| --- |
| license: other |
| license_name: per-file-upstream-licenses |
| license_link: LICENSE |
| tags: |
| - opencc |
| - automatic-speech-recognition |
| - translation |
| - model-resources |
| --- |
| |
| # OpenCC Models |
|
|
| Model resource distribution for OpenCC, AutoSubs Ultra and Subtitle Edit Ultra. |
| This repository is independent of the OpenStems / StemDeck model repository. |
|
|
| ## Publication status |
|
|
| Resource inventory and publication are in progress. Creating this repository is |
| not a statement that model files or the applications' public download integration |
| are ready. Production applications must only use a published, revision-pinned |
| resource manifest whose exact file lengths and SHA-256 digests have been verified. |
|
|
| Public upstream models may be obtained directly from verified public upstream |
| repositories. OpenCC-converted or packed artifacts retain their original model |
| identity, format and precision. Conversion and packaging are not model training. |
| Each published artifact must identify its upstream source, transformation and |
| applicable upstream terms. No repository-wide license replaces those terms. |
|
|
| HF and ModelScope origins for one logical artifact must provide identical final |
| file bytes. Similar names, model families or quantization labels are insufficient. |
|
|
| ## 中文 |
|
|
| 本仓库用于 OpenCC 及两个字幕应用的模型资源分发，与 StemDeck 仓库独立。 |
| 当前资源盘点和发布尚未完成，不能将空仓或示例地址当作可用下载源。 |
| 每个正式资源应固定版本、大小、SHA-256、来源、转换过程、精度和原许可。 |
| 两站回退必须提供相同的最终文件；格式或精度不同的模型不能互为镜像。 |
|
|
