mirror of
https://github.com/linyqh/NarratoAI.git
synced 2026-01-04 04:08:12 +00:00
062d317261
- 新增Azure Speech Services和Edge TTS引擎支持 - 重构语音设置界面，支持不同引擎的独立配置 - 添加引擎选择器和详细说明 - 更新requirements.txt添加azure-cognitiveservices-speech依赖 - 改进音色名称验证逻辑
122 lines
3.9 KiB
TOML
122 lines
3.9 KiB
TOML
[app]
|
||
project_version="0.6.8"
|
||
# 支持视频理解的大模型提供商
|
||
# gemini (谷歌, 需要 VPN)
|
||
# siliconflow (硅基流动)
|
||
# qwenvl (通义千问)
|
||
vision_llm_provider="gemini"
|
||
|
||
########## Gemini 视觉模型
|
||
vision_gemini_api_key = ""
|
||
vision_gemini_model_name = "gemini-2.0-flash-lite"
|
||
|
||
########## QwenVL 视觉模型
|
||
vision_qwenvl_api_key = ""
|
||
vision_qwenvl_model_name = "qwen2.5-vl-32b-instruct"
|
||
vision_qwenvl_base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1"
|
||
|
||
########## siliconflow 视觉模型
|
||
vision_siliconflow_api_key = ""
|
||
vision_siliconflow_model_name = "Qwen/Qwen2.5-VL-32B-Instruct"
|
||
vision_siliconflow_base_url = "https://api.siliconflow.cn/v1"
|
||
|
||
########## OpenAI 视觉模型
|
||
vision_openai_api_key = ""
|
||
vision_openai_model_name = "gpt-4.1-nano-2025-04-14"
|
||
vision_openai_base_url = "https://api.openai.com/v1"
|
||
|
||
########### NarratoAPI 微调模型 (未发布)
|
||
narrato_api_key = ""
|
||
narrato_api_url = ""
|
||
narrato_model = "narra-1.0-2025-05-09"
|
||
|
||
# 用于生成文案的大模型支持的提供商 (Supported providers):
|
||
# openai (默认, 需要 VPN)
|
||
# siliconflow (硅基流动)
|
||
# deepseek (深度求索)
|
||
# gemini (谷歌, 需要 VPN)
|
||
# qwen (通义千问)
|
||
# moonshot (月之暗面)
|
||
text_llm_provider="gemini"
|
||
|
||
########## OpenAI API Key
|
||
# Get your API key at https://platform.openai.com/api-keys
|
||
text_openai_api_key = ""
|
||
text_openai_base_url = "https://api.openai.com/v1"
|
||
text_openai_model_name = "gpt-4.1-mini-2025-04-14"
|
||
|
||
# 使用 硅基流动 第三方 API Key，使用手机号注册：https://cloud.siliconflow.cn/i/pyOKqFCV
|
||
# 访问 https://cloud.siliconflow.cn/account/ak 获取你的 API 密钥
|
||
text_siliconflow_api_key = ""
|
||
text_siliconflow_base_url = "https://api.siliconflow.cn/v1"
|
||
text_siliconflow_model_name = "deepseek-ai/DeepSeek-R1"
|
||
|
||
########## DeepSeek API Key
|
||
# 访问 https://platform.deepseek.com/api_keys 获取你的 API 密钥
|
||
text_deepseek_api_key = ""
|
||
text_deepseek_base_url = "https://api.deepseek.com"
|
||
text_deepseek_model_name = "deepseek-chat"
|
||
|
||
########## Gemini API Key
|
||
text_gemini_api_key=""
|
||
text_gemini_model_name = "gemini-2.0-flash"
|
||
text_gemini_base_url = "https://generativelanguage.googleapis.com/v1beta"
|
||
|
||
########## Qwen API Key
|
||
# 访问 https://bailian.console.aliyun.com/?tab=model#/api-key 获取你的 API 密钥
|
||
text_qwen_api_key = ""
|
||
text_qwen_model_name = "qwen-plus-1127"
|
||
text_qwen_base_url = "https://dashscope.aliyuncs.com/compatible-mode/v1"
|
||
|
||
########## Moonshot API Key
|
||
# 访问 https://platform.moonshot.cn/console/api-keys 获取你的 API 密钥
|
||
text_moonshot_api_key=""
|
||
text_moonshot_base_url = "https://api.moonshot.cn/v1"
|
||
text_moonshot_model_name = "moonshot-v1-8k"
|
||
|
||
# webui界面是否显示配置项
|
||
hide_config = true
|
||
|
||
[azure]
|
||
# Azure TTS 配置
|
||
speech_key = ""
|
||
speech_region = ""
|
||
|
||
[soulvoice]
|
||
# SoulVoice TTS API 密钥
|
||
api_key = ""
|
||
# 音色 URI（必需）
|
||
voice_uri = "speech:mcg3fdnx:clzkyf4vy00e5qr6hywum4u84:bzznlkuhcjzpbosexitr"
|
||
# API 接口地址（可选，默认值如下）
|
||
api_url = "https://tts.scsmtech.cn/tts"
|
||
# 默认模型（可选）
|
||
model = "FunAudioLLM/CosyVoice2-0.5B"
|
||
|
||
[ui]
|
||
# TTS引擎选择 (edge_tts, azure_speech, soulvoice)
|
||
tts_engine = "edge_tts"
|
||
|
||
# Edge TTS 配置
|
||
edge_voice_name = "zh-CN-XiaoyiNeural-Female"
|
||
edge_volume = 80
|
||
edge_rate = 1.0
|
||
edge_pitch = 0
|
||
|
||
# Azure Speech Services 配置
|
||
azure_voice_name = "zh-CN-XiaoyiNeural-Female"
|
||
azure_volume = 80
|
||
azure_rate = 1.0
|
||
azure_pitch = 0
|
||
|
||
[proxy]
|
||
# clash 默认地址：http://127.0.0.1:7890
|
||
http = ""
|
||
https = ""
|
||
enabled = false
|
||
|
||
[frames]
|
||
# 提取关键帧的间隔时间
|
||
frame_interval_input = 3
|
||
# 大模型单次处理的关键帧数量
|
||
vision_batch_size = 10
|