Vio350M

Vio350M is a small experimental masked discrete diffusion language model.

Training run

Parameters: 386,982,912
Final checkpoint step: 3459
Final trained tokens: 1700167680
Source checkpoint: checkpoints/latest/checkpoint.pt
Architecture: VioForMaskedDiffusionLM
Model type: vio350m_mdlm

This repository contains the exported inference weights from the local/Modal training artifact. Training checkpoints are intentionally not included by default.

Downloads last month: 19

Safetensors

Model size

0.4B params

Tensor type

F32

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support