forked from jiuyuan/InfiniTensor
7f6aec6c17
* fix(dist): 改善分布式脚本，只打印绝对误差 * feat(dist): 增加可导出onnx的pytorch运行脚本 * feat(front): 增加对Y值为-inf的where算子的图优化 * feat(kernel): 对b为常数的pow和div算子进行特判优化 * fix(front): 消除前端对global output形状信息的依赖，分布式脚本删除不必要的shape infer * feat(kernel): 针对matmul中bias为行向量时的expand操作的特化优化 * fix(kernel): 删除div pow const中不必要的同步 * Update expand.cu * fix: fix comments --------- Co-authored-by: Haojie Wang <haojie0429@gmail.com> Co-authored-by: Derui Yang <ydrml@hotmail.com> |
||
|---|---|---|
| .. | ||
| cuda_attention_kvcache.h | ||
| cuda_clip.h | ||
| cuda_common.h | ||
| cuda_element_wise.h | ||
| cuda_expand.h | ||
| cuda_kernel_wihtout_config.h | ||
| cuda_layernorm.h | ||
| cuda_pad_slice.h | ||
| cuda_rmsnorm.h | ||
| cuda_rope.h | ||
| cuda_runtime.h | ||
| cuda_softmax.h | ||
| cuda_split_concat.h | ||
| cuda_transpose.h | ||
| cuda_unary.h | ||
| cuda_utility.h | ||
| cuda_where.h | ||
| gather.h | ||
| gbmm_g2bmm.cuh | ||
| gbmm_g2bmm.h | ||
| nccl_communicator.h | ||
| operator_timer.h | ||
| resize.cuh | ||