Top transformers issues

BerTweet tokenizer issue

Modeling wontfix
Shiro-LK picture Shiro-LK·22 Jul 2020·11 Comments

Tabert

New model wontfix
flozi00 picture flozi00·5 Jul 2020·4 Comments

model.generate source code

wontfix
chmod-700 picture chmod-700·30 Jun 2020·3 Comments

BertForMaskedLM "labels" is an unexpected keyword

guoxuxu picture guoxuxu·20 Jun 2020·4 Comments

Is there a helper script to preprocess data for T5 for masked language modeling?

wontfix
abhisheksgumadi picture abhisheksgumadi·17 Jun 2020·4 Comments

Cannot load optimizer and lr_scheduler states with TPU training

misrasaurabh1 picture misrasaurabh1·12 Jun 2020·8 Comments

The purpose of files merges.txt, special_tokens_map.json, training_args.bin and add_tokens.json

wontfix
Aktsvigun picture Aktsvigun·5 Jun 2020·6 Comments
👍1

Use BART for longer documents

tuhinjubcse picture tuhinjubcse·11 May 2020·8 Comments
👍2

🐛 Summarization pipeline : T5-base much slower than BART-large

Pipeline
astariul-colanim picture astariul-colanim·3 Apr 2020·4 Comments
👍1

Bart.generate requires config.output_past=True

tuhinjubcse picture tuhinjubcse·30 Mar 2020·13 Comments
 Previous issues
Next issues