感谢以下开发者做出的贡献:
liuchao, yiyanzhi, hujiahui8, xuhui, yangshuo, nelson.lossing
- 支持CPU后端pool2d相关算子(avg/max/global)的自动生成流程。!763 (CPU)
- AKG集成符号切分算法。!764 (Ascend)
- GPU后端自动切分算法优化。 !784 !785 !783 (GPU)
- CPU后端Transpose算子性能优化。!811 (CPU)
- 修复了多filter场景中读写依赖分析时的bug。!778 (All)
- 修复了CPU后端batchMatmul识别轴流程中的bug。!807 (CPU)
- 修复了系统环境gcc版本低于7.3时,编译reduce算子时存在的bug。!817 (All)
感谢以下开发者做出的贡献:
yangsijia, polyhedral, zhangrenwei, yiyanzhi, hujiahui8, zhengzuohe, zhangzhaochuang, xuhui, liuchao, xiaruijie, DeshiChen, nelson.lossing, zhenzhang, chenlei_autodiff, wYann, hanhuifeng, gaoxiong, zichun_ye, chengbin, wangrao124, huangmengxi, harenome, huoxinyou, yanglinfeng, Etienne
- 自定义算子支持CPU编译流程。!612 (CPU)
- 支持高维CSR算子及其性能优化。!664 (GPU)
- 更新昇腾后端Poly重启动流程策略。!701 (Ascend)
- 支持CSRMM算子的自动生成。!709 (GPU)
- 新调度器支持调度原语辅助调度。!715 (All backends)
感谢以下开发者做出的贡献:
yangsijia, polyhedral, zhangrenwei, yiyanzhi, hujiahui8, zhengzuohe, zhangzhaochuang, xuhui, liuchao, xiaruijie, DeshiChen, nelson.lossing, zhenzhang,
chenlei_autodiff, wYann, hanhuifeng, gaoxiong, zichun_ye, chengbin, wangrao124, huangmengxi, harenome, huoxinyou, yanglinfeng, Etienne
- 为MindSpore HybridDSL特性提供循环原语支持。!560 (所有后端)
- 更新GPU后端的shared/local内存提升策略。!556 (GPU)
- 昇腾后端使用isl提供的新接口重构了pluto调度算法后的reschedule pass。!512 (Ascend)
感谢以下开发者做出的贡献:
yangsijia, polyhedral, zhangrenwei, yiyanzhi, hujiahui8, zhengzuohe, zhangzhaochuang, xuhui, liuchao, xiaruijie, DeshiChen, nelson.lossing, zhenzhang,
chenlei_autodiff, lingyunli63, wYann, hanhuifeng, gaoxiong, zichun_ye, chengbin, wangrao124, huangmengxi, harenome, lear, huoxinyou, yanglinfeng, Etienne, Trump