search

RTIInternational / gobbli

Deep learning with text doesn't have to be scary.
Apache License 2.0
275 stars 23 forks source link

Transformers gradient accumulation #13

Closed jasonnance closed 4 years ago

jasonnance commented 4 years ago

Description of Changes

Upgrades from pytorch-transformers to transformers and adds gradient accumulation to the Transformer model.