google-research/albert issues and pull requests

#97 - [ALBERT] Fine-tune SQUAD, got error in `convert_examples_to_features()`

Issue - State: open - Opened by wxp16 almost 5 years ago - 4 comments

#96 - [ALBERT]error "expected str instance, bytes found" to error "list index out of range"

Issue - State: closed - Opened by zsweet almost 5 years ago - 4 comments

#95 - Any plans on multilangual albert models release?

Issue - State: closed - Opened by maxbaluev almost 5 years ago - 1 comment

#94 - LookupError: No gradient defined for operation 'module_apply_tokens/bert/encoder/transformer/group_0_11/layer_11/inner_group_0/ffn_1/intermediate/output/dense/einsum/Ein$um' (op type: Einsum)

Issue - State: closed - Opened by MichaelCaohn almost 5 years ago - 22 comments

#93 - How to generate vocab.txt file?

Issue - State: closed - Opened by sunnysaxena almost 5 years ago - 4 comments

#92 - KeyError: '[CLS]' in tokenization.py

Issue - State: closed - Opened by acmilannesta almost 5 years ago - 2 comments

#91 - Multilingual ALBERT

Issue - State: closed - Opened by flaviussn almost 5 years ago - 5 comments

#90 - [ALBERT] Failed to load TF-Hub model on Google Colab

Issue - State: closed - Opened by xf05888 almost 5 years ago - 3 comments

#89 - [ALBERT] albert-xlarge V2 seems to have a different behavior than the other models

Issue - State: open - Opened by LysandreJik almost 5 years ago - 5 comments

#88 - [ALBERT] FullTokenizer inconsistency - do_lower_case ignored when spm_model_file is specified

Issue - State: open - Opened by kpe almost 5 years ago

#87 - Finetune ALBERT using pretained model

Issue - State: closed - Opened by MichaelCaohn almost 5 years ago - 6 comments

#86 - Fine-tuning Albert large

Issue - State: open - Opened by SohaKhazaeli almost 5 years ago - 8 comments

#85 - [ALBERT] question : unnecessary(or code error) in tokenization.py?

Issue - State: open - Opened by prokok almost 5 years ago - 3 comments

#84 - [ALBERT] How many value of 'dupe_factor' is good?

Issue - State: closed - Opened by ngoanpv almost 5 years ago - 2 comments

#83 - [ALBERT] How to know memory consumption on GPU/TPU of a model?

Issue - State: open - Opened by ngoanpv almost 5 years ago - 1 comment

#82 - tensorflow.python.framework.errors_impl.NotFoundError: Graph ops missing from the python registry ({'Einsum'}) are also absent from the c++ registry

Issue - State: closed - Opened by ntaherkhani almost 5 years ago - 5 comments

#81 - [UNK] token in v2 models

Issue - State: open - Opened by kkkppp almost 5 years ago - 5 comments

#80 - Incorrect English alphabet in line 402 in create_pretraining_data.py

Issue - State: open - Opened by tonyzhao6 almost 5 years ago - 1 comment

#79 - [ALBERT] Pre-training on TPU Pod

Issue - State: open - Opened by ngoanpv almost 5 years ago - 1 comment

#78 - [ALBERT] TFHub assets/albert_config.json has 0 for dropouts

Issue - State: open - Opened by kpe almost 5 years ago - 2 comments

#77 - Training time

Issue - State: open - Opened by peregilk almost 5 years ago - 1 comment

#76 - [ALBERT] Same amount of TPU memory consumption compared to BERT

Issue - State: open - Opened by jinmel almost 5 years ago - 1 comment

#75 - [ALBERT] what are the parameters setting for training data generation ?

Issue - State: open - Opened by wxp16 almost 5 years ago

#74 - [ALBERT] : Gradient for bert/embeddings/LayerNorm/gamma:0 is NaN : Tensor had NaN values [[node CheckNumerics_4 (defined at usr/local/lib/python3.5/dist-packages/tensorflow_core/python/framework/ops.py:1748) ]]

Issue - State: open - Opened by SUMMER1234 almost 5 years ago

#73 - How to train and test the ALBERT model.

Issue - State: closed - Opened by VikasRajashekar almost 5 years ago - 1 comment

#72 - Model for SQuAD Question Answering

Issue - State: closed - Opened by GaneshGS almost 5 years ago - 3 comments

#71 - [ALBERT] albert-xxlarge v1 `albert_config.josn/30k-clean.model` file

Issue - State: closed - Opened by lsq357 almost 5 years ago - 1 comment

#70 - Issues with codes in Google Colab-Tensorflow-2.0.0

Issue - State: closed - Opened by GaneshGS almost 5 years ago - 2 comments

#67 - race score using run_race.py with albert base v2 is so much higher than score in table(66.8)

Issue - State: open - Opened by seongwook-ham over 4 years ago - 10 comments

#64 - tokenization encode_pieces error

Issue - State: open - Opened by akakakakakaa over 4 years ago - 1 comment

#63 - Add scripts for evaluating RACE dataset.

Pull Request - State: closed - Opened by copybara-service[bot] over 4 years ago
Labels: cla: yes

#62 - Missing some weights parameters of the released albert_base_v2

Issue - State: closed - Opened by shenlanyilang over 4 years ago - 1 comment

#61 - ALBERT export_to_tfhub.py

Pull Request - State: closed - Opened by copybara-service[bot] over 4 years ago
Labels: cla: yes

#60 - `run_classifier_with_tfhub.py` is missing.

Issue - State: closed - Opened by bobloki over 4 years ago - 5 comments

#59 - Thank CLUE team for providing Chinese language training data

Pull Request - State: closed - Opened by copybara-service[bot] over 4 years ago
Labels: cla: yes

#58 - SentencePiece Models for Chinese Models Missing?

Issue - State: open - Opened by simonc256 over 4 years ago - 7 comments

#56 - impossible to reproduce glue result for mnli-m. did i do something wrong?

Issue - State: closed - Opened by seongwook-ham over 4 years ago - 4 comments

#55 - Finetune on RACE

Issue - State: closed - Opened by ViktorAlm over 4 years ago - 1 comment

#54 - do u have some contrastive datas between chinese Albert with other chinese pretraining models?

Issue - State: open - Opened by xiongma over 4 years ago - 3 comments

#53 - It seems that ALBERT didn't use sentencepiece in Chinese, but BERT's tokenizer?

Issue - State: open - Opened by freefuiiismyname over 4 years ago

#52 - adding Chinese models

Pull Request - State: closed - Opened by copybara-service[bot] over 4 years ago
Labels: cla: yes

#51 - max_predictions_per_seq=20 for seq_len=512

Issue - State: closed - Opened by YeDeming over 4 years ago - 4 comments

#50 - Where can I find the "spm_model_file" when run run_squad_v2.py

Issue - State: closed - Opened by wmmxk over 4 years ago - 9 comments

#49 - learning rate in optimization.py is so confusing

Issue - State: closed - Opened by seongwook-ham over 4 years ago - 2 comments

#48 - run_glue.sh script and documentation improvements

Pull Request - State: closed - Opened by copybara-service[bot] over 4 years ago
Labels: cla: yes

#47 - Necessary readings to understand concepts used In ALBERT

Issue - State: closed - Opened by satheeshravir over 4 years ago - 1 comment

#46 - Preventing estimator from deleting checkpoints

Pull Request - State: closed - Opened by copybara-service[bot] over 4 years ago
Labels: cla: yes