Mixture-of-experts models: many experts per layer, only a few active per token.
-
nkthebass/tinybrainbot-61m-a10m-v4.1-moe-base
Text Generation • 64.2M • Updated • 64 -
nkthebass/tinybrainbot-360m-a75m-v4.1-moe-base
Text Generation • 0.4B • Updated • 394 -
nkthebass/tinybrainbot-360m-a75m-v4.1-moe-instruct
Text Generation • 0.4B • Updated • 100 -
TinyBrainBot 360M MoE Chat
🧠Chat with a 383M MoE (75M active) in your browser