MiMo-V2.6-Flash-RL-UNCENSORED — standalone DFlash draft head
Draft head extracted verbatim from dealignai/MiMo-V2.6-Flash-RL-UNCENSORED (dflash/ subfolder) so vLLM can load it as a standalone speculative model (SpeculativeConfig.model requires a repo/dir with config.json at root; bundled subfolders are not discoverable).
Upstream: XiaomiMiMo/MiMo-V2.6-Flash-RL (MIT), uncensored edition by dealignai (MIT). No weight modifications.
- Downloads last month
- 22
Inference Providers NEW
This model isn't deployed by any Inference Provider. 🙋 Ask for provider support