Training & Pre-training
Months of thousands of GPUs predicting the next token, trillions of times.
Anda sedang melihat entri penjelas konsep: apa arti istilah ini dalam bahasa yang mudah dipahami, dan entri-entri yang dikaitkannya.
Pre-training runs a model over trillions of tokens, nudging weights after every prediction error, in synchronized runs across tens of thousands of accelerators — the workload the biggest clusters on this site were built for. A frontier run costs hundreds of millions of dollars in compute, fails in exotic ways (silent data corruption, stragglers, loss spikes), and is why checkpointing, interconnects and reliability engineering are front-page infrastructure topics.
Bagaimana ini terhubung ke bagian lain dari stack
Tautan catatan ini sekilas — klik grup untuk langsung menuju ke sana.
Terkait
Sumber
Llama 3 herd paper (training at scale) ↗
Rekaman ditinjau per 2026-01-15 — lihat Metodologi.