lm-deluge 0.0.67__py3-none-any.whl → 0.0.90__py3-none-any.whl

lm_deluge/server/models_anthropic.py

@@ -0,0 +1,172 @@
+"""
+Pydantic models for Anthropic-compatible API request/response formats.
+"""
+
+from __future__ import annotations
+
+import uuid
+from typing import Any, Literal
+
+from pydantic import BaseModel, Field
+
+
+# ============================================================================
+# Request Models
+# ============================================================================
+
+
+class AnthropicContentBlock(BaseModel):
+    """Content block in Anthropic message."""
+
+    type: Literal[
+        "text",
+        "image",
+        "tool_use",
+        "tool_result",
+        "document",
+        "container_upload",
+        "thinking",
+        "redacted_thinking",
+    ] = "text"
+
+    # Text content
+    text: str | None = None
+    citations: list[dict[str, Any]] | None = None
+    cache_control: dict[str, Any] | None = None
+
+    # Image/document content
+    source: dict[str, Any] | None = None
+    title: str | None = None
+    context: str | None = None
+
+    # Container upload content
+    file_id: str | None = None
+
+    # Tool use (assistant response)
+    id: str | None = None
+    name: str | None = None
+    input: dict[str, Any] | None = None
+    caller: dict[str, Any] | None = None
+
+    # Tool result (user message)
+    tool_use_id: str | None = None
+    content: str | list[dict[str, Any]] | None = None
+    is_error: bool | None = None
+
+    # Thinking content
+    thinking: str | None = None
+    signature: str | None = None
+
+    # Redacted thinking content
+    data: str | None = None
+
+
+class AnthropicMessage(BaseModel):
+    """Anthropic message format."""
+
+    role: Literal["user", "assistant"]
+    content: str | list[AnthropicContentBlock]
+
+
+class AnthropicTool(BaseModel):
+    """Tool definition for Anthropic."""
+
+    name: str
+    description: str | None = None
+    input_schema: dict[str, Any] | None = None
+
+
+class AnthropicMessagesRequest(BaseModel):
+    """Anthropic Messages API request format."""
+
+    model: str
+    max_tokens: int
+    messages: list[AnthropicMessage]
+    stream: bool = False
+
+    # System prompt (can be string or content blocks)
+    system: str | list[AnthropicContentBlock] | None = None
+
+    # Thinking configuration (Anthropic reasoning)
+    thinking: dict[str, Any] | None = None
+
+    # Sampling parameters
+    temperature: float | None = None
+    top_p: float | None = None
+    top_k: int | None = None
+
+    # Tool calling
+    tools: list[AnthropicTool] | None = None
+    tool_choice: dict[str, Any] | None = None
+
+    # Metadata
+    metadata: dict[str, Any] | None = None
+    stop_sequences: list[str] | None = None
+
+
+# ============================================================================
+# Response Models
+# ============================================================================
+
+
+class AnthropicResponseContentBlock(BaseModel):
+    """Content block in Anthropic response."""
+
+    type: Literal["text", "tool_use", "thinking", "redacted_thinking"]
+
+    # Text content
+    text: str | None = None
+    citations: list[dict[str, Any]] | None = None
+
+    # Tool use
+    id: str | None = None
+    name: str | None = None
+    input: dict[str, Any] | None = None
+
+    # Thinking content
+    thinking: str | None = None
+    signature: str | None = None
+
+    # Redacted thinking content
+    data: str | None = None
+
+
+class AnthropicUsage(BaseModel):
+    """Token usage for Anthropic."""
+
+    input_tokens: int = 0
+    output_tokens: int = 0
+    cache_creation_input_tokens: int | None = None
+    cache_read_input_tokens: int | None = None
+
+
+class AnthropicMessagesResponse(BaseModel):
+    """Anthropic Messages API response format."""
+
+    id: str = Field(default_factory=lambda: f"msg_{uuid.uuid4().hex[:24]}")
+    type: Literal["message"] = "message"
+    role: Literal["assistant"] = "assistant"
+    model: str
+    content: list[AnthropicResponseContentBlock]
+    stop_reason: str | None = None
+    stop_sequence: str | None = None
+    usage: AnthropicUsage
+
+
+# ============================================================================
+# Error Response
+# ============================================================================
+
+
+class AnthropicErrorDetail(BaseModel):
+    """Error detail for Anthropic."""
+
+    type: str
+    message: str
+
+
+class AnthropicErrorResponse(BaseModel):
+    """Anthropic error response format."""
+
+    type: Literal["error"] = "error"
+    error: AnthropicErrorDetail

lm_deluge/server/models_openai.py

@@ -0,0 +1,175 @@
+"""
+Pydantic models for OpenAI-compatible API request/response formats.
+"""
+
+from __future__ import annotations
+
+import time
+import uuid
+from typing import Any, Literal
+
+from pydantic import BaseModel, Field
+
+
+# ============================================================================
+# Request Models
+# ============================================================================
+
+
+class OpenAIMessage(BaseModel):
+    """OpenAI chat message format."""
+
+    role: Literal["system", "user", "assistant", "tool", "function", "developer"]
+    content: str | list[dict[str, Any]] | None = None
+    name: str | None = None
+    tool_call_id: str | None = None
+    tool_calls: list[dict[str, Any]] | None = None
+
+
+class OpenAIToolFunction(BaseModel):
+    """Function definition within a tool."""
+
+    name: str
+    description: str | None = None
+    parameters: dict[str, Any] | None = None
+
+
+class OpenAITool(BaseModel):
+    """Tool definition for function calling."""
+
+    type: Literal["function"] = "function"
+    function: OpenAIToolFunction
+
+
+class OpenAIChatCompletionsRequest(BaseModel):
+    """OpenAI Chat Completions API request format."""
+
+    model: str
+    messages: list[OpenAIMessage]
+    stream: bool = False
+
+    # Sampling parameters
+    max_tokens: int | None = None
+    max_completion_tokens: int | None = None
+    temperature: float | None = None
+    top_p: float | None = None
+    seed: int | None = None
+
+    # Tool calling
+    tools: list[OpenAITool] | None = None
+    tool_choice: str | dict[str, Any] | None = None
+
+    # Response formatting
+    response_format: dict[str, Any] | None = None
+
+    # Reasoning models
+    reasoning_effort: Literal["low", "medium", "high"] | None = None
+
+    # Other options (accepted but may be ignored)
+    n: int | None = None
+    stop: str | list[str] | None = None
+    presence_penalty: float | None = None
+    frequency_penalty: float | None = None
+    logit_bias: dict[str, float] | None = None
+    logprobs: bool | None = None
+    top_logprobs: int | None = None
+    user: str | None = None
+
+
+# ============================================================================
+# Response Models
+# ============================================================================
+
+
+class OpenAIFunctionCall(BaseModel):
+    """Function call within a tool call."""
+
+    name: str
+    arguments: str  # JSON string
+
+
+class OpenAIToolCall(BaseModel):
+    """Tool call in assistant message."""
+
+    id: str
+    type: Literal["function"] = "function"
+    function: OpenAIFunctionCall
+
+
+class OpenAIResponseMessage(BaseModel):
+    """Message in completion response."""
+
+    role: Literal["assistant"] = "assistant"
+    content: str | None = None
+    tool_calls: list[OpenAIToolCall] | None = None
+    refusal: str | None = None
+
+
+class OpenAIChoice(BaseModel):
+    """Choice in completion response."""
+
+    index: int = 0
+    message: OpenAIResponseMessage
+    finish_reason: str | None = None
+    logprobs: dict[str, Any] | None = None
+
+
+class OpenAIUsage(BaseModel):
+    """Token usage statistics."""
+
+    prompt_tokens: int = 0
+    completion_tokens: int = 0
+    total_tokens: int = 0
+
+
+class OpenAIChatCompletionsResponse(BaseModel):
+    """OpenAI Chat Completions API response format."""
+
+    id: str = Field(default_factory=lambda: f"chatcmpl-{uuid.uuid4().hex[:24]}")
+    object: Literal["chat.completion"] = "chat.completion"
+    created: int = Field(default_factory=lambda: int(time.time()))
+    model: str
+    choices: list[OpenAIChoice]
+    usage: OpenAIUsage | None = None
+    system_fingerprint: str | None = None
+
+
+# ============================================================================
+# Models List Response
+# ============================================================================
+
+
+class OpenAIModelInfo(BaseModel):
+    """Model information for /v1/models endpoint."""
+
+    id: str
+    object: Literal["model"] = "model"
+    created: int = Field(default_factory=lambda: int(time.time()))
+    owned_by: str = "lm-deluge"
+
+
+class OpenAIModelsResponse(BaseModel):
+    """Response for /v1/models endpoint."""
+
+    object: Literal["list"] = "list"
+    data: list[OpenAIModelInfo]
+
+
+# ============================================================================
+# Error Response
+# ============================================================================
+
+
+class OpenAIErrorDetail(BaseModel):
+    """Error detail object."""
+
+    message: str
+    type: str
+    param: str | None = None
+    code: str | None = None
+
+
+class OpenAIErrorResponse(BaseModel):
+    """Error response format."""
+
+    error: OpenAIErrorDetail