机场推荐官
共计 12695 个字符，预计需要花费 32 分钟才能阅读完成。
注册帐户
1. 我们需要在 Prodiap 32 注册一个账户, 并获取 PRODIA_API_KEY：
创建 Cloudflare Workers
2. 在 Cloud Flare 9 中的 Workers 和 Pages 中创建一个 Worker 项目并点击编辑代码：
编辑 Cloudflare Workers
3. 复制下方的 js 项目粘贴至 worker 中 修改相应参数后部署：(如何获取 cf_accountID 以及 cf_token 请自行查阅学习)
API_KEY 可自定义比如: sk-123456789
//prodia.com API_KEY
const PRODIA_API_KEY = "XXX";
// 本项目授权 api_key，防止被恶意调用
const API_KEY = "sk-XXX";
//cloudflare 账号列表，每次请求都会随机从列表里取一个账号
const CF_ACCOUNT_LIST = [
{account_id: "cf_accountID", token: "cf_token"}
];
// 在你输入的 prompt 中添加 ---ntl 可强制禁止提示词翻译、优化功能
// 在你输入的 prompt 中添加 ---tl 可强制开启提示词翻译、优化功能
// 是否开启提示词翻译、优化功能
const CF_IS_TRANSLATE = true;
// 示词翻译、优化模型
const CF_TRANSLATE_MODEL = "@cf/qwen/qwen1.5-14b-chat-awq";
// 模型映射，设置客户端可用的模型。one-api，new-api 在添加渠道时可使用 "获取模型列表" 功能，一键添加模型
const CUSTOMER_MODEL_MAP = { "animagineXLV3_v30.safetensors": "animagineXLV3_v30.safetensors [75f2f05b]",
"devlishphotorealism_sdxl15.safetensors": "devlishphotorealism_sdxl15.safetensors [77cba69f]",
"dreamshaperXL10_alpha2.safetensors": "dreamshaperXL10_alpha2.safetensors [c8afe2ef]",
"dynavisionXL_0411.safetensors": "dynavisionXL_0411.safetensors [c39cc051]",
"juggernautXL_v45.safetensors": "juggernautXL_v45.safetensors [e75f5471]",
"realismEngineSDXL_v10.safetensors": "realismEngineSDXL_v10.safetensors [af771c3f]",
"realvisxlV40.safetensors": "realvisxlV40.safetensors [f7fdcb51]",
"sd_xl_base_1.0.safetensors": "sd_xl_base_1.0.safetensors [be9edd61]",
"sd_xl_base_1.0_inpainting_0.1.safetensors": "sd_xl_base_1.0_inpainting_0.1.safetensors [5679a81a]",
"turbovisionXL_v431.safetensors": "turbovisionXL_v431.safetensors [78890989]"
};
/**
* Handles incoming requests to the Cloudflare Worker.
* @param {Request} request - The incoming request object.
* @returns {Response} - The response object.
* @throws {Error} - If the request is invalid or the response fails.
*/
async function handleRequest(request) {
try { if (request.method === "OPTIONS") {
return new Response("", {
status: 204,
headers: {
'Access-Control-Allow-Origin': '*',
"Access-Control-Allow-Headers": '*'
}
});
}
const authHeader = request.headers.get("Authorization");
if (!authHeader || !authHeader.startsWith("Bearer") || authHeader.split(" ")[1] !== API_KEY) { return new Response("Unauthorized", { status: 401});
}
if (request.url.endsWith("/v1/models")) { const arrs = [];
Object.keys(CUSTOMER_MODEL_MAP).map(element => arrs.push({ id: element, object: "model"}))
const response = {
data: arrs,
success: true
};
return new Response(JSON.stringify(response), {
headers: {
'Content-Type': 'application/json',
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Headers': '*'
}
});
}
if (request.method !== "POST") {
return new Response("Only POST requests are allowed", {
status: 405,
headers: {
'Access-Control-Allow-Origin': '*',
"Access-Control-Allow-Headers": '*'
}
});
}
if (!request.url.endsWith("/v1/chat/completions")) {
return new Response("Not Found", {
status: 404,
headers: {
'Access-Control-Allow-Origin': '*',
"Access-Control-Allow-Headers": '*'
}
});
}
const data = await request.json();
const messages = data.messages || [];
const model = CUSTOMER_MODEL_MAP[data.model] || CUSTOMER_MODEL_MAP["v1-5-inpainting.safetensors"];
const stream = data.stream || false;
const userMessage = messages.reverse().find((msg) => msg.role === "user")?.content;
if (!userMessage) { return new Response(JSON.stringify({ error: "未找到用户消息"}), {
status: 400,
headers: {
'Content-Type': 'application/json',
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Headers': '*'
}
});
}
const is_translate = extractTranslate(userMessage);
const originalPrompt = cleanPromptString(userMessage);
const translatedPrompt = is_translate ? await getPrompt(originalPrompt) : originalPrompt;
const imageUrl = await generateImageByText(model, translatedPrompt);
if (stream) { return handleStreamResponse(originalPrompt, translatedPrompt,"1024x1024", model, imageUrl);
} else { return handleNonStreamResponse(originalPrompt, translatedPrompt, "1024x1024", model, imageUrl);
}
} catch (error) {
return new Response("Internal Server Error:" + error.message, {
status: 500,
headers: {
'Content-Type': 'application/json',
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Headers': '*'
}
});
}
}
/**
* @description
* Translate a prompt into a stable diffusion prompt style.
* @param {string} prompt - The prompt to translate.
* @returns {Promise} The translated prompt.
* @throws {Error} If the translation fails.
*/
async function getPrompt(prompt) {
const requestBodyJson = {
messages: [
{
role: "system",
content: ` 作为 Stable Diffusion Prompt 提示词专家，您将从关键词中创建提示，通常来自 Danbooru 等数据库。
提示通常描述图像，使用常见词汇，按重要性排列，并用逗号分隔。避免使用 "-" 或 "."，但可以接受空格和自然语言。避免词汇重复。
为了强调关键词，请将其放在括号中以增加其权重。例如，"(flowers)" 将 'flowers' 的权重增加 1.1 倍，而 "(((flowers)))" 将其增加 1.331 倍。使用 "(flowers:1.5)" 将 'flowers' 的权重增加 1.5 倍。只为重要的标签增加权重。
提示包括三个部分：** 前缀 **（质量标签 + 风格词 + 效果器）+ ** 主题 **（图像的主要焦点）+ ** 场景 **（背景、环境）。
* 前缀影响图像质量。像 "masterpiece"、"best quality"、"4k" 这样的标签可以提高图像的细节。像 "illustration"、"lensflare" 这样的风格词定义图像的风格。像 "bestlighting"、"lensflare"、"depthoffield" 这样的效果器会影响光照和深度。
* 主题是图像的主要焦点，如角色或场景。对主题进行详细描述可以确保图像丰富而详细。增加主题的权重以增强其清晰度。对于角色，描述面部、头发、身体、服装、姿势等特征。
* 场景描述环境。没有场景，图像的背景是平淡的，主题显得过大。某些主题本身包含场景（例如建筑物、风景）。像 "花草草地"、"阳光"、"河流" 这样的环境词可以丰富场景。你的任务是设计图像生成的提示。请按照以下步骤进行操作：
1. 我会发送给您一个图像场景。需要你生成详细的图像描述
2. 图像描述必须是英文，输出为 Positive Prompt。
示例：
我发送：二战时期的护士。 您回复只回复： A WWII-era nurse in a German uniform, holding a wine bottle and stethoscope, sitting at a table in white attire, with a table in the background, masterpiece, best quality, 4k, illustration style, best lighting, depth of field, detailed character, detailed environment.`
},
{
role: "user",
content: prompt
}
]
};
const response = await postRequest(CF_TRANSLATE_MODEL, requestBodyJson);
if (!response.ok) { return prompt;}
const jsonResponse = await response.json();
const res = jsonResponse.result.response;
return res;
}
/**
* Generate an image from a given text prompt using the Prodia AI API
* @param {string} model - The name of the AI model to use for image generation
* @param {string} prompt - The text prompt to generate an image from
* @returns {string} - The URL of the generated image
* @throws {Error} - If the image generation fails
* @see https://docs.prodia.ai/docs/api-reference
*/
async function generateImageByText(model, prompt) {
// First request to generate the image
const generateOptions = {
method: 'POST',
headers: {
accept: 'application/json',
'content-type': 'application/json',
'X-Prodia-Key': PRODIA_API_KEY
},
body: JSON.stringify({
model: model,
prompt: prompt,
negative_prompt: 'low resolution, blurry, distorted features, wrong fingers, extra numbers, watermarks, ugly, distorted, deformed, deformed, repetitive, missing arms and legs, multiple hands and legs, incomplete limbs, long neck, cross-eyed, glazed eyes, lax eyes, squinting, deformed eyes',
steps: 20,
cfg_scale: 7,
seed: -1,
sampler: 'DPM++ 2M Karras',
width: 1024,
height: 1024
})
};
try { const generateResponse = await fetch('https://api.prodia.com/v1/sdxl/generate', generateOptions);
const generateData = await generateResponse.json();
if (generateData.status !== 'queued') { throw new Error('Failed to queue the job');
}
const jobId = generateData.job;
// Polling for the job status
const statusOptions = {
method: 'GET',
headers: {
accept: 'application/json',
'X-Prodia-Key': PRODIA_API_KEY
}
};
let statusData;
while (true) { const statusResponse = await fetch(`https://api.prodia.com/v1/job/${jobId}`, statusOptions);
statusData = await statusResponse.json();
if (statusData.status === 'succeeded') {
return statusData.imageUrl;
} else if (statusData.status === 'failed') { throw new Error('Image generation failed');
}
// Wait for a short period before checking again
await new Promise(resolve => setTimeout(resolve, 5000));
}
} catch (error) { return "图像生成或转换失败，请检查！" + error.message;}
}
/**
* Return a streaming response with the generated image.
*
* The response will contain the generated image as a base64 encoded string
* and the original and translated prompts as text. The response will be sent
* as a Server-Sent Event (SSE) stream, with the `data` event containing the
* response payload.
*
* @param {string} originalPrompt - The original prompt given to the model.
* @param {string} translatedPrompt - The translated prompt given to the model.
* @param {string} size - The size of the generated image.
* @param {string} model - The model used to generate the image.
* @param {string} imageUrl - The URL of the generated image.
* @returns {Response} - The response object.
*/
function handleStreamResponse(originalPrompt, translatedPrompt, size, model, imageUrl) { const uniqueId = `chatcmpl-${Date.now()}`;
const createdTimestamp = Math.floor(Date.now() / 1000);
const systemFingerprint = "fp_" + Math.random().toString(36).substr(2, 9);
const content = `🎨 原始提示词：${originalPrompt}n` +
`🌐 翻译后的提示词：${translatedPrompt}n` +
`📐 图像规格：${size}n` +
`🌟 图像生成成功！n` +
` 以下是结果：nn` +
``;
const responsePayload = {
id: uniqueId,
object: "chat.completion.chunk",
created: createdTimestamp,
model: model,
system_fingerprint: systemFingerprint,
choices: [
{
index: 0,
delta: {
content: content,
},
finish_reason: "stop",
},
],
};
const dataString = JSON.stringify(responsePayload);
return new Response(`data: ${dataString}nn`, {
status: 200,
headers: {
"Content-Type": "text/event-stream",
'Access-Control-Allow-Origin': '*',
"Access-Control-Allow-Headers": '*',
},
});
}
/**
* Return a non-streaming response with the generated image.
*
* The response will contain the generated image as a base64 encoded string
* and the original and translated prompts as text.
*
* @param {string} originalPrompt - The original prompt given to the model.
* @param {string} translatedPrompt - The translated prompt given to the model.
* @param {string} size - The size of the generated image (e.g. 1024x1024).
* * @param {string} model - The model used to generate the image (e.g. @cf/stabilityai/stable-diffusion-xl-base-1.0).
* @param {string} imageUrl - The URL of the generated image.
* @return {Response} - The response object with the generated image and prompts.
*/
function handleNonStreamResponse(originalPrompt, translatedPrompt, size, model, imageUrl) { const uniqueId = `chatcmpl-${Date.now()}`;
const createdTimestamp = Math.floor(Date.now() / 1000);
const systemFingerprint = "fp_" + Math.random().toString(36).substr(2, 9);
const content = `🎨 原始提示词：${originalPrompt}n` +
`🌐 翻译后的提示词：${translatedPrompt}n` +
`📐 图像规格：${size}n` +
`🌟 图像生成成功！n` +
` 以下是结果：nn` +
``;
const response = {
id: uniqueId,
object: "chat.completion",
created: createdTimestamp,
model: model,
system_fingerprint: systemFingerprint,
choices: [{
index: 0,
message: {
role: "assistant",
content: content
},
finish_reason: "stop"
}],
usage: {
prompt_tokens: translatedPrompt.length,
completion_tokens: content.length,
total_tokens: translatedPrompt.length + content.length
}
};
const dataString = JSON.stringify(response);
return new Response(dataString, {
status: 200,
headers: {
'Content-Type': 'application/json',
'Access-Control-Allow-Origin': '*',
'Access-Control-Allow-Headers': '*'
}
});
}
/**
* @description
* POST request to Cloudflare AI API
* @param {string} model - AI model name
* @param {object} jsonBody - JSON object to be sent in the body of the request
* @returns {Promise} - Response object
* @throws {Error} - If response status is not OK
*/
async function postRequest(model, jsonBody) { const cf_account = CF_ACCOUNT_LIST[Math.floor(Math.random() * CF_ACCOUNT_LIST.length)];
const apiUrl = `https://api.cloudflare.com/client/v4/accounts/${cf_account.account_id}/ai/run/${model}`;
const response = await fetch(apiUrl, {
method: 'POST',
headers: { 'Authorization': `Bearer ${cf_account.token}`,
'Content-Type': 'application/json'
},
body: JSON.stringify(jsonBody)
});
if (!response.ok) { throw new Error('Unexpected response' + response.status);
}
return response;
}
/**
* Extract translate flag from prompt string.
*
* This function will parse the flag from the given prompt string and return the
* translate flag. If the flag is not found, it will return the default translate
* flag set in CF_IS_TRANSLATE.
*
* @param {string} prompt The prompt string to parse the flag from.
* @return {boolean} The translate flag parsed from the prompt string.
*/
function extractTranslate(prompt) { const match = prompt.match(/---n?tl/);
if (match && match[0]) { if (match[0] == "---ntl") {
return false;
}
else if (match[0] == "---tl") {
return true;
}
}
return CF_IS_TRANSLATE;
}
/**
* Remove translate flag from prompt string.
*
* This function will remove the translate flag ("---ntl" or "---tl") from the
* given prompt string and return the cleaned prompt string.
*
* @param {string} prompt The prompt string to clean.
* @return {string} The cleaned prompt string.
*/
function cleanPromptString(prompt) { return prompt.replace(/---n?tl/, "").trim();}
addEventListener('fetch', event => { event.respondWith(handleRequest(event.request));
}); 设置自定义域
4. 在刚刚创建的 worker 中点击设置，在域和路由中添加自定义域，并填写自己的子域名，如 image.xxxx.com 1：(此处需自行查阅学习如何给 cf 添加自己的域名)
添加自定义模型
5. 复制自己的域名在任意对话平台或 API 中转站中添加自定义接口以及自定义模型：(此处以 ChatGPT-Next-Web 7 为例)
以下是所有模型的介绍，可以根据自己的需求来选择模型：
- animagineXLV3_v30：适合想要动画风格的创作者，可能在动画效果上表现优越。
- devlishphotorealism_sdxl15：如果你需要超真实的图像效果，这个模型可能是个好选择，特别适合用于摄影风格的创作。
- dreamshaperXL10_alpha2：注重创意和想象力，适合艺术风格较强的作品。
- juggernautXL_v45：可能在生成大型复杂场景方面表现出色，可以考虑。
- realismEngineSDXL_v10：专注于真实效果的生成，适合需要高度逼真图像的项目。
- sd_xl_base_1.0 和 sd_xl_base_1.0_inpainting_0.1：这些是基础模型，适合多种用途，也可以在需要时进行细微调整。
- turbovisionXL_v431：可能在速度和效率上有优势，适合需要快速生成的场景。
