richarddwang electra_pytorch issues

richarddwang / electra_pytorch

Pretrain and finetune ELECTRA with fastai and huggingface. (Results of the paper replicated !)

324 stars 41 forks source link

issues

Newest

Newest Most commented Recently updated Oldest Least commented Least recently updated

Learner class missing?

#42 RalfKellner opened 2 months ago
0
bug: fix accuracy metric and add `test_with_y=True`

#41 EllingtonKirby opened 8 months ago
0
Restarting from previous checkpoint

#40 jbrry opened 11 months ago
0
Python version

#39 alexpancheva opened 1 year ago
0
License info is required

#38 Ab1992ao opened 1 year ago
0
Custom Dataset

#37 asparius closed 1 year ago
0
NameError: name 'sort_by_run' is not defined

#36 tomo-logbii closed 1 year ago
2
Relative importance of different "tricks" in README

#35 vmurahari3 closed 2 years ago
1
SST-2 accuracy is 50% after finetuning

#34 xiamengzhou closed 2 years ago
10
Use different tokenizer (and specify special tokens)

#33 ghost closed 2 years ago
1
Pyarrow dataloading issue

#32 skaulintel closed 2 years ago
2
Training time and ++ version

#31 Luuk99 closed 2 years ago
1
如何多卡并行训练模型（How to train multi-card models in parallel？）

#30 MrSworder closed 2 years ago
1
How to load the cached data from ELECTRAProcessor?

#29 cooelf closed 3 years ago
1
Is it possible to perform the fine-tuning within the HuggingFace library?

#28 lucaguarro closed 3 years ago
3
Is multi_task.py in a working state and if so how should one use it?

#27 lucaguarro closed 3 years ago
1
How can I pretrain ELECTRA starting from weights from google ?

#26 richarddwang closed 3 years ago
9
How do I continue language model training?

#25 PhilipMay closed 3 years ago
3
Small typo in the README.md

#24 AminTaheri23 closed 3 years ago
1
Fixing typo mistake

#23 ngoquanghuy99 closed 3 years ago
0
Sequence length too long for `ELECTRADataProcessor`.

#22 PhilipMay closed 3 years ago
3
GPU utilization falls back to 0% when training with multiple GPUs

#21 PhilipMay closed 3 years ago
5
How do I extract and save the discriminator from the checkpoint?

#20 PhilipMay closed 3 years ago
3
Pretrain does not save the model.

#19 PhilipMay closed 3 years ago
2
Assignment without usage.

#18 PhilipMay closed 3 years ago
1
How do I pretrain on multiple GPUs?

#17 PhilipMay closed 3 years ago
2
Error: `ELECTRAModel` object has no attribute `to_native_fp16`

#16 PhilipMay closed 3 years ago
1
Is it right that your input data is not sentence splitted?

#15 PhilipMay closed 3 years ago
2
Description to use "just text files".

#14 PhilipMay closed 3 years ago
4
ipython package is missing

#13 PhilipMay closed 3 years ago
1
Question about the use of with no grad

#12 xiayandi closed 3 years ago
1
Questtion about ELECTRADataProcessor or ExampleBuilder

#11 miyamonz closed 3 years ago
2
RAM usage during training

#10 Tiiiger closed 3 years ago
7
how long to train electra base? (i have 30GB text, 1 GPU v100)

#9 BinhMinhs10 closed 3 years ago
1
back-propagate the discriminator loss through the generator

#8 mehdimashayekhi closed 3 years ago
1
Cannot run

#7 wangyu1997 closed 3 years ago
2
Glue score during pre-training

#6 Albert-Ma closed 3 years ago
1
multi-GPU training

#5 gchlodzinski closed 3 years ago
11
duration of training

#4 amy-hyunji closed 4 years ago
1
Training from scratch with other datasets (other languages)

#3 dumitrescustefan closed 4 years ago
3
Error in store_attr()

#2 glakshmidhar closed 4 years ago
1
providing datasets

#1 amy-hyunji closed 4 years ago
2