dsh-podwave 0.10.15 → 0.10.20
This diff represents the content of publicly available package versions that have been released to one of the supported registries. The information contained in this diff is provided for informational purposes only and reflects changes between package versions as they appear in their respective public registries.
- package/README.md +1 -0
- package/dist/client.js +117 -69
- package/package.json +2 -2
- package/scripts/asr_sherpa_batch.py +3 -0
- package/src/downloader.js +101 -4
- package/src/host.js +49 -7
- package/src/setup.js +15 -5
- package/src/transcribe.js +104 -16
- package/src/translate.js +29 -1
package/README.md
CHANGED
|
@@ -180,4 +180,5 @@ CI(GitHub Actions)在每次 push / PR 跑 `npm test` 与 `npm run build`。
|
|
|
180
180
|
|
|
181
181
|
- 安装中心会自动下载 ffmpeg(约 202 MB)、llama.cpp(约 25 MB)、SenseVoice(约 239 MB)与 Hy-MT2-Q4_K_M(约 1.1 GB),合计约 1.6 GB,建议预留 3 GB 磁盘。
|
|
182
182
|
- Windows 10 及以上自带 `tar` 与 `powershell`,解压和文件夹选择器无需额外安装。
|
|
183
|
+
- 安装时会校验下载文件的 sha256(GitHub Release 官方 digest / Hugging Face LFS),并拒绝含 `..` 或绝对路径的压缩包;macOS 没有 ffmpeg 时会尝试通过 Homebrew 安装。
|
|
183
184
|
- 首次安装完成后,设置页会列出每项下载内容与体积;模型在 `~/.podwave/models`,运行时在 `~/.podwave/runtime`。
|