Add model card: runtime builds + mirrored ACE-Step 1.5 GGUF weights
Browse files
README.md
ADDED
|
@@ -0,0 +1,62 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 1 |
+
---
|
| 2 |
+
license: mit
|
| 3 |
+
base_model:
|
| 4 |
+
- ACE-Step/Ace-Step1.5
|
| 5 |
+
tags:
|
| 6 |
+
- gguf
|
| 7 |
+
- text-to-music
|
| 8 |
+
- acestep
|
| 9 |
+
---
|
| 10 |
+
|
| 11 |
+
# acestep — runtime + GGUF weights mirror
|
| 12 |
+
|
| 13 |
+
Files needed by the local music generation feature of the KakaoWork desktop app.
|
| 14 |
+
This repository is a **mirror kept for build reproducibility**: the app pins a
|
| 15 |
+
sha256 for every file, so it needs a location whose bytes never change.
|
| 16 |
+
|
| 17 |
+
## Contents
|
| 18 |
+
|
| 19 |
+
### Runtime (built by us)
|
| 20 |
+
|
| 21 |
+
`acestep.cpp` has no upstream release artifacts, so we build it ourselves.
|
| 22 |
+
|
| 23 |
+
| File | Platform |
|
| 24 |
+
| --- | --- |
|
| 25 |
+
| `acestep-darwin-arm64.zip` | macOS arm64 (ad-hoc signed) |
|
| 26 |
+
| `acestep-win-x64.zip` | Windows x64 (signed) |
|
| 27 |
+
|
| 28 |
+
Each archive is flat: `ace-lm`, `ace-synth` and the ggml shared libraries sit at
|
| 29 |
+
the top level. The executables resolve the libraries next to them, so keep them
|
| 30 |
+
in one directory.
|
| 31 |
+
|
| 32 |
+
### Weights (mirrored, byte-identical)
|
| 33 |
+
|
| 34 |
+
Re-uploaded unchanged from
|
| 35 |
+
[Serveurperso/ACE-Step-1.5-GGUF](https://huggingface.co/Serveurperso/ACE-Step-1.5-GGUF).
|
| 36 |
+
Only the four files our app loads are mirrored; that repository has many more
|
| 37 |
+
quantizations and is the place to look for them.
|
| 38 |
+
|
| 39 |
+
| File | Size | sha256 |
|
| 40 |
+
| --- | --- | --- |
|
| 41 |
+
| `acestep-v15-turbo-Q4_K_M.gguf` | 1.45 GB | `55b4d8514850f3d0f82536f37e99673aaf48df802b5ae5b153eea32a2e2daa5e` |
|
| 42 |
+
| `acestep-5Hz-lm-4B-Q5_K_M.gguf` | 3.03 GB | `938ed7067c8897f66acf4c3a86fc1fa8113d5cd1a5f13e6edec2e03207514e2d` |
|
| 43 |
+
| `Qwen3-Embedding-0.6B-Q8_0.gguf` | 784 MB | `972f23255e46adfe744a0eb9a0039f3c63988f65753b0968d776e8b27168c321` |
|
| 44 |
+
| `vae-BF16.gguf` | 337 MB | `0599862ac5d15cd308e1d2e368373aea6c02e25ebd1737ad4a4562a0901b0ef8` |
|
| 45 |
+
|
| 46 |
+
The sha256 values are the upstream ones — verify against the source repository
|
| 47 |
+
if you want to confirm nothing was altered in transit.
|
| 48 |
+
|
| 49 |
+
## Credits and license
|
| 50 |
+
|
| 51 |
+
- Original model weights: **ACE Studio** and **StepFun** —
|
| 52 |
+
[ACE-Step/Ace-Step1.5](https://huggingface.co/ACE-Step/Ace-Step1.5). All model
|
| 53 |
+
weights are theirs; we changed nothing.
|
| 54 |
+
- GGUF conversion and quantization: **Serveurperso** —
|
| 55 |
+
[Serveurperso/ACE-Step-1.5-GGUF](https://huggingface.co/Serveurperso/ACE-Step-1.5-GGUF)
|
| 56 |
+
(MIT), whose model card also credits Aryanne and jeankassio for specific
|
| 57 |
+
weight merges.
|
| 58 |
+
- The mirrored weights are redistributed under the MIT license of the source
|
| 59 |
+
repository. Please refer to the upstream repositories for the terms that apply
|
| 60 |
+
to the original model.
|
| 61 |
+
- The runtime archives are builds of
|
| 62 |
+
[acestep.cpp](https://github.com/ACE-Step/acestep.cpp); its own license applies.
|