Fill-Mask • Updated • 5.16k • • 14
ELECTRA release
updated
This collection regroups the ELECTRA models released by the Google team.
Note Smallest generator model. 12 Layers, 1024 intermediate size, 256 hidden size, 4 attention heads.
Updated • 212k • 38Note Smallest discriminator model. 12 Layers, 1024 intermediate size, 256 hidden size, 4 attention heads.
Fill-Mask • Updated • 2.14k • • 10Note Base generator model. 12 Layers, 1024 intermediate size, 256 hidden size, 4 attention heads.
Updated • 41.3M • 129Note Base discriminator model. 12 Layers, 3072 intermediate size, 768 hidden size, 12 attention heads.
Fill-Mask • Updated • 524 • 9Note Largest generator model. 24 Layers, 1024 intermediate size, 1024 hidden size, 4 attention heads.
Updated • 27.2k • 18Note Largest discriminator model. 24 Layers, 4096 intermediate size, 1024 hidden size, 16 attention heads.