Show HN: GPT-2 124M trained from scratch – 2.48B tokens, 2.8h on 4 A40s
This project is a GPT‑2 124M model trained from scratch on 2.48 billion tokens in 2.8 hours using four A40 GPUs. It’s useful for prototypes and internal workflo
Original sourceThis project is a GPT‑2 124M model trained from scratch on 2.48 billion tokens in 2.8 hours using four A40 GPUs. It’s useful for prototypes and internal workflo
Original source