Sync model card with upstream GitHub inference README
Browse files
README.md
CHANGED
|
@@ -12,7 +12,7 @@ license: apache-2.0
|
|
| 12 |
# LocalVQE
|
| 13 |
|
| 14 |
[](https://huggingface.co/spaces/LocalAI-io/LocalVQE-demo)
|
| 15 |
-
[](https://www.apache.org/licenses/LICENSE-2.0)
|
| 17 |
|
| 18 |
**Local Voice Quality Enhancement** — a compact neural model for joint
|
|
@@ -29,7 +29,7 @@ Try it live: <https://huggingface.co/spaces/LocalAI-io/LocalVQE-demo>.
|
|
| 29 |
|
| 30 |
This page is the Hugging Face model card — it hosts the published weights.
|
| 31 |
Source code, build system, tests, and training pipeline live in the GitHub
|
| 32 |
-
repository: <https://github.com/
|
| 33 |
|
| 34 |
The current release is **v1.1**, which fixes intermittent crackling the
|
| 35 |
previous release produced under heavy background noise.
|
|
@@ -135,11 +135,11 @@ Full 800-clip eval on the
|
|
| 135 |
## Building the C++ Inference Engine
|
| 136 |
|
| 137 |
Source, build system, and tests live at
|
| 138 |
-
<https://github.com/
|
| 139 |
compiler. A [Nix](https://nixos.org/) flake is provided:
|
| 140 |
|
| 141 |
```bash
|
| 142 |
-
git clone --recursive https://github.com/
|
| 143 |
cd LocalVQE
|
| 144 |
|
| 145 |
# With Nix:
|
|
@@ -245,10 +245,10 @@ calibration have been worked through.
|
|
| 245 |
It is provided for verification, ablation, and downstream research — not
|
| 246 |
for end-user inference, which should go through the GGML build above. The
|
| 247 |
model definition lives under `pytorch/` in the
|
| 248 |
-
[GitHub repo](https://github.com/
|
| 249 |
|
| 250 |
```bash
|
| 251 |
-
git clone https://github.com/
|
| 252 |
cd LocalVQE/pytorch
|
| 253 |
pip install -r requirements.txt
|
| 254 |
```
|
|
@@ -256,7 +256,7 @@ pip install -r requirements.txt
|
|
| 256 |
## Citing LocalVQE
|
| 257 |
|
| 258 |
If you use LocalVQE in academic work, please cite the repository via the
|
| 259 |
-
`CITATION.cff` at <https://github.com/
|
| 260 |
a "Cite this repository" button that produces APA and BibTeX entries
|
| 261 |
automatically.
|
| 262 |
|
|
|
|
| 12 |
# LocalVQE
|
| 13 |
|
| 14 |
[](https://huggingface.co/spaces/LocalAI-io/LocalVQE-demo)
|
| 15 |
+
[](https://github.com/localai-org/LocalVQE)
|
| 16 |
[](https://www.apache.org/licenses/LICENSE-2.0)
|
| 17 |
|
| 18 |
**Local Voice Quality Enhancement** — a compact neural model for joint
|
|
|
|
| 29 |
|
| 30 |
This page is the Hugging Face model card — it hosts the published weights.
|
| 31 |
Source code, build system, tests, and training pipeline live in the GitHub
|
| 32 |
+
repository: <https://github.com/localai-org/LocalVQE>.
|
| 33 |
|
| 34 |
The current release is **v1.1**, which fixes intermittent crackling the
|
| 35 |
previous release produced under heavy background noise.
|
|
|
|
| 135 |
## Building the C++ Inference Engine
|
| 136 |
|
| 137 |
Source, build system, and tests live at
|
| 138 |
+
<https://github.com/localai-org/LocalVQE>. Requires CMake ≥ 3.20 and a C++17
|
| 139 |
compiler. A [Nix](https://nixos.org/) flake is provided:
|
| 140 |
|
| 141 |
```bash
|
| 142 |
+
git clone --recursive https://github.com/localai-org/LocalVQE.git
|
| 143 |
cd LocalVQE
|
| 144 |
|
| 145 |
# With Nix:
|
|
|
|
| 245 |
It is provided for verification, ablation, and downstream research — not
|
| 246 |
for end-user inference, which should go through the GGML build above. The
|
| 247 |
model definition lives under `pytorch/` in the
|
| 248 |
+
[GitHub repo](https://github.com/localai-org/LocalVQE):
|
| 249 |
|
| 250 |
```bash
|
| 251 |
+
git clone https://github.com/localai-org/LocalVQE.git
|
| 252 |
cd LocalVQE/pytorch
|
| 253 |
pip install -r requirements.txt
|
| 254 |
```
|
|
|
|
| 256 |
## Citing LocalVQE
|
| 257 |
|
| 258 |
If you use LocalVQE in academic work, please cite the repository via the
|
| 259 |
+
`CITATION.cff` at <https://github.com/localai-org/LocalVQE> — GitHub renders
|
| 260 |
a "Cite this repository" button that produces APA and BibTeX entries
|
| 261 |
automatically.
|
| 262 |
|