kaia on Nostr: I tried Omni and converting a full book would take 65mins at 8 step inference and 200 ...

I tried Omni and converting a full book would take 65mins at 8 step inference and 200 chunk size. now trying Qwen3 1.7B to compare.