Upload README.md with huggingface_hub
Browse files
README.md
CHANGED
|
@@ -18,6 +18,8 @@ base_model:
|
|
| 18 |
|
| 19 |
High-quality multi-speaker TTS, 7B Qwen2 backbone. INT8/4 LM.
|
| 20 |
|
|
|
|
|
|
|
| 21 |
**On-device performance (M4 Max, Core AI):** 2.37× RTF.
|
| 22 |
|
| 23 |
> ⚠️ **Beta artifacts.** These `.aimodel` bundles are compiled for macOS 27 / Xcode 27 beta (Core AI). They may need re-export on the GA toolchain. The original weights are Microsoft VibeVoice (see upstream for the model license).
|
|
@@ -84,4 +86,4 @@ Resolve assets by role via `manifest.json` (`default` = recommended variant):
|
|
| 84 |
"decode_compute": "gpu",
|
| 85 |
"sem_compute": "gpu"
|
| 86 |
}
|
| 87 |
-
```
|
|
|
|
| 18 |
|
| 19 |
High-quality multi-speaker TTS, 7B Qwen2 backbone. INT8/4 LM.
|
| 20 |
|
| 21 |
+
**Source & export pipeline:** [github.com/gafiatulin/vibevoice-coreai](https://github.com/gafiatulin/vibevoice-coreai)
|
| 22 |
+
|
| 23 |
**On-device performance (M4 Max, Core AI):** 2.37× RTF.
|
| 24 |
|
| 25 |
> ⚠️ **Beta artifacts.** These `.aimodel` bundles are compiled for macOS 27 / Xcode 27 beta (Core AI). They may need re-export on the GA toolchain. The original weights are Microsoft VibeVoice (see upstream for the model license).
|
|
|
|
| 86 |
"decode_compute": "gpu",
|
| 87 |
"sem_compute": "gpu"
|
| 88 |
}
|
| 89 |
+
```
|