and @stas00,\r\nI have made changes in the following way,\r\n| Earlier | Now |\r\n| ---- | ---- |\r\n| test | predict |\r\n| test_examples | predict_examples |\r\n| test_dataset | predict_dataset |\r\n| max_test_samples | max_predict_samples |\r\n| val | eval |\r\n| val_examples | eval_examples |\r\n| val_dataset | eval_dataset |\r\n| max_val_samples | max_eval_samples |\r\n\r\n* I have also made changes in the trainer code for the above variables\r\n* I have modified the template file accordingly\r\n* `examples/pytorch/question-answering/run_qa_no_trainer.py` don't have code complete code for predict stage. Shall we need to add? \r\n\r\n", "Hi @sgugger,\r\nI was trying to test [run_qa_no_trainer.py](https://github.com/huggingface/transformers/blob/master/examples/pytorch/question-answering/run_qa_no_trainer.py) and I came to know that when I passed `--test_file` it was giving me an error since we don't have such [argument](https://github.com/huggingface/transformers/blob/master/examples/pytorch/question-answering/run_qa_no_trainer.py#L87) in the file.\r\n\r\nWhen i give `--do_predict` it gives an error at this [line](https://github.com/huggingface/transformers/blob/master/examples/pytorch/question-answering/run_qa_no_trainer.py#L507) since I was using squad data and it doesn't have `test` set\r\n\r\nBut as you said it will work perfectly for the dataset with `testset`.", "Ah! now I get it. We can add support for a `--test_file` in another PR, yes!", "I am exploring other files, I will add more changes in docs and other files", "You're doing a very painful task, @bhadreshpsavani, as @sgugger commented we unfortunately can't normalize these 2 names fully both because of the backcompat and also where in some cases something is called validation/test split :( Thank you for doing this important work and bearing with all these setbacks.", "Hi @stas00, \r\nIt's totally fine. \r\nI was expecting more suggestions because I did a lot of changes for all files in a single go. I am enjoying this coding work that's important for me!", "Hi @stas00,\r\nCan we use this command?\r\n`git push --force-with-lease origin myfeature`\r\nI read that this is safe than `--force` and [many organizations even using this.](https://stackoverflow.com/questions/41283955/github-keeps-saying-this-branch-is-x-commits-ahead-y-commits-behind)", "In general all PRs are isolated until they are merged, so if you make a mistake on your own PR, in the worst case you will make a mess of your own changes, but it won't impact the master. So feel free to experiment.\r\n\r\n**edit:** this is for sure for when you don't have write access to upstream master, or if you work in your own fork. I'm not sure if it's the same when one does have a write access and is working directly on the source. I find it much safer to always do all the work in my own fork and merge upstream via PRs.\r\n\r\nI have never used this particular flag before, so if it's safer go for it.", "`run_tests_torch` CI job has been flakey as of recent, You can always force a CI restart if you see one or more CI jobs are failing unrelated to your commit with an empty commit:\r\n\r\n```\r\ngit commit --allow-empty -m \"Trigger CI\"\r\ngit push\r\n```", "This is a cool command. I will definitely need this. I will note this down!\r\nThanks", "Hi @sgugger and @stas00 ,\r\nI have reverted trainer changes and updated the example pytorch readme.\r\nPlease let me know if we need to make more changes ", "@sgugger, should these not be covered too?\r\n\r\n```\r\n$ grep -Ir max_val_samples\r\nexamples/tensorflow/text-classification/run_text_classification.py: max_val_samples: Optional[int] = field(\r\nexamples/tensorflow/text-classification/run_text_classification.py: if data_args.max_val_samples is not None:\r\nexamples/tensorflow/text-classification/run_text_classification.py: eval_dataset = eval_dataset.select(range(data_args.max_val_samples))\r\nexamples/research_projects/wav2vec2/run_common_voice.py: max_val_samples: Optional[int] = field(\r\nexamples/research_projects/wav2vec2/run_common_voice.py: if data_args.max_val_samples is not None:\r\nexamples/research_projects/wav2vec2/run_common_voice.py: eval_dataset = eval_dataset.select(range(data_args.max_val_samples))\r\nexamples/research_projects/wav2vec2/run_common_voice.py: max_val_samples = data_args.max_val_samples if data_args.max_val_samples is not None else len(eval_dataset)\r\nexamples/research_projects/wav2vec2/run_common_voice.py: metrics[\"eval_samples\"] = min(max_val_samples, len(eval_dataset))\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: max_val_samples: Optional[int] = field(\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: if data_args.max_val_samples is not None:\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: eval_dataset = eval_dataset.select(range(data_args.max_val_samples))\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: max_val_samples = data_args.max_val_samples if data_args.max_val_samples is not None else len(eval_dataset)\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: metrics[\"eval_samples\"] = min(max_val_samples, len(eval_dataset))\r\n```\r\n\r\n```\r\n$ grep -Ir max_test_samples\r\nexamples/tensorflow/text-classification/run_text_classification.py: max_test_samples: Optional[int] = field(\r\nexamples/tensorflow/text-classification/run_text_classification.py: if data_args.max_test_samples is not None:\r\nexamples/tensorflow/text-classification/run_text_classification.py: test_dataset = test_dataset.select(range(data_args.max_test_samples))\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: max_test_samples: Optional[int] = field(\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: if data_args.max_test_samples is not None:\r\ntests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py: test_dataset = test_dataset.select(range(data_args.max_test_samples))\r\n```", "This PR should be rebased on master and deal with `examples/tensorflow/text-classification/run_text_classification.py` that was added recently yes.\r\n\r\n`examples/research_projects/wav2vec2/run_common_voice.py` is a research-project so is not actively maintained and pinned to a version of transformers, I would leave it out of this PR.\r\n\r\nFor `tests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py`, I wouldn't touch it since it's a test script (which will ultimately be replaced by a TF example) but I'll let @philschmid decide on this one.", "> For `tests/sagemaker/scripts/pytorch/run_glue_model_parallelism.py`, I wouldn't touch it since it's a test script (which will ultimately be replaced by a TF example) but I'll let @philschmid decide on this one.\r\n\r\nIs a custom version for `SageMakerTrainer`, this will be removed after the deprecation of `SageMakerTrainer`, there is no need for adjustments. ", "Thanks for adjusting! I think this is good to merge, @stas00 if you agree I'll let you click on the button :-) " ]

[]

[ { "login": "Rocketknight1", "id": 12866554, "node_id": "MDQ6VXNlcjEyODY2NTU0", "avatar_url": "https://avatars.githubusercontent.com/u/12866554?v=4", "gravatar_id": "", "url": "https://api.github.com/users/Rocketknight1", "html_url": "https://github.com/Rocketknight1", "followers_url": "https://api.github.com/users/Rocketknight1/followers", "following_url": "https://api.github.com/users/Rocketknight1/following{/other_user}", "gists_url": "https://api.github.com/users/Rocketknight1/gists{/gist_id}", "starred_url": "https://api.github.com/users/Rocketknight1/starred{/owner}{/repo}", "subscriptions_url": "https://api.github.com/users/Rocketknight1/subscriptions", "organizations_url": "https://api.github.com/users/Rocketknight1/orgs", "repos_url": "https://api.github.com/users/Rocketknight1/repos", "events_url": "https://api.github.com/users/Rocketknight1/events{/privacy}", "received_events_url": "https://api.github.com/users/Rocketknight1/received_events", "type": "User", "site_admin": false } ]

[]

[]

[]

[ "Awesome job @GeetDsa - this looks good to me :-) \r\n\r\nAlso pinging @patil-suraj and @Narsil for review.", "@Narsil - do you think we need a test here or should this be fine without?", "Hi @patrickvonplaten , Most of the errors are caused because of \"max_length\" attribute used while creating `beam_scorer` object in the test cases. These errors are arising as the beam_scorer attribute is removed from the source code. Infact, I had modified the files under `tests/*` to take account for this.(This can be found in my commits. Sorry that I did not had different commits for the actual source code and the tests). So, may be can you review the files that I have modified under `tests/*` as well?", "I think we shouldn´t modify the tests if possible (thats what make backward compatiblity enforced).\r\n\r\nInstead, we should probably : \r\n- add some warning for users that used the raw components/functions\r\n- point the towards a better solution\r\n- modify the tests that DO raise warnings to make sure we do raise them (and also swallow them in the logs) (they become backward compatibility tests their names should probably reflect that too) (We can remove some backward tests if they are very redundant btw)\r\n- Have new tests that point show new API.\r\n- Higher level tests (within models) should not be affected.\r\n@GeetDsa I can do this for you if you want.", "> I think we shouldn´t modify the tests if possible (thats what make backward compatiblity enforced).\r\n> \r\n> Instead, we should probably :\r\n> \r\n> * add some warning for users that used the raw components/functions\r\n> * point the towards a better solution\r\n> * modify the tests that DO raise warnings to make sure we do raise them (and also swallow them in the logs) (they become backward compatibility tests their names should probably reflect that too) (We can remove some backward tests if they are very redundant btw)\r\n> * Have new tests that point show new API.\r\n> * Higher level tests (within models) should not be affected.\r\n> @GeetDsa I can do this for you if you want.\r\n\r\nHi @Narsil, I think it would be better if you can do it, as I don't really know how to deal with tests.", "> > I think we shouldn´t modify the tests if possible (thats what make backward compatiblity enforced).\r\n> > Instead, we should probably :\r\n> > \r\n> > * add some warning for users that used the raw components/functions\r\n> > * point the towards a better solution\r\n> > * modify the tests that DO raise warnings to make sure we do raise them (and also swallow them in the logs) (they become backward compatibility tests their names should probably reflect that too) (We can remove some backward tests if they are very redundant btw)\r\n> > * Have new tests that point show new API.\r\n> > * Higher level tests (within models) should not be affected.\r\n> > @GeetDsa I can do this for you if you want.\r\n\r\n\r\n Hi @Narsil, I think it would be better if you can do it, as I don't really know how to deal with tests.\r\n\r\n", "I'll take care of it", "@Narsil - I think we don't need to add a test here. By design the PR prevents errors such as the one mentioned in the issue - It's not possible anymore to pass `max_length` to `BeamSearchScorer` and thus a test doesn't make much sense here IMO.", "Will wait for https://github.com/huggingface/transformers/pull/11442 to be merged before rebasing this one to merge", "@patrickvonplaten But if users were using the BeamScorer object, that's a breaking change, isn't it ?", "> @patrickvonplaten But if users were using the BeamScorer object, that's a breaking change, isn't it ?\r\n\r\nYes true. IMO, no functionality is lost though because:\r\n\r\n- Previously, if one had passed `max_length` to both `def beam_search(...)` and `BeamSearchScorer(...)`, then there would have been a bug (see issue). The correct way of fixing the bug (while still allowing `BeamSearchScorer(...)` to accept `max_length`) would have been to overwrite `BeamSearchScorer's` max_length with `beam_search(...)`'s max_length. On the other hand it's never possible to **not** pass `max_length` to `def beam_search(...)` => therefore I think either way the `max_length` arg to `BeamSearchScorer` is useless (the `max_length` value of `beam_search(...)` would have been preferred in any way. \r\n\r\n=> However, we could/should probably add `**kwargs` to `BeamSearchScorer` that throws a warning if `max_length` is passed & says that's it's deprecated. This would be cleaner overall and have no breaking changes - what do you think @Narsil ?", "Yep, that's what I had in mind, just accept it, raise a warning (and ignore it, exactly as it used to if I understand correctly) ", "> Yep, that's what I had in mind, just accept it, raise a warning (and ignore it, exactly as it used to if I understand correctly)\r\n\r\nDone! Could you review one last time and merge if ok for you? @Narsil ", "@patrickvonplaten, thank you for taking care of it. I found a small issue in the warning message. The message provided in the latest commit is \"\"`max_length` should be passed directly to `beam_search(...)`, `beam_sample(...)`\"\", shouldn't you include \"generate(..)\" as well, as sometime, `generate(...)` inherently calls `beam_search(..,)` or `group_beam_search(..,)`, or other relevant functions.", "> @patrickvonplaten, thank you for taking care of it. I found a small issue in the warning message. The message provided in the latest commit is \"\"`max_length` should be passed directly to `beam_search(...)`, `beam_sample(...)`\"\", shouldn't you include \"generate(..)\" as well, as sometime, `generate(...)` inherently calls `beam_search(..,)` or `group_beam_search(..,)`, or other relevant functions.\r\n\r\nHey @GeetDsa, \r\n\r\nIt should be fine, since the warning is impossible to be triggered by `generate(...)` since when one calls `generate(...)`, one cannot pass `max_length` to `BeamScorer` " ]

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

[]

[ "Think we still need to fill out a couple of things before publishing it :-) ", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored.", "unstale", "@patil-suraj @patrickvonplaten \r\n\r\nI have started implementation of GLM model in HF. Need some input on implementation part.\r\n\r\n1. Original GLM model is implemented using torch parallel distribution , In HF implementation, are we going to keep torch parallel distributed or convert to normal architecture(without parallelism )\r\n\r\n2. There are six version of models (like GLM -base, GLM- large…etc), in which two diffrent tokenization is used Wordpiece and BPE. Like GLM-base and large is having wordpiece and 'GLE-Roberta'and 'GLE-large' is having BPE.\r\n\r\n" ]

[]

[]

[ "Hey @cceyda - thanks for the issue!\r\n\r\nYes, this was an intentional choice since those parameters seemed to be the same in all experiments after studying the paper. If people want to change those params, we could always add them in a later version.", "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[]

[]

[ "This issue has been automatically marked as stale because it has not had recent activity. If you think this still needs to be addressed please comment on this thread.\n\nPlease note that issues that do not follow the [contributing guidelines](https://github.com/huggingface/transformers/blob/master/CONTRIBUTING.md) are likely to be ignored." ]

[ { "id": 2934977194, "node_id": "MDU6TGFiZWwyOTM0OTc3MTk0", "url": "https://api.github.com/repos/huggingface/transformers/labels/Flax", "name": "Flax", "color": "4862AD", "default": false, "description": "" } ]

[]

[]

[]

[]

[]