microsoft / MASS

MASS: Masked Sequence to Sequence Pre-training for Language Generation
https://arxiv.org/pdf/1905.02450.pdf
Other
1.11k stars 206 forks source link

How many Gpu needed for Text Summarization of CNNDM fine-tuning ? #134

Closed fseasy closed 4 years ago

fseasy commented 4 years ago

Hi, Thanks for the great work! I found the summarization guid but it lack the info about gpu num. I searched the issues and found gigaword use 8 gpu? So would CNNDM also need 8 gpus with the given script to reproduce the result?

nlpyang commented 4 years ago

Hi, I think they used 8 gpus for CNNDM, you can parse the released pt file to see their training arguments.

fseasy commented 4 years ago

Hi, I think they used 8 gpus for CNNDM, you can parse the released pt file to see their training arguments.

Great! Thanks