fastapi 基于硅基流模型文字转语音保存至腾讯OSS

最新推荐文章于 2025-06-04 16:29:08 发布

weixin_41865469

最新推荐文章于 2025-06-04 16:29:08 发布

阅读量1.1k

点赞数 7

文章标签： fastapi python AI 文字转语音腾讯云OSS存储硅基流TTS

本文链接：https://blog.youkuaiyun.com/weixin_41865469/article/details/145488290

版权

硅基流
https://cloud.siliconflow.cn/account/ak
这里我们用到了TTS语音模型
在这里插入图片描述

腾讯云oss
https://console.cloud.tencent.com/
在这里插入图片描述

from fastapi import FastAPI, HTTPException, Query
import json
import requests
import time
from pydantic import BaseModel, ValidationError
import datetime
import random
import os
from qcloud_cos import CosConfig
from qcloud_cos import CosS3Client
app = FastAPI()
import os
# print("当前脚本路径:", os.path.abspath(__file__))
# print("工作目录:", os.getcwd())

class VideoSubmission(BaseModel):
    prompt: str
    model: str


class AudioSubmission(BaseModel):
    input: str
    model: str
    voice: str

def load_config(config_file):
    base_path = os.getcwd()  # 获取当前工作目录
    full_path = os.path.join(base_path, "tool_cal", config_file)  # 正确拼接路径
    with open(full_path, 'r', encoding='utf-8') as file:
        return json.load(file)

config = load_config('config.json')
api_url = config["api_url"]
auth_token = config["authorization_token"]
region = config["region"]
secret_id = config["secret_id"]
secret_key = config["secret_key"]
bucket = config["bucket"]
output_path = config["output_path"]
audiomodel = config["audiomodel"]
voice=config["voice"]

def submit_video_job(api_url, auth_token, model, prompt):
    submit_url = f"{
     api_url}/video/submit"
    payload = json.dumps({
   
        "model": model,
        "prompt": prompt