8000 Commits · goog00/InfiniCore · GitHub

Repository navigation

Commits

Commits on Jun 5, 2026
- Merge pull request #1204 from InfiniTensor/issue/1202
  wooway777
  authored
  0098309
- Merge pull request #1203 from InfiniTensor/issue/1201
  wooway777
  authored
  1e14371
- issue/1202 - fix kunlunxin softplus calculation dispatch
  wooway777
  committed
  b850c44
- issue/1201 - support int64 in kunlunxin add
  wooway777
  committed
  5961a54
- Merge pull request #1200 from rubik-hua/llama3_rope
  wooway777
  authored
  4c227c7
Commits on Jun 4, 2026
Commits on Jun 3, 2026
- Merge pull request #1181 from rubik-hua/refactor_rope
  wooway777
  authored
  30cf323
Commits on Jun 2, 2026
- Merge pull request #1188 from InfiniTensor/issue/1187
  wooway777
  authored
  c87ec32
- add random_sample to adapt ascend BF16 format
  Jingbo-gao
  committed
  9122613
- Merge pull request #1190 from InfiniTensor/issue/1189
  PanZezhong1725
  authored
  a026e7a
- issue/1180 refactor(nn): decouple RoPE scaling logic with polymorphic interfaces
  rubik
  committed
  91c67d9
- issue/1189 add sigmoid in infinicore
  PanZezhong1725
  committed
  d8f0b6f
- Merge pull request #1176 from InfiniTensor/issue/1175
  wooway777
  authored
  9980bf6
- Merge pull request #1174 from InfiniTensor/issue/1173
  wooway777
  authored
  010f0d9
Commits on Jun 1, 2026
- issue/1187: success qy device
  xgqdut2016
  committed
  60bcd44
- Merge pull request #1184 from Jingbo-gao/swiglu-ascend-adapt
  wooway777
  authored
  945445e
Commits on May 29, 2026
- Merge pull request #1186 from InfiniTensor/issue/1185
  wooway777
  authored
  8973222
- issue/1185 - ascend aclnn add
  
  Jingbo-gao
  authored and
  wooway777
  committed
  4e59b0e
- add swiglu to adapt ascend BF16 format
  Jingbo-gao
  committed
  3ab2d41
- Merge pull request #1183 from InfiniTensor/issue/1182
  wooway777
  authored
  9b3c2b7
- issue/1182 - fix ascend stream
  wooway777
  committed
  e57b202
Commits on May 27, 2026
- issue/1173 add nv macro guard
  PanZezhong1725
  committed
  4bae1c7
- issue/1175 performance: remove conv algo choosing
  PanZezhong1725
  committed
  97d3762
- issue/1173 implement mha op with flash_attn::mha_fwd
  PanZezhong1725
  committed
  a6422fc
Commits on May 26, 2026
- Merge pull request #1171 from InfiniTensor/issue/1170
  wooway777
  authored
  5e68803
- issue/1170 - update zeros test file docs
  wooway777
  committed
  22c5ddb
Commits on May 25, 2026
- issue/1170 InfinirtMemset, infinicore.zeros
  PanZezhong1725
  committed
  fafd723
Commits on May 19, 2026
- Merge pull request #1165 from InfiniTensor/issue/1142
  wooway777
  authored
  6d07ea6
- issue/1142 - feat: add swiglu referencing nvidia for moore gpu
  spike-zhu
  committed
  d2ec67b
  Copy full SHA for d2ec67b
Commits on May 12, 2026
- Merge pull request #1162 from qinyiqun/linear_refactor
  wooway777
  authored
  90fd438
Commits on May 11, 2026
- feat: add Module::children() to expose direct submodules
  qinyiqun
  committed
  7002314
Commits on May 9, 2026
- Merge pull request #1159 from InfiniTensor/issue/1158
  wooway777
  authored
  5b623d1

0