返回Ssignal1908/19 16:00·8天前·1 signalsArchived topic · 归档话题,来源已停止追踪Same effective batch does not mean same training time with gradient accumulation, tested on LoRA on T4 and L4 [D]收藏分享导出 Markdown热度趋势趋势数据积累中百分比基于当前可用热度信号,而非评论数或独立用户人数。