Pre-trained checkpoints for the Gated Recurrent Transformer (GRT). arXiv:2608.15062
-
Gated Recurrent Transformers: Expressive Depth through Recurrent Modulation
Paper • 2608.15062 • Published • 10 -
Amr-Hegazy/grt-large-isoflop
Text Generation • 0.3B • Updated • 216 -
Amr-Hegazy/grt-large-isoparam
Text Generation • 0.8B • Updated • 198 -
Amr-Hegazy/grt-medium-isoflop
Text Generation • 0.4B • Updated • 202
