From 52208f5b78fd8cc31d01f440b5f5e250896c1e64 Mon Sep 17 00:00:00 2001
From: Morgan Funtowicz <funtowiczmo@gmail.com>
Date: Mon, 4 Nov 2024 23:24:50 +0100
Subject: [PATCH] misc(backend): decrease log verbosity in callback

---
 backends/llamacpp/src/backend.rs | 4 ++--
 1 file changed, 2 insertions(+), 2 deletions(-)

diff --git a/backends/llamacpp/src/backend.rs b/backends/llamacpp/src/backend.rs
index 08fac675..62b4743d 100644
--- a/backends/llamacpp/src/backend.rs
+++ b/backends/llamacpp/src/backend.rs
@@ -17,7 +17,7 @@ use tokenizers::Tokenizer;
 use tokio::sync::mpsc::{unbounded_channel, UnboundedSender};
 use tokio::time::Instant;
 use tokio_stream::wrappers::UnboundedReceiverStream;
-use tracing::{error, info};
+use tracing::{debug, error, info};
 
 type InferResult = Result<InferStreamResponse, InferError>;
 
@@ -113,7 +113,7 @@ fn llama_generate_callback(
     is_final: bool,
     n_generated_tokens: usize,
 ) -> bool {
-    info!("Generated token: {new_token_id} -> logits={new_token_logit}, is_final={is_final} ({n_generated_tokens})");
+    debug!("Generated token: {new_token_id} -> logits={new_token_logit}, is_final={is_final} ({n_generated_tokens})");
 
     let ctx = unsafe { &mut *ctx };