🧠 两小时从 0 训出你的 64M 大模型。Train your own 64M LLM from scratch in 2 hours.
update readme
gongjy54af1ddd030a
diff --git a/README.md b/README.mdindex 3d9420c..5705574 100644--- a/README.md+++ b/README.md@@ -193,13 +193,13 @@ streamlit run fast_inference.py🍭 【Tip】预训练和全参微调pretrain和full_sft均支持多卡加速-* 单机N卡启动训练(ddp)+* 单机N卡启动训练(DDP)```bashtorchrun --nproc_per_node N 1-pretrain.py# andtorchrun --nproc_per_node N 3-full_sft.py```-* 单机N卡启动训练(deepspeed)+* 单机N卡启动训练(DeepSpeed)```bashdeepspeed --master_port 29500 --num_gpus=N 1-pretrain.py# and