Skip to content

webui: drop parameter controls the engine never reads #414

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Closed
CryptVenture wants to merge 1 commit into from
Closed

webui: drop parameter controls the engine never reads #414

Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
27 changes: 9 additions & 18 deletions webui/configs/model_params.json
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -5,8 +5,7 @@
{"name": "speaker_guidance_scale", "type": "slider", "label": "speaker_guidance_scale", "label_en": "Speaker guidance", "default": 8.0, "minimum": 0.0, "maximum": 15.0, "step": 0.1},
{"name": "truncation_factor", "type": "slider", "label": "truncation_factor", "label_en": "Noise truncation", "default": 0.8, "minimum": 0.0, "maximum": 1.0, "step": 0.05},
{"name": "guidance_interval", "type": "slider", "label": "guidance_interval", "label_en": "Guidance interval", "default": 1, "minimum": 1, "maximum": 3, "step": 1, "precision": 0, "info": "Refresh the unconditional CFG lanes every Nth guided step. Higher is faster and works best with more steps; 1 is highest fidelity."},
{"name": "reference_duration_sec", "type": "slider", "label": "reference_duration_sec", "label_en": "Reference trim (s)", "default": 15.0, "minimum": 1.0, "maximum": 60.0, "step": 1.0, "info": "Trim the speaker reference before encoding. Around 10 s usually clones best."},
{"name": "seed", "type": "number", "label": "seed", "label_en": "Seed", "default": 0, "minimum": 0, "step": 1, "precision": 0}
{"name": "reference_duration_sec", "type": "slider", "label": "reference_duration_sec", "label_en": "Reference trim (s)", "default": 15.0, "minimum": 1.0, "maximum": 60.0, "step": 1.0, "info": "Trim the speaker reference before encoding. Around 10 s usually clones best."}
],
"_comment": "WebUI TTS 高级参数控件配置：按模型 family 动态生成控件（gr.render）。每项字段：name=选项键(随请求 options 透传给模型)；type=slider|number|bool|text|choice；label/info=显示文案；default=默认值(应等于模型默认，已按 src/models/<family>/*.cpp 校对)；minimum/maximum/step=数值范围；precision=0 表示整数；choices=下拉候选。规则：只有被用户改动过的控件值才会随请求发送；seed/max_tokens 已有专用输入框，勿在此重复；参考文本用『参考文本』框(reference_text)；文件路径/parity 类参数(如 *_noise_file)未纳入，可用『其它参数(JSON)』兜底框传。",

Expand Down Expand Up @@ -50,7 +49,7 @@
{"name": "top_k", "type": "number", "label": "top_k", "default": 50, "minimum": 0, "step": 1, "precision": 0},
{"name": "top_p", "type": "slider", "label": "top_p", "default": 1.0, "minimum": 0.0, "maximum": 1.0, "step": 0.01},
{"name": "repetition_penalty", "type": "slider", "label": "repetition_penalty", "default": 1.0, "minimum": 1.0, "maximum": 1.5, "step": 0.01},
{"name": "best_of_n", "type": "number", "label": "best_of_n（候选数，>1 自动开启）", "default": 1, "minimum": 1, "maximum": 8, "step": 1, "precision": 0}
{"name": "miotts.best_of_n", "type": "number", "label": "miotts.best_of_n（候选数，>1 自动开启）", "label_en": "best_of_n (candidate count; >1 enables best-of-N selection)", "default": 1, "minimum": 1, "maximum": 8, "step": 1, "precision": 0}
],

"chatterbox": [
Expand Down Expand Up @@ -134,17 +133,12 @@
],

"ace_step": [
{"name": "route", "type": "choice", "label": "route（操作类型）", "default": "text2music", "choices": ["text2music", "complete", "lego", "extract", "cover", "cover-nofsq", "repaint", "remix"], "info": "cover/remix=换词翻唱，非 text2music 需上传源音频；详见 webui/README.md"},
{"name": "route", "type": "choice", "label": "route（操作类型）", "label_en": "route (task)", "default": "text2music", "choices": ["text2music", "complete", "lego", "extract", "cover", "cover-nofsq", "repaint"], "info": "cover/cover-nofsq=换词翻唱；非 text2music 需上传源音频；详见 webui/README.md", "info_en": "cover and cover-nofsq re-sing new lyrics. Every route except text2music needs a source audio upload; see webui/README.md."},
{"name": "num_inference_steps", "type": "number", "label": "num_inference_steps", "default": 8, "minimum": 1, "maximum": 20, "step": 1, "precision": 0, "info": "扩散步数（turbo 上限 20）；remix 路由不填时默认 16，其他路由默认 8"},
{"name": "shift", "type": "slider", "label": "shift（时间步弯曲）", "default": 3.0, "minimum": 1.0, "maximum": 5.0, "step": 0.5, "info": "原版 turbo 默认 3.0；1.0 会明显劣化 remix 换词咬字"},
{"name": "guidance_scale", "type": "slider", "label": "guidance_scale", "default": 1.0, "minimum": 0.0, "maximum": 5.0, "step": 0.1},
{"name": "audio_cover_strength", "type": "slider", "label": "【cover】audio_cover_strength", "default": 1.0, "minimum": 0.0, "maximum": 1.0, "step": 0.05, "info": "1=贴近原曲，0=自由发挥；建议 0.5"},
{"name": "cover_noise_strength", "type": "slider", "label": "【cover】cover_noise_strength", "default": 0.0, "minimum": 0.0, "maximum": 1.0, "step": 0.05, "info": "保旋律强度；推荐 0.1~0.25"},
{"name": "source_caption", "type": "text", "label": "【remix】source_caption", "default": "", "placeholder": "源歌曲描述；『🔍 分析』自动填"},
{"name": "source_lyrics", "type": "text", "lines": 4, "label": "【remix】source_lyrics", "default": "", "placeholder": "源歌曲原歌词；『🔍 分析』自动填"},
{"name": "flow_edit_n_min", "type": "slider", "label": "【remix】flow_edit_n_min", "default": 0.0, "minimum": 0.0, "maximum": 1.0, "step": 0.05, "info": "调大更保源曲、换词更弱"},
{"name": "flow_edit_n_max", "type": "slider", "label": "【remix】flow_edit_n_max", "default": 1.0, "minimum": 0.0, "maximum": 1.0, "step": 0.05, "info": "唱不出新歌词时降到 0.7~0.9"},
{"name": "flow_edit_n_avg", "type": "number", "label": "【remix】flow_edit_n_avg", "default": 2, "minimum": 1, "maximum": 4, "step": 1, "precision": 0, "info": "每步多次采样取平均（remix 默认 2）；1=最快"},
{"name": "bpm", "type": "number", "label": "【曲谱】BPM", "default": 0, "minimum": 0, "step": 1, "precision": 0, "info": "0=不指定"},
{"name": "keyscale", "type": "text", "label": "【曲谱】keyscale", "default": "", "placeholder": "如 F major"},
{"name": "timesignature", "type": "text", "label": "【曲谱】timesignature", "default": "", "placeholder": "如 4"}
Expand Down Expand Up @@ -177,9 +171,9 @@
{"name": "route", "type": "choice", "label": "route（转换路径）", "default": "", "choices": ["", "v2_vc", "v1_whisper_bigvgan_vc", "v1_xlsr_hift_vc", "v1_svc"], "info": "留空=按任务默认"},
{"name": "num_inference_steps", "type": "number", "label": "num_inference_steps", "default": 30, "minimum": 1, "step": 1, "precision": 0, "info": "CFM 扩散步数"},
{"name": "length_adjust", "type": "slider", "label": "length_adjust（时长伸缩）", "default": 1.0, "minimum": 0.5, "maximum": 2.0, "step": 0.05},
{"name": "intelligibility_cfg_rate", "type": "slider", "label": "intelligibility_cfg_rate（仅 v2_vc）", "default": 0.7, "minimum": 0.0, "maximum": 1.0, "step": 0.05},
{"name": "similarity_cfg_rate", "type": "slider", "label": "similarity_cfg_rate（仅 v2_vc）", "default": 0.7, "minimum": 0.0, "maximum": 1.0, "step": 0.05},
{"name": "inference_cfg_rate", "type": "slider", "label": "inference_cfg_rate（仅 v1 路径）", "default": 0.7, "minimum": 0.0, "maximum": 1.0, "step": 0.05}
{"name": "intelligibility_guidance_scale", "type": "slider", "label": "intelligibility_guidance_scale（仅 v2_vc）", "label_en": "Intelligibility guidance (v2_vc only)", "default": 0.7, "minimum": 0.0, "maximum": 1.0, "step": 0.05},
{"name": "similarity_guidance_scale", "type": "slider", "label": "similarity_guidance_scale（仅 v2_vc）", "label_en": "Similarity guidance (v2_vc only)", "default": 0.7, "minimum": 0.0, "maximum": 1.0, "step": 0.05},
{"name": "inference_guidance_scale", "type": "slider", "label": "inference_guidance_scale（仅 v1 路径）", "label_en": "Inference guidance (v1 paths only)", "default": 0.7, "minimum": 0.0, "maximum": 1.0, "step": 0.05}
],

"rvc": [
Expand Down Expand Up @@ -233,13 +227,13 @@
],

"index_tts2": [
{"name": "lang", "type": "choice", "label": "lang（语种提示, 仅 IndexTTS2.5 模型）", "label_en": "lang (language hint, IndexTTS2.5 models only)", "default": "auto", "choices": ["auto", "zh", "en", "ja", "es", "ar"], "info": "仅对 IndexTTS2.5（多语种）模型生效：auto 含汉字按中文，否则按英文；日/西/阿建议显式选择", "info_en": "Only applies to IndexTTS2.5 (multilingual) models: auto picks zh when the text contains Han characters, otherwise en; set ja/es/ar explicitly"},
{"name": "emotion_text", "type": "text", "label": "emotion_text（情绪参考文本）", "label_en": "emotion_text (emotion reference text)", "default": "", "placeholder": "例：你吓死我了！你是鬼吗？", "placeholder_en": "e.g. You scared me to death!", "info": "填写后自动开启情感条件（use_emotion_text）", "info_en": "Setting this enables emotion conditioning."},
{"name": "emotion_alpha", "type": "slider", "label": "emotion_alpha（情感强度）", "label_en": "emotion_alpha", "default": 1.0, "minimum": 0.0, "maximum": 1.0, "step": 0.05},
{"name": "use_emotion_text", "type": "bool", "label": "use_emotion_text（从朗读文本推断情感）", "label_en": "use_emotion_text (infer from text)", "default": false},
{"name": "use_random_emotion", "type": "bool", "label": "use_random_emotion（随机情感）", "label_en": "use_random_emotion", "default": false},
{"name": "interval_silence_ms", "type": "number", "label": "interval_silence_ms（分段间静音）", "label_en": "interval_silence_ms", "default": 200, "minimum": 0, "step": 50, "precision": 0},
{"name": "duration_factor", "type": "slider", "label": "duration_factor（语速/时长倍率，>1 更慢，<1 更快）", "label_en": "duration_factor (duration multiplier; >1 slower, <1 faster)", "default": 1.0, "minimum": 0.5, "maximum": 2.0, "step": 0.05, "info": "对齐官方 IndexTTS2.5 的 duration_factor：缩放输出时长，不改变音色/内容", "info_en": "Matches official IndexTTS2.5 duration_factor: scales output duration without changing timbre or content"}
{"name": "duration_factor", "type": "slider", "label": "duration_factor（语速/时长倍率，>1 更慢，<1 更快）", "label_en": "duration_factor (duration multiplier; >1 slower, <1 faster)", "default": 1.0, "minimum": 0.5, "maximum": 2.0, "step": 0.05, "info": "对齐官方 IndexTTS2.5 的 duration_factor：缩放输出时长，不改变音色/内容", "info_en": "Matches official IndexTTS2.5 duration_factor: scales output duration without changing timbre or content"},
{"name": "language", "type": "choice", "label": "language（语种提示, 仅 IndexTTS2.5 模型）", "label_en": "language (language hint, IndexTTS2.5 models only)", "default": "auto", "choices": ["auto", "zh", "en", "ja", "es", "ar"], "info": "仅对 IndexTTS2.5（多语种）模型生效：auto 含汉字按中文，否则按英文；日/西/阿建议显式选择", "info_en": "Only applies to IndexTTS2.5 (multilingual) models: auto picks zh when the text contains Han characters, otherwise en; set ja/es/ar explicitly"}
],

"irodori_tts": [
Expand Down Expand Up @@ -317,10 +311,7 @@
{"name": "language", "type": "choice", "label": "language", "default": "zh", "choices": ["zh", "en"]},
{"name": "num_inference_steps", "type": "number", "label": "num_inference_steps", "default": 10, "minimum": 1, "step": 1, "precision": 0},
{"name": "guidance_scale", "type": "slider", "label": "guidance_scale", "default": 2.0, "minimum": 0.0, "maximum": 10.0, "step": 0.1},
{"name": "max_new_audio_steps", "type": "number", "label": "max_new_audio_steps", "default": 750, "minimum": 1, "step": 1, "precision": 0},
{"name": "top_k", "type": "number", "label": "top_k", "default": 20, "minimum": 0, "step": 1, "precision": 0},
{"name": "top_p", "type": "slider", "label": "top_p", "default": 0.8, "minimum": 0.0, "maximum": 1.0, "step": 0.01},
{"name": "temperature", "type": "slider", "label": "temperature", "default": 0.7, "minimum": 0.0, "maximum": 2.0, "step": 0.05}
{"name": "max_new_audio_steps", "type": "number", "label": "max_new_audio_steps", "default": 750, "minimum": 1, "step": 1, "precision": 0}
],

"firered-audio-vdesign": [
Expand Down
Loading