Question about retraining trimodal gesture models on BEAT #3
Description
Great work! I noticed you compared multiple models by retraining them on english_15_141 and reporting FGD scores. I’m curious about how trimodal gesture generation models were retrained. Could you share details on the process? Thanks