Bound Ollama embedding batches for Yggdrasil GPU

This commit is contained in:
Meta Vens 2026-07-16 23:26:58 +01:00
commit 586147a751

View file

@ -13,7 +13,7 @@ use serde_json::{Value, json};
use std::time::Duration;
const POINT_BATCH_SIZE: usize = 128;
const EMBED_BATCH_SIZE: usize = 32;
const EMBED_BATCH_SIZE: usize = 4;
#[derive(Clone, Debug)]
pub(crate) struct QdrantConfig {