apunkt/plano
1
0
Fork 0
mirror of https://github.com/katanemo/plano.git synced 2026-05-08 15:22:43 +02:00
Code Issues Projects Releases Packages Wiki Activity Actions

adding support for model aliases in archgw (#566)

Browse source 
* adding support for model aliases in archgw

* fixed PR based on feedback

* removing README. Not relevant for PR

---------

Co-authored-by: Salman Paracha <salmanparacha@MacBook-Pro-136.local>
This commit is contained in:
Salman Paracha 2025-09-16 11:12:08 -07:00 • committed by GitHub
parent 1e8c81d8f6
commit 4eb2b410c5
No known key found for this signature in database
GPG key ID: B5690EEEBB952194
12 changed files with 634 additions and 14 deletions
Download patch file Download diff file Expand all files Collapse all files

19
crates/hermesllm/src/apis/openai.rs
View file

@ -104,6 +104,20 @@ pub struct ChatCompletionsRequest {
// pub web_search: Option<bool>, // GOOD FIRST ISSUE: Future support for web search
}
impl ChatCompletionsRequest {
/// Suppress max_tokens if the model is o3, o3-*, openrouter/o3, or openrouter/o3-*
pub fn suppress_max_tokens_if_o3(&mut self) {
let model = self.model.as_str();
let is_o3 = model == "o3"
|| model.starts_with("o3-")
|| model == "openrouter/o3"
|| model.starts_with("openrouter/o3-");
if is_o3 {
self.max_tokens = None;
}
}
}
// ============================================================================
// CHAT COMPLETIONS API TYPES
// ============================================================================
@ -530,7 +544,10 @@ impl TryFrom<&[u8]> for ChatCompletionsRequest {
type Error = OpenAIStreamError;
fn try_from(bytes: &[u8]) -> Result<Self, Self::Error> {
serde_json::from_slice(bytes).map_err(OpenAIStreamError::from)
let mut req: ChatCompletionsRequest = serde_json::from_slice(bytes).map_err(OpenAIStreamError::from)?;
// Use the centralized suppression logic
req.suppress_max_tokens_if_o3();
Ok(req)
}
}