Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection

Devs

Scaling Laws for Forgetting during Finetuning with Pretraining Data Injection | Read Paper on Bytez