I want to generate a sentence from a machine translation model with constrained decoding that requires a custom BeamScorer. Is there a way how to replace the standard BeamSearchScorer while using the high-level API such as the Translate pipeline or the generate method without explicitly calling the beam_search method explicitly?
Asked
Active
Viewed 8 times
1
Jindřich
- 10,270
- 2
- 23
- 44