Enable loading ckpt for t0 finetuning

bigscience-workshop / Megatron-DeepSpeed

Ongoing research training transformer language models at scale, including: BERT & GPT-2

Other

1.31k stars 213 forks source link

Open Muennighoff opened 2 years ago