MoE Diffusion Transformer for non-autoregressive text generation, powered natively by Rust and Burn 🔥.
This repo goals is experiments of applying exisiting MoE models to dLLM design and inspired by LLaDa MoE v2 article
by @getStRiCtd
| Name | Name | Last commit date | ||
|---|---|---|---|---|
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
 |  | |||
MoE Diffusion Transformer for non-autoregressive text generation, powered natively by Rust and Burn 🔥.
This repo goals is experiments of applying exisiting MoE models to dLLM design and inspired by LLaDa MoE v2 article
by @getStRiCtd