mirror of
https://github.com/Memo-2023/mana-monorepo.git
synced 2026-05-18 04:09:41 +02:00
e757470cb0
Extends the chat-completions surface so callers can ask any provider to call named functions and get structured tool_calls back. Wired through all three provider adapters so the planner and companion can switch off the fragile JSON-parsing pathway. - Request: tools[], tool_choice, assistant tool_calls, tool-role messages with tool_call_id. - Response: MessageResponse.tool_calls, Choice.finish_reason adds "tool_calls", DeltaContent streams tool_calls. - Google provider: Tool(function_declarations=...) build, result normalised (args dict → JSON string), function_response parts on a user turn for tool-role messages. - OpenAI-compat: 1:1 passthrough of the OpenAI spec. - Ollama: /api/chat passthrough; model-level capability check via a TOOL_CAPABLE_OLLAMA_PATTERNS whitelist (llama3.1+, qwen2.5+, mistral, command-r, …) — unsupported models rejected rather than silently falling back to prose. - Router: model_supports_tools() check upfront for both streaming and non-streaming paths; ProviderCapabilityError bubbles as 400. No silent downgrade. Missing tool support = explicit error. Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>
120 lines
2.9 KiB
Python
120 lines
2.9 KiB
Python
"""Response models for OpenAI-compatible API."""
|
|
|
|
import time
|
|
import uuid
|
|
from typing import Literal
|
|
|
|
from pydantic import BaseModel, Field
|
|
|
|
|
|
class Usage(BaseModel):
|
|
"""Token usage information."""
|
|
|
|
prompt_tokens: int = 0
|
|
completion_tokens: int = 0
|
|
total_tokens: int = 0
|
|
|
|
|
|
class ToolCallFunction(BaseModel):
|
|
"""Function portion of a tool_call the assistant produced."""
|
|
|
|
name: str
|
|
arguments: str # JSON-encoded (OpenAI spec)
|
|
|
|
|
|
class ToolCall(BaseModel):
|
|
"""A tool invocation the assistant decided to make."""
|
|
|
|
id: str
|
|
type: Literal["function"] = "function"
|
|
function: ToolCallFunction
|
|
|
|
|
|
class MessageResponse(BaseModel):
|
|
"""Response message from the model."""
|
|
|
|
role: Literal["assistant"] = "assistant"
|
|
content: str | None = None
|
|
tool_calls: list[ToolCall] | None = None
|
|
|
|
|
|
class Choice(BaseModel):
|
|
"""A single completion choice."""
|
|
|
|
index: int = 0
|
|
message: MessageResponse
|
|
finish_reason: (
|
|
Literal["stop", "length", "content_filter", "tool_calls"] | None
|
|
) = "stop"
|
|
|
|
|
|
class ChatCompletionResponse(BaseModel):
|
|
"""Response from chat completions endpoint (non-streaming)."""
|
|
|
|
id: str = Field(default_factory=lambda: f"chatcmpl-{uuid.uuid4().hex[:12]}")
|
|
object: Literal["chat.completion"] = "chat.completion"
|
|
created: int = Field(default_factory=lambda: int(time.time()))
|
|
model: str
|
|
choices: list[Choice]
|
|
usage: Usage = Field(default_factory=Usage)
|
|
|
|
|
|
class DeltaContent(BaseModel):
|
|
"""Delta content for streaming responses."""
|
|
|
|
role: Literal["assistant"] | None = None
|
|
content: str | None = None
|
|
tool_calls: list[ToolCall] | None = None
|
|
|
|
|
|
class StreamChoice(BaseModel):
|
|
"""A single streaming choice."""
|
|
|
|
index: int = 0
|
|
delta: DeltaContent
|
|
finish_reason: (
|
|
Literal["stop", "length", "content_filter", "tool_calls"] | None
|
|
) = None
|
|
|
|
|
|
class ChatCompletionStreamResponse(BaseModel):
|
|
"""Response chunk from chat completions endpoint (streaming)."""
|
|
|
|
id: str = Field(default_factory=lambda: f"chatcmpl-{uuid.uuid4().hex[:12]}")
|
|
object: Literal["chat.completion.chunk"] = "chat.completion.chunk"
|
|
created: int = Field(default_factory=lambda: int(time.time()))
|
|
model: str
|
|
choices: list[StreamChoice]
|
|
|
|
|
|
class ModelInfo(BaseModel):
|
|
"""Information about a model."""
|
|
|
|
id: str
|
|
object: Literal["model"] = "model"
|
|
created: int = Field(default_factory=lambda: int(time.time()))
|
|
owned_by: str = "mana-llm"
|
|
|
|
|
|
class ModelsResponse(BaseModel):
|
|
"""Response from models endpoint."""
|
|
|
|
object: Literal["list"] = "list"
|
|
data: list[ModelInfo]
|
|
|
|
|
|
class EmbeddingData(BaseModel):
|
|
"""A single embedding result."""
|
|
|
|
object: Literal["embedding"] = "embedding"
|
|
index: int = 0
|
|
embedding: list[float]
|
|
|
|
|
|
class EmbeddingResponse(BaseModel):
|
|
"""Response from embeddings endpoint."""
|
|
|
|
object: Literal["list"] = "list"
|
|
data: list[EmbeddingData]
|
|
model: str
|
|
usage: Usage = Field(default_factory=Usage)
|