🚀🚀 「大模型」2小时完全从0训练64M的小参数GPT！🌏 Train a 64M-parameter GPT from scratch in just 2h!
Fix: support loading DDP-saved LoRA weights for inference
W
whitesword committed
3a18fdd666a93852d9779cdccb0165b5ea209f91
Parent: fe24501
🚀🚀 「大模型」2小时完全从0训练64M的小参数GPT！🌏 Train a 64M-parameter GPT from scratch in just 2h!