Skip to content

“Quen 3's pre-training occurred in three stages. In stage…” — Y Combinator Startup Podcast at 4:36

“Quen 3's pre-training occurred in three stages. In stage 1, the general stage, models were trained on over 30 trillion tokens covering 119 languages at a sequence length of 4096 tokens. In stage 2, the reasoning stage, models were trained on an additional 5 trillion higher-quality tokens featuring more STEM…”

0:00 12:31