simonkoson/military-tech-voice3.0
1
0
Fork 0
Code Issues Pull Requests Actions Packages Projects Releases Wiki Activity
15 Commits 1 Branch 0 Tags
effcceb44b4f281008c7823910b8370d2dba1134
Commit Graph

15 Commits

This Branch
This Branch
All Branches
Author SHA1 Message Date
lanhao effcceb44b fix: Fish Audio走东京VPS反向代理，解决国内服务器无法直连问题 
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-07-02 23:58:52 +08:00
lanhao 433903365c feat: 接入 Fish Audio TTS 引擎 
- 新增 fish_audio_service.py，使用 REST API 调用
- voice_id: 34e841f0f4ff44aca6e396debf2776d4（蓝皓音色克隆）
- 支持语速/音量控制，情绪通过内嵌标签实现
- 前端引擎下拉菜单新增 Fish Audio 选项

Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-07-02 23:29:15 +08:00
lanhao 1ad214f5d5 fix: MiniMax音色更换为lanhao_military_v2，使用12.wav重新克隆 
Co-Authored-By: Claude Opus 4.6 <noreply@anthropic.com>
 2026-07-02 22:06:37 +08:00
lanhao c44fa711a7 fix: CosyVoice补全语速/声调/音量参数 
三个引擎的滑块参数现在全部生效

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 21:44:30 +08:00
lanhao dbdb032adb fix: 豆包引擎补全pitch_ratio和volume_ratio参数 
前端声调/音量滑块现在对豆包引擎也生效了

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 21:34:49 +08:00
lanhao ab50ca9157 fix: 豆包API加model_type=4指定ICL 2.0 + 优化音频参数 
之前缺少model_type参数可能回退到ICL 1.0导致音质不如平台调试

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 21:26:34 +08:00
lanhao 54cb8d14ba fix: 豆包音色切换为 S_LK0j18H72 
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 21:17:12 +08:00
lanhao cf3c719c35 feat: 接入豆包 Seed-ICL 2.0 语音合成引擎 
- 新增 doubao_service.py 调用火山引擎 TTS API
- 路由加 doubao 引擎分支
- 前端引擎下拉菜单加豆包选项

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 21:11:23 +08:00
lanhao cb2b37436e fix: admin默认显示名改为刘通 
Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 19:02:28 +08:00
lanhao 223761e717 feat: 登录验证 + 邀请码注册 + 用量监控后台 
- 新增用户认证系统（登录/登出/邀请码注册）
- 新增管理后台（邀请码管理/用户管理/使用记录/用量统计）
- 合成接口加登录验证，每次调用记录用户和稿件内容
- SQLite存储用户数据和使用日志
- 默认admin账号: simonkoson

Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>
 2026-07-02 18:51:23 +08:00
lanhao 3b6019d0f2 fix(frontend): use ENGINE_CONFIG instead of hardcoded engine names 
- Replace ternary engine name checks in generateAudio and downloadAudio
- Use ENGINE_CONFIG[engine].name for consistent engine naming
 2026-07-01 01:27:37 +08:00
lanhao 48f5bf3623 fix(frontend): toolbar layout + engine-select styling 
- Restructure toolbar to use toolbar-row for horizontal layout
- Fix emotion emoji buttons wrapping to vertical
- Enlarge engine-select with better padding, font-size, min-width
- Add toolbar-divider between engine and emotion sections
- Add hover/focus box-shadow to engine-select
- Remove toolbar-hint text to save space
 2026-07-01 01:04:28 +08:00
lanhao a0346d05ba feat(frontend): engine dropdown + emotion dual mode 
- Replace engine buttons with <select> dropdown
- Add ENGINE_CONFIG for centralized engine management
- Add global/per-sentence emotion UI mode switching
- Add #engine-select dark theme styles
- Remove old .switch-btn and .engine-icon styles
- Reset emotion state on engine switch
 2026-07-01 00:48:05 +08:00
lanhao 81460f06d7 feat: 字数限制2000 + MiniMax emotion参数 + 前端情绪全局选择模式 2026-06-30 23:47:23 +08:00
simonkoson d6edc69c65 Initial commit: 军事科技AI配音系统 v1.0 
- Flask backend with MiniMax and CosyVoice TTS engines
- Frontend UI with emotion tags, speed/pitch/volume controls
- Nginx reverse proxy configuration
- Dual engine support (MiniMax speech-2.8-hd, CosyVoice v3.5-flash)
 2026-06-30 21:00:20 +08:00