Commits · 422a4e1e158c38b89541089b0ba0663fbfa5d78d · jerrrrry / infinicore

12 Jan, 2026 1 commit
- issue/867 fix page caching api, paged attn support more head dims · 96551cb7
  PanZezhong authored Jan 12, 2026
  
  96551cb7
09 Jan, 2026 2 commits
- issue/867 pass total kv lens as paged attn args · 499b1dc6
  PanZezhong authored Jan 09, 2026
  
  499b1dc6
- issue/810 add common graph op macros · 0fa8805e
  PanZezhong authored Jan 09, 2026
  
  0fa8805e
08 Jan, 2026 1 commit
- issue/867 - feat: adjust paged_attention_prefill interface naming · 0a2839a2
  zhushuang authored Jan 07, 2026
  
  0a2839a2
06 Jan, 2026 1 commit
- issue/810 static compute graph infra · 006d530c
  PanZezhong authored Jan 06, 2026
  
  006d530c
30 Dec, 2025 3 commits
- issue/847 paged attention prefill一段式接口 · 99b940b2
  PanZezhong authored Dec 30, 2025
  
  99b940b2
- issue/847 correct cache_lens naming · e13ad8f9
  PanZezhong authored Dec 30, 2025
  
  e13ad8f9
- issue/848 - feat: add paged attention prefill for nvidia gpu with test pass · 1ba0bcfa
  zhushuang authored Dec 30, 2025
  
  1ba0bcfa
29 Dec, 2025 2 commits
- issue/847-paged caching和atention添加infinicore的接口和测试 · 38078981
  pengcheng888 authored Dec 29, 2025
  
  38078981
- issue/834 - feat: add paged attention for nvidia gpu with test pass · 17299923
  zhushuang authored Dec 29, 2025
  
  17299923
26 Dec, 2025 3 commits
- 在 `include/infinicore/ops.hpp` 中加入 `#include "ops/random_sample.hpp"` · ca6e759f
  Jiacheng Huang authored Dec 26, 2025
  
  ca6e759f
- Issue/840: 英伟达支持Int8 Gemm (#852) · ed04d3e6
  qinyiqun authored Dec 26, 2025
```
* can commit

* can exec sm_90a

* can exec < sm_90

* fix format

* fix format

* 增加测试，测试对标sglang test

* fix format 1

* fix format 2

* add compile option to disable cutlass
```
  ed04d3e6
- Revert "Issue/840: 英伟达Int8 Gemm (#841)" · 458bb997
  PanZezhong1725 authored Dec 26, 2025
```
This reverts commit 25258029.
```
  458bb997
25 Dec, 2025 1 commit
- Issue/840: 英伟达Int8 Gemm (#841) · 25258029
  qinyiqun authored Dec 25, 2025
  
  25258029
24 Dec, 2025 2 commits
- Unify add_rms_norm to always return (normalized_result, add_result) pair. · 2a432b34
  zhuyue authored Dec 24, 2025
  
  2a432b34
- 增加cpu的add rms_norm算子,c++和python接口 · 7d60e5b8
  zhuyue authored Dec 23, 2025
  
  7d60e5b8
22 Dec, 2025 1 commit
- issue/821 添加squeeze算子，完善unsqueeze算子测试 · 8e25feb0
  PanZezhong authored Dec 22, 2025
  
  8e25feb0
18 Dec, 2025 1 commit
- issue/798 - fix operator device handling · 3720127c
  wooway777 authored Dec 17, 2025
  
  3720127c
10 Dec, 2025 1 commit

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 1 commit
- issue/713 - 为c++添加 RowParallelLinear 和 ColParallelLinear · 92472152
  pengcheng888 authored Dec 08, 2025
  
  92472152
06 Dec, 2025 1 commit
- issue/719 add more logs · 11aa0c14
  PanZezhong1725 authored Dec 06, 2025
  
  11aa0c14
05 Dec, 2025 2 commits
- Issue/714 - feat(random_sample): add batch processing interface. · ff84910c
  zhuyue authored Dec 04, 2025
  
  ff84910c
- issue/719 Module 支持单个张量加载 · 420369bd
  PanZezhong authored Dec 05, 2025
  
  420369bd
04 Dec, 2025 1 commit
- issue/709 nn::RoPE forward支持inplace，添加unsqueeze · e238ace2
  PanZezhong authored Dec 04, 2025
  
  e238ace2
03 Dec, 2025 2 commits
- issue/700 support default tensor · 6dd619dd
  PanZezhong1725 authored Dec 03, 2025
  
  6dd619dd
- issue/697: fix load_state_dict · 0c6803c6
  Ceng23333 authored Dec 03, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
  0c6803c6
02 Dec, 2025 1 commit
- issue/682 Parameter supports TP, fix copy_from · ee43bda6
  PanZezhong authored Dec 02, 2025
  
  ee43bda6
01 Dec, 2025 1 commit
- issue/680-为mulmat添加alpha参数算子 · 4916ff8a
  pengcheng888 authored Dec 01, 2025
  
  4916ff8a
21 Nov, 2025 1 commit

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

18 Nov, 2025 1 commit

Issue/571 DeviceEvent (#583) · 79b70e58

thatPepe authored Nov 18, 2025



* issue/571 - introduced the DeviceEvent feature

---------
Co-authored-by: Jiacheng Huang <huangjiacheng0709@outlook.com>

79b70e58

17 Nov, 2025 3 commits
- issue/612 - differentiate mempcy sync/async · ce5ddb69
  wooway777 authored Nov 17, 2025
  
  ce5ddb69
- issue/584-修改变量名,文件名;添加#include<optional>;修改测试代码 · 5c4747cd
  pengcheng888 authored Nov 17, 2025
  
  5c4747cd
- issue/581 - 调整py的linear的实现位置; 测试时,修改为新的调用方式 · 9dbf020e
  pengcheng888 authored Nov 17, 2025
  
  9dbf020e
14 Nov, 2025 1 commit
- issue/591 - fix operator context mismatch · 16854aed
  wooway777 authored Nov 14, 2025
  
  16854aed
13 Nov, 2025 1 commit
- Add random_sample python interface and tests. · 7005dd1b
  zhuyue authored Nov 12, 2025
  
  7005dd1b
12 Nov, 2025 1 commit
- Add mul python interface and tests. · a565b363
  zhuyue authored Nov 12, 2025
  
  a565b363
07 Nov, 2025 1 commit
- issue/545 nn::module::Rope · 289d4002
  Ceng23333 authored Nov 07, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
  289d4002
31 Oct, 2025 2 commits
- do assertion at load_parameter && update Module definition with macros · 777b3233
  Ceng23333 authored Oct 31, 2025
```
Signed-off-by: Ceng23333 <441651826@qq.com>
```
  777b3233
- Add gemm operator python interface and tests. · 001141c9
  zhuyue authored Oct 30, 2025
  
  001141c9