verl-project/verl

Support Generative Reward Model (GenRM)

Chiusa

#229 aperta il 9 feb 2025

 (17 commenti) (0 reazioni) (1 assegnatario)Python (4426 fork)auto 404
enhancementgood first issue

Metriche repository

Star
 (23.023 stelle)
Metriche merge PR
 (Merge medio 3g 6h) (125 PR mergiate in 30 g)

Descrizione

According to the documentation, veRL only supports AutoModelForSequenceClassification. What would be the best way to implement generative reward model (GenRM) for veRL? I tried looking at FSDP Workers and Megatron-LM Workers but they no longer exist.

Guida contributor