curl --request POST \
--url https://api.ppio.com/openai/v1/files \
--header 'Authorization: <authorization>' \
--header 'Content-Type: <content-type>' \
--data '
{
"purpose": "<string>"
}
'{
"id": "<string>",
"object": "<string>",
"bytes": 123,
"created_at": 123,
"filename": "<string>",
"purpose": "<string>",
"metadata": {
"total_requests": 123
}
}curl --request POST \
--url https://api.ppio.com/openai/v1/files \
--header 'Authorization: <authorization>' \
--header 'Content-Type: <content-type>' \
--data '
{
"purpose": "<string>"
}
'{
"id": "<string>",
"object": "<string>",
"bytes": 123,
"created_at": 123,
"filename": "<string>",
"purpose": "<string>",
"metadata": {
"total_requests": 123
}
}application/json.jsonl,其中每行描述一个 API 推理请求的详细信息。每个请求必须包含一个唯一的 custom_id,以便在批处理完成后在输出文件中查找相应的推理结果。每行的 body 字段中的参数将作为实际推理请求参数发送到指定的 endpoint。{"custom_id": "request-1", "body": {"model": "deepseek/deepseek-v3-0324", "messages": [{"role": "user", "content": "Hello, world!"}], "max_tokens": 400}}
{"custom_id": "request-2", "body": {"model": "deepseek/deepseek-v3-0324", "messages": [{"role": "system", "content": "You are a helpful assistant."},{"role": "user", "content": "Hello world!"}],"max_tokens": 1000}}
batch。枚举值: batchfile。