EngineX-Hygon/sglang
5
0
Fork 0
Code Issues Pull Requests Actions 7 Projects Releases Wiki Activity

[router][grpc] Support E2E non-stream chat completions (#10980)

Browse Source
This commit is contained in:
Chang Su
2025-09-26 22:02:06 -07:00
committed by GitHub
parent bd95944cf6
commit 37f3325b06
8 changed files with 325 additions and 136 deletions
Download Patch File Download Diff File Expand all files Collapse all files

17
sgl-router/src/protocols/spec.rs
View File

@@ -423,10 +423,25 @@ pub struct ChatCompletionResponse {
pub system_fingerprint: Option<String>,
}
/// Response message structure for ChatCompletionResponse (different from request ChatMessage)
#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct ChatCompletionMessage {
pub role: String, // Always "assistant" for responses
#[serde(skip_serializing_if = "Option::is_none")]
pub content: Option<String>,
#[serde(skip_serializing_if = "Option::is_none")]
pub tool_calls: Option<Vec<ToolCall>>,
/// Reasoning content for O1-style models (SGLang extension)
#[serde(skip_serializing_if = "Option::is_none")]
pub reasoning_content: Option<String>,
// Note: function_call is deprecated and not included
// Note: refusal, annotations, audio are not added yet
}
#[derive(Debug, Clone, Deserialize, Serialize)]
pub struct ChatChoice {
pub index: u32,
pub message: ChatMessage,
pub message: ChatCompletionMessage,
#[serde(skip_serializing_if = "Option::is_none")]
pub logprobs: Option<ChatLogProbs>,
pub finish_reason: Option<String>, // "stop", "length", "tool_calls", "content_filter", "function_call"