Top fairseq issues

Implement Reducing Transformer Depth on Demand With Structred Dropout

gvskalyan picture gvskalyan·16 Oct 2019·6 Comments

BLEU score from generate.py is cased BLEU or not?

shawnkx picture shawnkx·1 Oct 2019·2 Comments

Resume training by specifying the model you'd like to resume from using `--restore-file <path to checkpoint>`.

aastha19 picture aastha19·27 Sep 2019·7 Comments

RoBERTa and 514

LysandreJik picture LysandreJik·26 Sep 2019·4 Comments

Need more information about distributed training

jalajthanaki picture jalajthanaki·19 Sep 2019·4 Comments

Using sentencepiece tokenizer in RoBERTa pretraining regiment

nitinnairk picture nitinnairk·3 Sep 2019·2 Comments

Fine-tuning RoBERTa on custom dataset.

PhilippeMarcotte picture PhilippeMarcotte·20 Aug 2019·1 Comments

RoBERTa RACE preprocessing doesn't handle newlines

nelson-liu picture nelson-liu·12 Aug 2019·4 Comments

RuntimeError: Expected object of backend CPU but got backend CUDA for argument #4 'mat1'

sam3oh5 picture sam3oh5·7 Aug 2019·6 Comments
👍1

Facebook FAIR’s WMT19 News Translation Task Submission

vince62s picture vince62s·29 Jul 2019·8 Comments
 Previous issues
Next issues