huggingface/text-generation-inference
5
0
Fork 0
mirror of https://github.com/huggingface/text-generation-inference.git synced 2025-09-11 12:24:53 +00:00
Code Issues Packages Projects Releases Wiki Activity

feat: emit params in logs for each request

Browse Source
This commit is contained in:
drbh 2024-04-17 01:18:09 +00:00
parent 593c443b45
commit bd28c36815
1 changed files with 26 additions and 18 deletions
Show Stats Download Patch File Download Diff File Expand all files Collapse all files

14
router/src/server.rs
View File

@ -657,6 +657,7 @@ async fn completions(
format!("{}-{}", info.version, info.docker_label.unwrap_or("native"));
let infer_clone = infer.clone();
let compute_type_clone = compute_type.clone();
let params_clone = generate_request.parameters.clone();
// Create a future for each generate_stream_internal call.
let generate_future = async move {
@ -690,7 +691,8 @@ async fn completions(
let (header_tx, header_rx) = oneshot::channel();
let (sse_tx, sse_rx) = tokio::sync::mpsc::unbounded_channel();
tokio::spawn(async move {
tokio::spawn(
async move {
let (header_map, sse) = generate_stream_internal(
infer_clone.clone(),
compute_type_clone.clone(),
@ -710,7 +712,11 @@ async fn completions(
break;
}
}
});
}
.instrument(
tracing::info_span!("request", index = %index, parameters = ?params_clone),
),
);
(header_rx, sse_rx)
};
@ -796,6 +802,7 @@ async fn completions(
for (index, generate_request) in generate_requests.into_iter().enumerate() {
let infer_clone = infer.clone();
let compute_type_clone = compute_type.clone();
let params_clone = generate_request.parameters.clone();
let response_future = async move {
let result = generate(
Extension(infer_clone),
@ -804,7 +811,8 @@ async fn completions(
)
.await;
result.map(|(headers, generation)| (index, headers, generation))
};
}
.instrument(tracing::info_span!("request", index = %index, parameters = ?params_clone));
responses.push(response_future);
}
let generate_responses = responses.try_collect::<Vec<_>>().await?;