Commits · f7c5939980170e30dc0332cd62dfaa4802d68f1f · jerrrrry / infinicore

06 Mar, 2026 7 commits
- issue/1031 T1-1-9 · f7c59399
  PanZezhong authored Mar 06, 2026
  
  f7c59399
- Merge pull request #1054 from InfiniTensor/issue/1033c · e7a1b121
  thatPepe authored Mar 06, 2026
```
issue/1033 - fix xmake.lua for fa target
```
  e7a1b121
- issue/1033 - fix xmake.lua for fa target · 74b05cdb
  wooway777 authored Mar 06, 2026
  
  74b05cdb
- Merge pull request #1053 from InfiniTensor/issue/1033xmake · dce99862
  thatPepe authored Mar 06, 2026
```
Issue/1033 patch aten and fa adaptations
```
  dce99862
- issue/1033 - fix nv lua for default backend · d6e44e84
  wooway777 authored Mar 06, 2026
  
  d6e44e84
- issue/1033 - fix argmax compile on hygon and qy · 05d05fd7
  wooway777 authored Mar 06, 2026
  
  05d05fd7
- issue/1033 fix search python lib · e7c743ab
  PanZezhong authored Mar 06, 2026
  
  e7c743ab
05 Mar, 2026 16 commits
- issue/1033 - update readme · 11d8d0bb
  wooway777 authored Mar 05, 2026
  
  11d8d0bb
- issue/1033 - fix format · d790f7b4
  wooway777 authored Mar 05, 2026
  
  d790f7b4
- issue/1033 - replace __C with __INFINI_C · b1ee0a8a
  wooway777 authored Mar 05, 2026
  
  b1ee0a8a
- issue/1033 - fix arg with cudart version · c17e3b83
  wooway777 authored Mar 05, 2026
  
  c17e3b83
- issue/1033 - fix nvidia lua · 9015e384
  wooway777 authored Mar 05, 2026
  
  9015e384
- issue/1033 add fpic ldflag for infinirt and fix python link · e6b6fba5
  PanZezhong authored Mar 05, 2026
  
  e6b6fba5
- issue/1033 add flash-attn compile target · 06362c94
  PanZezhong authored Mar 05, 2026
  
  06362c94
- issue/1033 support stream guard · 515e1eca
  PanZezhong authored Mar 05, 2026
  
  515e1eca
- issue/1033 fix mha_varlen test · 8ab073b4
  PanZezhong authored Mar 05, 2026
  
  8ab073b4
- issue/1033 support flash_attn lib with aten adaptor · f6496d44
  PanZezhong authored Mar 05, 2026
  
  f6496d44
- Merge pull request #1051 from InfiniTensor/issue/1050 · 8d99a8f5
  pengcheng888 authored Mar 05, 2026
```
issue/1050 - fix paged caching and paged prefill on metax
```
  8d99a8f5
- Merge pull request #1049 from InfiniTensor/issue/1048 · 3e5cad10
  thatPepe authored Mar 05, 2026
```
issue/1048: 修复QY机器上的paged_attention_prefill和swiglu编译问题
```
  3e5cad10
- Merge pull request #1047 from InfiniTensor/issue/1035i · 69a74ac5
  thatPepe authored Mar 05, 2026
```
issue/1035 - support both int32 and int64 in kv caching
```
  69a74ac5
- issue/1050 - fix paged caching and paged prefill on metax · 9a0f2505
  wooway777 authored Mar 05, 2026
  
  9a0f2505
- issue/1048: debug qy swiglu and paged_attention_prefill · 832e607e
  xgqdut2016 authored Mar 05, 2026
  
  832e607e
- issue/1035 - support both int32 and int64 in kv caching · a0cbae66
  wooway777 authored Mar 05, 2026
  
  a0cbae66
04 Mar, 2026 6 commits
- Merge pull request #1044 from InfiniTensor/issue/1042 · a9503148
  thatPepe authored Mar 04, 2026
```
issue/1042 - fix address misaligned error of rearrange
```
  a9503148
- Merge pull request #1039 from InfiniTensor/issue/1035 · 83e28ec3
  thatPepe authored Mar 04, 2026
```
issue/1035: 添加NVIDIA平台上的kv_caching算子
```
  83e28ec3
- issue/1035 - kv caching on ali, ilu, hygon, qy, metax · af394a32
  wooway777 authored Mar 03, 2026
  
  af394a32
- Merge pull request #1045 from InfiniTensor/issue/1041 · 811ffab3
  spike-zhu authored Mar 04, 2026
```
issue/1041 - feat: use template to replace int64_t in paged_attention_prefill kernel with test pass
```
  811ffab3
- issue/1042 - fix address misaligned error of rearrange · 75a9b69e
  pengcheng888 authored Mar 04, 2026
  
  75a9b69e
- issue/1041 - feat: use template to replace int64_t in paged_attention_prefill kernel with test pass · f06d6465
  zhushuang authored Mar 03, 2026
  
  f06d6465
03 Mar, 2026 4 commits
- Merge pull request #1046 from InfiniTensor/issue/1032i · b2f915cb
  thatPepe authored Mar 03, 2026
```
issue/1032 - fix block size on iluvatar
```
  b2f915cb
- issue/1032 - fix block size on iluvatar · c3946bf0
  wooway777 authored Mar 03, 2026
  
  c3946bf0
- issue/1035: kv caching on nvidia · c70805c9
  xgqdut2016 authored Mar 02, 2026
  
  c70805c9
- Merge pull request #1038 from InfiniTensor/issue/1032n · abd45713
  thatPepe authored Mar 03, 2026
```
Issue/1032n - cuda swiglu
```
  abd45713
02 Mar, 2026 2 commits
- issue/1032 - adjust swiglu operator.cc · 1443aa67
  wooway777 authored Mar 02, 2026
  
  1443aa67
- Merge pull request #1037 from InfiniTensor/issue/1036 · 90cb1b54
  thatPepe authored Mar 02, 2026
```
Issue/1036 paged caching support strides
```
  90cb1b54
28 Feb, 2026 5 commits
- issue/1036 fix some warnings · 3e1ef507
  PanZezhong authored Feb 28, 2026
  
  3e1ef507
- issue/1036 paged caching support strides · 67425576
  PanZezhong authored Feb 28, 2026
  
  67425576
- Merge pull request #1024 from InfiniTensor/issue/1021 · d8176086
  spike-zhu authored Feb 28, 2026
```
Issue/1021： issue/1021 - feat: support bf16 in infiniccl with mccl
```
  d8176086
- issue/1021 - feat: support bf16 in infiniccl on moore_gpu_arch mp_31 with mccl · 93d261ed
  zhushuang authored Feb 12, 2026
  
  93d261ed
- issue/1032 - follow nv changes on metax (swiglu) · ae49716d
  wooway777 authored Feb 28, 2026
  
  ae49716d