Russ Salakhutdinov / @rsalakhu (RSS Feed) on Nostr: Check out our new work on High-Modality Multimodal Transformer -- a single model that ...

Check out our new work on High-Modality Multimodal Transformer -- a single model that scales up to 10 modalities (text, image, audio, video, sensors, proprioception, speech, time-series, sets, tables).

This work also demonstrates a crucial scaling behaviour: performance…

nitter.moomoo.me/pliang279/status/1668673236383707136#m (https://nitter.moomoo.me/pliang279/status/1668673236383707136#m)

https://nitter.moomoo.me/rsalakhu/status/1668894210638970880#m