Commits · 7c397dd250b439286caf6b976c96a4122cc1f193 · jerrrrry / infinicore

20 Nov, 2025 2 commits

Issue/636 - Fix metax compile error of fp8 with maca (#639) · 685b6e03

crapromer authored Nov 20, 2025

* issue/636 - add support for fp8 with maca sdk

* issue/636 - add functional header to support Fn

* issue/636 - format code with clang

685b6e03

Issue/445 沐曦平台添加macaSDK支持 (#468) · ed012302

crapromer authored Nov 20, 2025

* initial add mc support for meta

* add command description for maca compilation

* rebase metax maca support to main

* issue/445 - clang format code on ubuntu

* issue//445 - change config from use_mc to use-mc and format code

ed012302

23 Oct, 2025 1 commit
- issue/473 - the ones and zeros operators · 9b8de584
  pengcheng888 authored Oct 23, 2025
```
Co-authored-by: pengcheng888 <pengcheng@example.com>
```
  9b8de584
29 Sep, 2025 2 commits
- feat: add AWQ dequantize in moore gpu, with test pass · 8896615a
  zhushuang authored Sep 29, 2025
  
  8896615a
- issue/486 Adapt seven operators to Hygon machines. · e698ef6b
  gongchensu authored Sep 29, 2025
```
Co-authored-by: zhuyue <zhuyue@qiyuanlab.com>
```
  e698ef6b
17 Sep, 2025 1 commit
- issue/436: 支持9g7b 4b模型 · 3bdd832e
  zhangyue authored Sep 17, 2025
  
  3bdd832e
15 Sep, 2025 2 commits
- issue/450: remove indexToReducedOffset() in all platforms · 9ef02a16
  Ziminli authored Sep 15, 2025
  
  9ef02a16
- issue/450: change indexToReducedOffset() to indexToOffset in elementwise... · 5e581b8e
  Ziminli authored Sep 15, 2025
```
issue/450: change indexToReducedOffset() to indexToOffset in elementwise framework on CPU, NVIDIA, Cambricon, Metax, Moore, and Kunlun
```
  5e581b8e
03 Sep, 2025 1 commit
- issue/342: delete to · 1cadb2a1
  xgqdut2016 authored Sep 03, 2025
  
  1cadb2a1
02 Sep, 2025 1 commit
- issue/418: 解决 p800 上手写算子引用 sm 上指针的报错问题 · 20488ee7
  zhangyue authored Sep 02, 2025
  
  20488ee7
28 Aug, 2025 1 commit
- fix: implement mudnn handle use to prevent mudnn handle mismatching · f34d4e3b
  zhushuang authored Aug 28, 2025
  
  f34d4e3b
27 Aug, 2025 1 commit
- issue/340: xblas · b92ecc31
  zhangyue authored Aug 27, 2025
  
  b92ecc31
26 Aug, 2025 4 commits
- issue/340: kunlun cublas gemm · a9acf208
  xgqdut2016 authored Aug 05, 2025
  
  a9acf208
- issue/404 修改 #define · 376065bc
  zhangyue authored Aug 26, 2025
  
  376065bc
- issue/404: 删除无用注释 · b95d7c00
  zhangyue authored Aug 26, 2025
  
  b95d7c00
- issue/404: kunlun_common.h 和 kunlun_handle.h 解耦合 · 2fe7052f
  zhangyue authored Aug 26, 2025
  
  2fe7052f
25 Aug, 2025 1 commit
- issue/390: kunlun p800 causal softmax · c5167eb7
  zhangyue authored Aug 22, 2025
  
  c5167eb7
20 Aug, 2025 4 commits
- issue/278 - Cambricon Random Sample · c24a52ea
  thatPepe authored Aug 20, 2025
  
  c24a52ea
- issue/385 删除无用注释 · a2a463db
  zhangyue authored Aug 20, 2025
  
  a2a463db
- issues/385 p800 rmsnorm 支持多精度 · eb3972eb
  zhangyue authored Aug 20, 2025
  
  eb3972eb
- issue/246: add Elementwise and SwiGLU in moore gpu · d4b03cf7
  spike-zhu authored Aug 20, 2025
  
  d4b03cf7
18 Aug, 2025 1 commit
- feat: Add BF16 support gemm in moore gpu and rename existing 'musa' to 'moore' in some files · 66a8eb93
  zhushuang authored Aug 18, 2025
  
  66a8eb93
15 Aug, 2025 1 commit
- fix: add bf16 support and resolve build issues for rms_norm in moore gpu · e1cba119
  zhushuang authored Aug 15, 2025
  
  e1cba119
13 Aug, 2025 1 commit
- issue/214 - Elementwise Support for Cambricon Bang · adbda4c4
  thatPepe authored Aug 13, 2025
  
  adbda4c4
07 Aug, 2025 2 commits

issue/349 P800-elementwise · 72c4dc7c

zhangyue authored Aug 07, 2025

* 适配p800软件栈

* 支持p800上编译手写算子，重构elementwise 算子组件

* fix format

* delete include xtdk_io

* 支持 elementwise float16 bfloat16 数据类型

* 修改自定义 size_t 和 ptrdiff_t 中基础数据类型

* p800 拆分 kernel

72c4dc7c

修改自定义 size_t 和 ptrdiff_t 中基础数据类型 · 1c7792f4
zhangyue authored Aug 07, 2025

1c7792f4

06 Aug, 2025 4 commits
- 支持 elementwise float16 bfloat16 数据类型 · 46a15a1a
  zhangyue authored Aug 06, 2025
  
  46a15a1a
- delete include xtdk_io · 3a7633ba
  zhangyue authored Aug 06, 2025
  
  3a7633ba
- fix format · c3633ba8
  zhangyue authored Aug 06, 2025
  
  c3633ba8
- 支持p800上编译手写算子，重构elementwise 算子组件 · f7e7c7ba
  zhangyue authored Aug 06, 2025
  
  f7e7c7ba
16 Jul, 2025 1 commit
- issue/158/fix: Update handle.cc · c78ca243
  Derui Yang authored Jul 16, 2025
  
  c78ca243
14 Jul, 2025 3 commits
- issue/158/refactor: 支持天数的 pytorch 测试 · d9de5133
  YdrMaster authored Jul 14, 2025
```
Signed-off-by: YdrMaster <ydrml@hotmail.com>
```
  d9de5133
- issue/158/refactor: 分离天数和英伟达的算子定义 · e09a0b7c
  YdrMaster authored Jul 14, 2025
```
Signed-off-by: YdrMaster <ydrml@hotmail.com>
```
  e09a0b7c
- issue/291/style: 根据实际情况将 cuda 改为 nvidia · ceb57c2a
  YdrMaster authored Jul 14, 2025
```
Signed-off-by: YdrMaster <ydrml@hotmail.com>
```
  ceb57c2a
11 Jul, 2025 2 commits
- issue/291/style: 所有 maca 改为 metax · 507be07e
  YdrMaster authored Jul 11, 2025
```
Signed-off-by: YdrMaster <ydrml@hotmail.com>
```
  507be07e
- issue/291/refactor: 适配沐曦 · 05247bb7
  PanZezhong authored Jul 10, 2025
```
Signed-off-by: PanZezhong <panzezhong@qiyuanlab.com>
```
  05247bb7
10 Jul, 2025 1 commit
- issue/291/fix: 兼容 bf16 · 27a13460
  YdrMaster authored Jul 09, 2025
```
Signed-off-by: YdrMaster <ydrml@hotmail.com>
```
  27a13460
09 Jul, 2025 3 commits
- issue/291/build: 为 cudnn 增加一个编译选项，并允许 cuda 使用标记弃用的函数 · f789a52f
  YdrMaster authored Jun 30, 2025
```
Signed-off-by: YdrMaster <ydrml@hotmail.com>
```
  f789a52f
- issue291/refactor: ENABLE_CUDA_API 宏全部替换为 ENABLE_NVIDIA_API · 14a53dee
  PanZezhong authored Jun 30, 2025
```
Signed-off-by: PanZezhong <panzezhong@qiyuanlab.com>
```
  14a53dee
- issue/312: 沐曦BF16编译问题 · 9e0773f6
  Catheriany authored Jul 09, 2025
  
  9e0773f6