ELECTRA is a transformer with a new pre-training approach which trains two transformer models: the generator and the discriminator. The generator replaces tokens in the sequence - trained as a masked language model - and the discriminator (the ELECTRA contribution) attempts to identify which tokens are replaced by the generator in the sequence. This pre-training task is called replaced token detection, and is a replacement for masking the input.
Source: ELECTRA: Pre-training Text Encoders as Discriminators Rather Than GeneratorsPaper | Code | Results | Date | Stars |
---|
Task | Papers | Share |
---|---|---|
Language Modelling | 31 | 14.62% |
Sentence | 16 | 7.55% |
Question Answering | 16 | 7.55% |
Reading Comprehension | 10 | 4.72% |
Natural Language Inference | 7 | 3.30% |
Natural Language Understanding | 6 | 2.83% |
Text Classification | 5 | 2.36% |
Named Entity Recognition (NER) | 5 | 2.36% |
Multi-Task Learning | 5 | 2.36% |