metadata
license: apache-2.0
library_name: eva01
pipeline_tag: text-to-3d
EVA01 Unified All-Task Generation
This repository contains the generation-only adapter for the EVA01 Unified MoT runtime. It supports text-to-3D and image-to-3D inference through the native SS -> Shape -> Material cascade.
The adapter contains only EVA01 generation parameters. Frozen Qwen3-VL, TRELLIS, TRELLIS.2, tokenizer, normalization, and decoder assets must be supplied separately as described in the OpenEVA repository.
This repository does not include training code, training data, edit parameters, decoder weights, or base model weights.