请用pydantic 给出OpenAI的响应格式
- User
请用pydantic 给出OpenAI的响应格式
Assistant
以下是使用 Pydantic 定义的 OpenAI API 响应格式:
from pydantic import BaseModel, Field from typing import List, Optional, Union, Literal from datetime import datetime # Chat Completion 相关模型 class ChatCompletionMessage(BaseModel): role: Literal["system", "user", "assistant", "function"] content: Optional[str] = None name: Optional[str] = None function_call: Optional[dict] = None class ChatCompletionChoice(BaseModel): index: int message: ChatCompletionMessage finish_reason: Optional[Literal["stop", "length", "function_call", "content_filter"]] = None class Usage(BaseModel): prompt_tokens: int completion_tokens: int total_tokens: int class ChatCompletionResponse(BaseModel): id: str object: Literal["chat.completion"] created: int model: str choices: List[ChatCompletionChoice] usage: Usage system_fingerprint: Optional[str] = None # Streaming 响应 class ChatCompletionStreamChoice(BaseModel): index: int delta: dict finish_reason: Optional[Literal["stop", "length", "function_call", "content_filter"]] = None class ChatCompletionStreamResponse(BaseModel): id: str object: Literal["chat.completion.chunk"] created: int model: str choices: List[ChatCompletionStreamChoice] system_fingerprint: Optional[str] = None # Embeddings 响应 class EmbeddingData(BaseModel): object: Literal["embedding"] index: int embedding: List[float] class EmbeddingResponse(BaseModel): object: Literal["list"] data: List[EmbeddingData] model: str usage: dict # Completions 响应(旧版 API) class CompletionChoice(BaseModel): text: str index: int logprobs: Optional[dict] = None finish_reason: Optional[str] = None class CompletionResponse(BaseModel): id: str object: Literal["text_completion"] created: int model: str choices: List[CompletionChoice] usage: Usage # 图像生成响应 class ImageData(BaseModel): url: Optional[str] = None b64_json: Optional[str] = None revised_prompt: Optional[str] = None class ImageResponse(BaseModel): created: int data: List[ImageData] # 错误响应 class ErrorDetail(BaseModel): message: str type: str param: Optional[str] = None code: Optional[str] = None class ErrorResponse(BaseModel): error: ErrorDetail # 模型列表响应 class ModelData(BaseModel): id: str object: Literal["model"] created: int owned_by: str permission: Optional[List[dict]] = None root: Optional[str] = None parent: Optional[str] = None class ModelListResponse(BaseModel): object: Literal["list"] data: List[ModelData] # 文件相关响应 class FileResponse(BaseModel): id: str object: Literal["file"] bytes: int created_at: int filename: str purpose: str status: Optional[Literal["uploaded", "processed", "error"]] = None status_details: Optional[str] = None # Fine-tuning 相关 class FineTuningJobResponse(BaseModel): id: str object: Literal["fine_tuning.job"] created_at: int finished_at: Optional[int] = None model: str fine_tuned_model: Optional[str] = None organization_id: str status: Literal["validating_files", "queued", "running", "succeeded", "failed", "cancelled"] hyperparameters: dict training_file: str validation_file: Optional[str] = None result_files: Optional[List[str]] = None trained_tokens: Optional[int] = None error: Optional[dict] = None # Assistant API 响应 class AssistantResponse(BaseModel): id: str object: Literal["assistant"] created_at: int name: Optional[str] = None description: Optional[str] = None model: str instructions: Optional[str] = None tools: Optional[List[dict]] = None file_ids: Optional[List[str]] = None metadata: Optional[dict] = None # Thread 响应 class ThreadResponse(BaseModel): id: str object: Literal["thread"] created_at: int metadata: Optional[dict] = None # Message 响应 class MessageContent(BaseModel): type: Literal["text", "image_file"] text: Optional[dict] = None image_file: Optional[dict] = None class MessageResponse(BaseModel): id: str object: Literal["thread.message"] created_at: int thread_id: str role: Literal["user", "assistant"] content: List[MessageContent] assistant_id: Optional[str] = None run_id: Optional[str] = None file_ids: Optional[List[str]] = None metadata: Optional[dict] = None # Run 响应 class RunResponse(BaseModel): id: str object: Literal["thread.run"] created_at: int thread_id: str assistant_id: str status: Literal["queued", "in_progress", "requires_action", "cancelling", "cancelled", "failed", "completed", "expired"] required_action: Optional[dict] = None last_error: Optional[dict] = None expires_at: Optional[int] = None started_at: Optional[int] = None cancelled_at: Optional[int] = None failed_at: Optional[int] = None completed_at: Optional[int] = None model: str instructions: Optional[str] = None tools: Optional[List[dict]] = None file_ids: Optional[List[str]] = None metadata: Optional[dict] = None usage: Optional[Usage] = None这些 Pydantic 模型涵盖了 OpenAI API 的主要响应格式,包括:
- Chat Completions(聊天补全)
- Embeddings(嵌入)
- Images(图像生成)
- Files(文件管理)
- Fine-tuning(微调)
- Assistants API(助手 API)
使用这些模型可以方便地进行类型检查和数据验证,确保与 OpenAI API 的交互更加健壮和类型安全。
Other models, same or similar prompt
Responses were produced in different, uncontrolled contexts (system prompts, settings, prior turns). This is not a benchmark.