Chinese-Vicuna: A Chinese Instruction-following LLaMA-based Model —— 一个中文低资源的llama+lora方案,结构参考alpaca
4.14k
stars
425
forks
source link
简单的问题,finetune_other_continue.sh中step = 样本量/(MICRO_BATCH_SIZE*GRADIENT_ACCUMULATION_STEPS)。多卡的时候,是不是应该得是 step = 样本量/batch/(MICRO_BATCH_SIZE*GRADIENT_ACCUMULATION_STEPS*gpu数量)数量呢?这边一直不是很理解 #233
Closed
niuhuluzhihao closed 1 year ago