hf_text-generation-inference/backends/v3/benches/prefix_cache.rs

48 lines
1.4 KiB
Rust
Raw Permalink Normal View History

use std::sync::Arc;
use criterion::{black_box, criterion_group, criterion_main, Criterion};
use rand::Rng;
use text_generation_router_v3::block_allocator::Allocator;
use text_generation_router_v3::radix::RadixAllocator;
fn prefix_cache_benchmark(c: &mut Criterion) {
// let prefixes: Vec<Vec<u32>> = (0..8192)
// .chunks(256)
// .into_iter()
// .map(|c| c.collect())
// .collect();
let mut cache = RadixAllocator::new(1, 262144, None);
c.bench_function("Radix allocator", |b| {
b.iter_batched(
|| {
//prefixes
// .choose_multiple(&mut rand::thread_rng(), 5)
// .fold(Vec::new(), |mut v, s| {
// v.extend(s);
// v
// })
(0..7936)
.map(|_| rand::thread_rng().gen_range(0..1024))
.collect::<Vec<u32>>()
},
|prefill| {
let alloc = cache.allocate(
prefill.len() as u32 + 13,
Some(Arc::new(black_box(prefill))),
);
if let Some(alloc) = alloc {
cache.free(alloc.blocks.clone(), alloc.allocation_id);
}
},
criterion::BatchSize::SmallInput,
);
});
}
criterion_group!(benches, prefix_cache_benchmark);
criterion_main!(benches);