Commits · 215d1932485d8bbde199e304bb389aa7bddbd96d · jerrrrry / infinicore

19 Dec, 2025 1 commit
- issue/563 - 调整#include位置 · 812f6726
  pengcheng888 authored Dec 19, 2025
  
  812f6726
18 Dec, 2025 1 commit
- issue/798 - fix operator device handling · 3720127c
  wooway777 authored Dec 17, 2025
  
  3720127c
17 Dec, 2025 1 commit
- Issue/794 - Use native musaEventElapsedTime API. · 8739920b
  zhuyue authored Dec 17, 2025
  
  8739920b
11 Dec, 2025 2 commits
- issue/744: kunlun softplus · f817c394
  zhangyue authored Dec 09, 2025
```
issue/744: softplus

issue/744: kunlun softplus

issue/744: delete F64
```
  f817c394
- issue/753: kunlun gelu kernel; delete template kernel Instantiation · 579bb1bf
  zhangyue authored Dec 11, 2025
  
  579bb1bf
10 Dec, 2025 2 commits

issue/746: 修复causal_softmax在长宽在1024边缘的计算错误 · 0b6bdab0
Ceng23333 authored Dec 10, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
0b6bdab0

Issue/739 在cpu, nvidia, metax, moore threads支持batched rope (#743) · 51beebc6

thatPepe authored Dec 10, 2025

* issue/739 - support batched RoPE on Nvidia and CPU

* issue/739 - metax, moore batched rope

* issue/739 - adjust metax flags

* issue/739 - added a rope module interface to forward inplace in output tensor

51beebc6

09 Dec, 2025 1 commit
- issue/741 暴露 infinicore_cpp_api 类继承 · c74dfaea
  PanZezhong authored Dec 09, 2025
  
  c74dfaea
08 Dec, 2025 2 commits
- issue/722 - adjusted cuda rearrange for shape (8, 4, 20, 64) · 0eb27e6e
  wooway777 authored Dec 05, 2025
  
  0eb27e6e
- issue/713 - 为c++添加 RowParallelLinear 和 ColParallelLinear · 92472152
  pengcheng888 authored Dec 08, 2025
  
  92472152
06 Dec, 2025 2 commits
- issue/730 - optimized debug · 16a80bc6
  wooway777 authored Dec 06, 2025
  
  16a80bc6
- issue/719 add more logs · 11aa0c14
  PanZezhong1725 authored Dec 06, 2025
  
  11aa0c14
05 Dec, 2025 1 commit
- issue/719 Module 支持单个张量加载 · 420369bd
  PanZezhong authored Dec 05, 2025
  
  420369bd
04 Dec, 2025 3 commits
- issue/709 nn::RoPE forward支持inplace，添加unsqueeze · e238ace2
  PanZezhong authored Dec 04, 2025
  
  e238ace2
- Issue/705 - Refactor infinirt multi-device support. · 35e73b83
  zhuyue authored Dec 04, 2025
  
  35e73b83
- issue/704 - add ccl supprt for maca with mc api · 6433bf2b
  crapromer authored Dec 04, 2025
  
  6433bf2b
03 Dec, 2025 3 commits

issue/700 算子执行时根据张量 set device， CPU时无操作 · fa149d69
PanZezhong authored Dec 03, 2025

fa149d69
issue/697: fix load_state_dict · 0c6803c6
Ceng23333 authored Dec 03, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
0c6803c6

issue/695 c++ infinicore::nn::module支持bf16 · 986bb179

Ceng authored Dec 03, 2025



* issue/695 c++ infinicore::nn::module支持bf16
Signed-off-by: Ceng23333 <441651826@qq.com>

* embedding support f16
Signed-off-by: Ceng23333 <441651826@qq.com>

* resolve comments
Signed-off-by: Ceng23333 <441651826@qq.com>

* fix format
Signed-off-by: Ceng23333 <441651826@qq.com>

---------
Signed-off-by: Ceng23333 <441651826@qq.com>

986bb179

02 Dec, 2025 1 commit
- issue/682 Parameter supports TP, fix copy_from · ee43bda6
  PanZezhong authored Dec 02, 2025
  
  ee43bda6
01 Dec, 2025 4 commits
- issue/691: 修复InfiniLM依赖问题 · c973b0b9
  Ceng23333 authored Dec 01, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
  c973b0b9
- issue/688: 修复issue 634引入编译问题 · 1c25a902
  Ceng23333 authored Dec 01, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
  1c25a902
- issue/634: InfiniCore 支持InfiniLM Llama模型适配 (#668) · 2f3f4076
  Ceng authored Dec 01, 2025
```
* issue/634: InfiniCore 支持InfiniLM Llama模型适配
Signed-off-by: Ceng23333 <441651826@qq.com>

* .
Signed-off-by: Ceng23333 <441651826@qq.com>

---------
Signed-off-by: Ceng23333 <441651826@qq.com>
```
  2f3f4076
- issue/680-为mulmat添加alpha参数算子 · 4916ff8a
  pengcheng888 authored Dec 01, 2025
  
  4916ff8a
29 Nov, 2025 1 commit
- issue/563 Add metax support for topkrouter · a15aa367
  Zhao Shijie authored Nov 28, 2025
  
  a15aa367
28 Nov, 2025 6 commits
- issue/676: fixing · 6cf95c95
  zhangyue authored Nov 28, 2025
  
  6cf95c95
- issue/676: fix format · ce10d777
  zhangyue authored Nov 28, 2025
  
  ce10d777
- issue/676 format · 74aeb4f4
  zhangyue authored Nov 28, 2025
  
  74aeb4f4
- issue/676: format · c1af9783
  zhangyue authored Nov 28, 2025
  
  c1af9783
- issue/676: kunlun topkrouter · 5584035d
  zhangyue authored Nov 28, 2025
  
  5584035d
- issue/614: fix device class in python · 5f0f80d6
  zhangyue authored Nov 18, 2025
  
  5f0f80d6
26 Nov, 2025 1 commit
- Issue/670 - fix: correct macro for mccub/hccub conditional compilation. · 904a9254
  zhuyue authored Nov 26, 2025
  
  904a9254
22 Nov, 2025 3 commits
- Issue/661 - Support run.py --bench · ed95eb04
  zhuyue authored Nov 22, 2025
  
  ed95eb04
- Issue/658 - Add Moore platform support for add, mul, and silu operations · 33d0f769
  zhuyue authored Nov 22, 2025
```
- Implement Moore backend for add, mul, and silu elementwise operations
- Filter unsupported dtypes (BF16, F64) for Moore platform in tests
```
  33d0f769
- issue/656 - updated readme · 5e2b003c
  wooway777 authored Nov 22, 2025
  
  5e2b003c
21 Nov, 2025 5 commits
- Issue/654 - Update CUB API usage for CUDA 12.9+ compatibility · 10572e55
  zhuyue authored Nov 21, 2025
  
  10572e55
- Issue/654 - Update CUB API usage for CUDA 12.9+ compatibility · 9a9f0982
  zhuyue authored Nov 21, 2025
  
  9a9f0982
- Issue/648 - fix: fix metax compilation for tanh and layer_norm operations. · d18b77a0
  zhuyue authored Nov 21, 2025
  
  d18b77a0
- Issue/645 - Fix metax add rms_norm operators. · d93e352b
  zhuyue authored Nov 21, 2025
  
  d93e352b
- ISSUE/628 适配QY C610 GPU，增加编译选项，适配已有算子。添加bge类模型所需的算子， (#629) · 85bc98ac
  qinyiqun authored Nov 21, 2025
```
* ISSUE/628 适配QY C610 GPU，增加编译选项，适配已有算子。添加bge类模型所需的算子，包括gelu,layer_norm，lp_norm(支持l1，l2 norm)，relu，softmax，tanh。

---------
Co-authored-by: xgqdut2016 <kenan_gewei@163.com>
Co-authored-by: xgqdut2016 <140036308+xgqdut2016@users.noreply.github.com>
```
  85bc98ac