GitLab

issue/834: add paged attention for nvidia gpu

* issue/818 - Create a test api and add load test case from file feature

* issue/818 - Rename class

issue/860: 在 `include/infinicore/ops.hpp` 中加入 `#include "ops/random_sample.hpp"`

issue/850 - Tensor和device类访问不存在的属性,则抛出异常

* 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

issue/809 add more size classes

* issue/853 - Bug fix: import TestCase and utils class from new file

* issue/853 - import class from framework

Revert "Issue/840: 英伟达Int8 Gemm"

This reverts commit 25258029.

issue/787 - Split run ops test logic and fix kwargs name in report

issue/809 支持锁定、缓存的内存分配器

issue/824 - reduce test sizes

issue/824 - removed redundant batch handling

issue/821 添加squeeze算子，完善unsqueeze算子测试

Issue/563 Add metax support for topkrouter

issue/812: 增加CUDA_ARCH选项与CUTLASS支持

issue/798 - fix operator device handling

issue/792 - add moore threads device sync to tests

Issue/794 - Use native musaEventElapsedTime API.