FrontierModels/fm-render

LTX-2 + MOSS-TTSD environment

A generic open-source ML environment image for running LTX-2 video generation and MOSS-TTSD text-to-speech on RTX PRO 6000 (Blackwell) hardware.

Dual Python environment:

• —System Python — torch 2.9.1+cu130, transformers 4.57.6, ltx-pipelines, no xformers.
• —MOSS-TTSD venv (/opt/moss-venv) — --system-site-packages, transformers 5.0.0.

The image contains no application code or data — it is meant to be used as a runtime environment with the workload script and model weights supplied via a mounted volume.