--- license: other license_name: alphacinema-model-pack license_link: https://huggingface.co/alpha-studio/alphacinema-models/blob/main/LICENSES/NOTICE.txt library_name: gguf pipeline_tag: other tags: - text-to-motion - kimodo - llama-3 - blender --- # Alpha Cinema model pack The models Alpha Cinema (a Blender add-on, formerly AlphaMotion) generates body motion with, from text and constraints. The add-on's Download Models fetches this pack at a pinned commit and checks every file against `manifest.json`; its own C++/CUDA cores run them on an NVIDIA GPU. **These are not alpha-studio's models.** They are other people's work, one of them converted here, each under its own licence. Read them in `LICENSES/` before using any file on its own. | File | What | From | Licence | | --- | --- | --- | --- | | `motion/Kimodo-SOMA-RP-v1.1/` | Kimodo motion diffusion model, SOMA skeleton, with its config and statistics | NVIDIA [Kimodo-SOMA-RP-v1.1](https://huggingface.co/nvidia/Kimodo-SOMA-RP-v1.1), unchanged | NVIDIA Open Model License | | `text/Llama-3-Kimodo-AlphaCinema-Q8_0.gguf` | Kimodo's text encoder: LLM2Vec over Llama 3 8B Instruct | LocalAI-io [Llama-3-Kimodo-GGML](https://huggingface.co/LocalAI-io/Llama-3-Kimodo-GGML) BF16, base matrices and token embedding converted to Q8_0 here | Meta Llama 3 Community License; LLM2Vec adapters MIT | | `text/tokenizer.gguf` | Llama 3 tokenizer | LocalAI-io Llama-3-Kimodo-GGML, as published | Meta Llama 3 Community License | | `data/skeletons/` | SOMA skeletons: joints, parents, rest positions | NVIDIA Kimodo's code (`kimodo/assets/skeletons`), the numbers exported to JSON | Apache-2.0 | Kimodo-SOMA-RP-v1.1: Licensed by NVIDIA Corporation under the NVIDIA Open Model License. The NVIDIA Open Model License is in `LICENSES/NVIDIA-Open-Model-License.txt`. The text encoder: Built with Meta Llama 3. Meta Llama 3 is licensed under the Meta Llama 3 Community License, Copyright (c) Meta Platforms, Inc. All Rights Reserved, `LICENSES/LICENSE-META-LLAMA-3.txt`; its use is subject to the Llama 3 Acceptable Use Policy. The GPU runtime (cuBLAS) is not here: the add-on takes it from NVIDIA's own `nvidia-cublas-cu12` wheel on PyPI. `manifest.json` lists every file with its size and SHA-256. alpha-studio is not affiliated with, or endorsed by, NVIDIA, Meta, McGill NLP or LocalAI.