File size: 2,396 Bytes
cd5be23
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
 
1
2
3
4
5
6
7
8
9
10
11
12
13
14
15
16
17
18
19
20
21
22
23
24
25
26
27
28
29
30
31
32
33
34
35
36
37
38
39
40
41
42
43
44
45
46
47
---

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.