strata serve: prompt 4022 tokens = 0 reused + 4022 read in 12765 ms (315.1 tok/s), 256 generated in 5653 ms (45.3 tok/s), drafts accepted 166 of 220, 1 checkpoints
strata serve: decode expert cache hit rate: 81.9% (107474 hits / 131199 lookups)
strata serve: KV streaming: 99.35% of 1953336 block reads hit VRAM, 51.3 MiB read from RAM
strata serve: suffix drafts: 5 windows, 14 of 15 drafts accepted
