- 16 Apr, 2021 1 commit
-
-
Jing Zhang authored
-
- 15 Apr, 2021 2 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
- 14 Apr, 2021 8 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
- 13 Apr, 2021 5 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
Chao Liu authored
* overhaul vector_type, make int8x4_t real vector instead of aliasing from int32_t
-
Jing Zhang authored
-
Chao Liu authored
* initial implementation for magic number division and DynamicMerge_v2_magic_division that uses it * turn off DynamicMerge_v2_magic_division that use magic number division by default
-
- 12 Apr, 2021 2 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
- 09 Apr, 2021 2 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
- 08 Apr, 2021 1 commit
-
-
Jing Zhang authored
-
- 07 Apr, 2021 3 commits
-
-
zjing14 authored
* Hybrid direct + implicit GEMM forward convolution NCHWc v5r1. Input tensor bypass LDS. Support fp32/fp16/int8
-
Jing Zhang authored
-
Jing Zhang authored
-
- 06 Apr, 2021 8 commits
-
-
Chao Liu authored
-
Chao Liu authored
* use address_space(4) in kernel signature to fix performance issue when passing tensor descriptor from host to kernel by (void) pointers * remove passing by pointer* option (only use pass by value or void*)
-
zjing14 authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
- 05 Apr, 2021 2 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
- 02 Apr, 2021 4 commits
-
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
Jing Zhang authored
-
- 01 Apr, 2021 2 commits
-
-
root authored
-
Jing Zhang authored
-