0
代码
代码
Issues
Pull Requests
讨论
Wiki
分析
0
ascend-robotascend-robot[fix] int64 could cause overflow
9306b9f8创建于 4月2日历史提交
文件最后提交记录最后更新时间
AbsKernelNpu.cpp
!2298 [4/N] cleancode Merge pull request !2298 from zhangqiongwen/master_cleancode_0318_02 1 年前
AcosKernelNpu.cpp
!2233 Fix cleancode. Merge pull request !2233 from 王夏夏/master 1 年前
AcoshKernelNpu.cpp
!2305 modify cleancode/4 Merge pull request !2305 from 史佳馨/master 1 年前
AdaptiveAvgPool2dBackwardKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
AdaptiveAvgPool2dKernelNpu.cpp
!2291 Fix codecheck Merge pull request !2291 from 刘嘉巍/dev 1 年前
AdaptiveAvgPool3dBackwardKernelNpu.cpp
!2215 [Codecheck] Fix codecheck. Merge pull request !2215 from 刘嘉巍/x1 1 年前
AdaptiveAvgPool3dKernelNpu.cpp
!2305 modify cleancode/4 Merge pull request !2305 from 史佳馨/master 1 年前
AdaptiveMaxPool2dBackwardKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
AdaptiveMaxPool2dKernelNpu.cpp
!2263 [Codecheck] Fix codecheck. Merge pull request !2263 from 刘嘉巍/deformable 1 年前
AddKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
AddLayerNormBackwardKernelNpu.cpp
!2247 cleancode for ops. Merge pull request !2247 from 王夏夏/master 1 年前
AddLayerNormNpu.cpp
!1713 fix ops * fix ops 1 年前
AddReluKernelNpu.cpp
!2328 cleancode Merge pull request !2328 from DaiFu/2503CleanCode2 1 年前
AddRmsNormKernelNpu.cpp
!2258 [Task] cleancode 1/N. Merge pull request !2258 from will-devil/cleancode-0 1 年前
AddbmmKernelNpu.cpp
!2247 cleancode for ops. Merge pull request !2247 from 王夏夏/master 1 年前
AddcdivKernelNpu.cpp
!2247 cleancode for ops. Merge pull request !2247 from 王夏夏/master 1 年前
AddcmulKernelNpu.cpp
!2305 modify cleancode/4 Merge pull request !2305 from 史佳馨/master 1 年前
AddmmKernelNpu.cpp
!2450 recover storage_size when enqueue throw exception Merge pull request !2450 from 詹昊/master 1 年前
AddmvKernelNpu.cpp
!2204 clean code : infershape file Merge pull request !2204 from wgb/clean_code 1 年前
AddrKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
AffineGridGeneratorBackwardKernelNpu.cpp
!2212 [codecheck] fix codecheck. Merge pull request !2212 from 王夏夏/master 1 年前
AffineGridGeneratorKernelNpu.cpp
!2328 cleancode Merge pull request !2328 from DaiFu/2503CleanCode2 1 年前
AllKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
AmpForeachNonFiniteCheckAndUnscaleKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
AnchorResponseFlagsKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
AnyKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
ApplyAdamKernelNpu.cpp
!2299 adding cleancode modify Merge pull request !2299 from 史佳馨/master 1 年前
ApplyAdamWKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
ArangeKernelNpu.cpp
!3086 The aclop for torch._dim_arange, torch._unsafe_index, torch.index_add, and torch.nn.functional.adaptive_avg_pool1d has been switched to aclnn. Merge pull request !3086 from 史佳馨/master 8 个月前
ArgmaxKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
ArgminKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
ArgsortKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
AsStridedKernelNpu.cpp
[fix] int64 could cause overflow Co-authored-by: chenshuai<chenshuai81@huawei.com> # message auto-generated for no-merge-commit merge: !4632 merge cherry-pick-mr-4534-1775121094636-auto into 26.0.0 [fix] int64 could cause overflow Created-by: culechan Commit-by: chenshuai Merged-by: ascend-robot Description: <!-- PR描述模板更新日期：20260203 --> # 【合入来源】 > <font color="red">**如有社区issue，请关联issue链接**</font>\ > <font color="red">**请勿携带内部流程信息（需求链接、问题单、内部issue等）**</font> - [ ] 需求 - [x] 问题单 A1 run torch_npu test_indexing.py -v -k test_index_put_accumulate_large_tensor failed NativeOpApiFunctions:clone->copy_d2d_dtype_baseformat->npu_stride_copy_out cause overflow - [ ] issue/工单 - [ ] 重构优化 - [ ] 资料更新 # 【修改方案】 change int with int64_t # 【资料变更】 > 请确认是否涉及资料变更。如涉及，需要在PR中体现，并简要说明修改内容。如不涉及，需填写“不涉及” # 【接口变更】 > 请确认是否涉及跨代码仓或者客户面可见的接口变更。如涉及，需要详细说明接口以及对应的变更内容，同时需要在资料中体现。如不涉及，需填写“不涉及” # 【功能验证】 ![image.png](https://raw.gitcode.com/user-images/assets/7403085/d50599e2-bdf5-4e54-ba49-16970bd9a7f2/image.png 'image.png') # 【CheckList】 > PR提交人对以下CheckList自检项进行全量自检，自检通过或不涉及，均修改 [ ] 为 [x] - [ ] 代码注释完备，正确记录错误日志 - [ ] 代码实现进行了返回值、空指针等校验 - [ ] PR标题正确使用类型标签，如：feat、fix、refactor、docs、test等 - [ ] PR持续集成流水线（CI）执行通过，代码检查无异常 See merge request: Ascend/op-plugin!46321 个月前
AsinKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
AsinhKernelNpu.cpp
!2282 Fix codecheck. Merge pull request !2282 from 刘嘉巍/cleancode 1 年前
Atan2KernelNpu.cpp
!2233 Fix cleancode. Merge pull request !2233 from 王夏夏/master 1 年前
AtanKernelNpu.cpp
!2275 fix cleancode Merge pull request !2275 from 詹昊/cleancode1 1 年前
AtanhKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
AvgPool2dBackwardKernelNpu.cpp
!2305 modify cleancode/4 Merge pull request !2305 from 史佳馨/master 1 年前
AvgPool2dKernelNpu.cpp
!2263 [Codecheck] Fix codecheck. Merge pull request !2263 from 刘嘉巍/deformable 1 年前
AvgPool3dBackwardKernelNpu.cpp
!2309 fix cleancode Merge pull request !2309 from 詹昊/master 1 年前
AvgPool3dKernelNpu.cpp
!2218 [1/N]cleancode Format Merge pull request !2218 from zhangqiongwen/master_cleancode_01 1 年前
BaddbmmKernelNpu.cpp
!2309 fix cleancode Merge pull request !2309 from 詹昊/master 1 年前
BatchNMSKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
BatchNormBackwardElemtKernelNpu.cpp
!2209 [Codecheck] Fix codecheck. Merge pull request !2209 from 刘嘉巍/cleancode 1 年前
BatchNormBackwardKernelNpu.cpp
!2326 [Task] cleancode 2/N. Merge pull request !2326 from will-devil/cleancode-1 1 年前
BatchNormBackwardReduceKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BatchNormElemtKernelNpu.cpp
!2282 Fix codecheck. Merge pull request !2282 from 刘嘉巍/cleancode 1 年前
BatchNormGatherStatsUpdateKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BatchNormGatherStatsWithCountsKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BatchNormKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BatchNormReduceKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BatchNormStatsKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BernoulliKernelNpu.cpp
!2305 modify cleancode/4 Merge pull request !2305 from 史佳馨/master 1 年前
BertApplyAdamKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
BinaryCrossEntropyBackwardKernelNpu.cpp
!2265 fix cleancode Merge pull request !2265 from 詹昊/cleancode1 1 年前
BinaryCrossEntropyKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BinaryCrossEntropyWithLogitsKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BincountKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BitwiseAndKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
BitwiseNotKernelNpu.cpp
!2328 cleancode Merge pull request !2328 from DaiFu/2503CleanCode2 1 年前
BitwiseOrKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
BitwiseXorKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
BmmKernelNpu.cpp
!2233 Fix cleancode. Merge pull request !2233 from 王夏夏/master 1 年前
BmmV2KernelNpu.cpp
[950] npu_fuesed_matmul接口新增 & npu_grouped_matmul_add_接口新增参数 Co-authored-by: eternityk<taoyikun@huawei.com> # message auto-generated for no-merge-commit merge: !3446 merge master into master [950] npu_fuesed_matmul接口新增 & npu_grouped_matmul_add_接口新增参数 Created-by: eternityk Commit-by: eternityk Merged-by: ascend-robot Description: **【新功能提交】** **功能描述** npu_fuesed_matmul接口新增 & npu_grouped_matmul_add_接口新增参数 **相关Issue** https://gitee.com/ascend/op-plugin-for-lingqu2.0/commit/5afab9f02e512b1f38362364e2ac2a3b60a9659a?w=1 https://gitee.com/ascend/op-plugin-for-lingqu2.0/commit/426ed212a59383739cbd19bb245aa4858fbbfed7?w=1 https://gitee.com/ascend/op-plugin-for-lingqu2.0/pulls/161/files?w=1 https://gitee.com/ascend/op-plugin-for-lingqu2.0/pulls/175/files?w=1 https://gitee.com/ascend/op-plugin-for-lingqu2.0/pulls/180/files?w=1 **Checklist** [x] 符合编程规范&&代码红线 [x] 具备UT测试用例看护&&测试用例为有效用例（若无测试用例看护请说明原因） [x] 是否涉及对外接口变更 是 [x] 是否涉及官网文档修改 不涉及文档修改，A5资料暂不开放 See merge request: Ascend/op-plugin!34466 个月前
BoundingBoxDecodeKernelNpu.cpp
!2327 modify cleancode/5_1 Merge pull request !2327 from 史佳馨/master 1 年前
BoundingBoxEncodeKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
BroadcastKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
CastKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
CatKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
CdistBackwardKernelNpu.cpp
!2328 cleancode Merge pull request !2328 from DaiFu/2503CleanCode2 1 年前
CdistKernelNpu.cpp
fix the bug in cdist and add the test instance Co-authored-by: jjt_3<1926550527@qq.com> # message auto-generated for no-merge-commit merge: !4085 merge fix_cdist_bug_master into master fix the bug in cdist and add the test instance Created-by: jjt_3 Commit-by: jjt_3 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: **Special notes for your reviewers**: See merge request: Ascend/op-plugin!40854 个月前
CeilKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
CeluKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
ChannelShuffleKernelNpu.cpp
!2247 cleancode for ops. Merge pull request !2247 from 王夏夏/master 1 年前
CiouKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
ClampKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
Col2imKernelNpu.cpp
!3126 [Feature] Delete code for v1.11.0 and v2.0 (2/n) Merge pull request !3126 from 刘嘉巍/foreach 8 个月前
ComplexKernelNpu.cpp
!2327 modify cleancode/5_1 Merge pull request !2327 from 史佳馨/master 1 年前
ConfusionTransposeKernelNpu.cpp
!3126 [Feature] Delete code for v1.11.0 and v2.0 (2/n) Merge pull request !3126 from 刘嘉巍/foreach 8 个月前
ConstantPadNdKernelNpu.cpp
!2093 aclop constantPadNd aicore分支输入维度限制开放至8维 Merge pull request !2093 from weinachuan/master 1 年前
Conv2dBackwardKernelNpu.cpp
!2291 Fix codecheck Merge pull request !2291 from 刘嘉巍/dev 1 年前
Conv2dKernelNpu.cpp
!2079 [Fix] Fix special case in conv1d. Merge pull request !2079 from 刘嘉巍/master 1 年前
Conv3dBackwardKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
Conv3dKernelNpu.cpp
!3101 slow_conv3d Merge pull request !3101 from DaiFu/2508aclop2aclnn_1 8 个月前
ConvDepthWise2dKernelNpu.cpp
!2149 TORCH MAIN SYNC : remove use_const_ref_for_mutable_tensors Merge pull request !2149 from dilililiwhy/main_sync_20250125 1 年前
ConvDepthwise2dBackwardKernelNpu.cpp
!2212 [codecheck] fix codecheck. Merge pull request !2212 from 王夏夏/master 1 年前
ConvTbcBackwardKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
ConvTranspose2dBackwardKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
ConvTranspose2dKernelNpu.cpp
!2204 clean code : infershape file Merge pull request !2204 from wgb/clean_code 1 年前
ConvTranspose3dBackwardKernelNpu.cpp
!2245 [Codecheck] Fix codecheck. Merge pull request !2245 from 刘嘉巍/x1 1 年前
ConvolutionKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
ConvtbcKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
CosKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
CoshKernelNpu.cpp
!2328 cleancode Merge pull request !2328 from DaiFu/2503CleanCode2 1 年前
CountNonZeroKernelNpu.cpp
!2327 modify cleancode/5_1 Merge pull request !2327 from 史佳馨/master 1 年前
CropAndResizeKernelNpu.cpp
!1713 fix ops * fix ops 1 年前
CtcLossBackwardKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
CtcLossKernelNpu.cpp
【950】cast && quant类算子回合 Co-authored-by: 李根<ligen75@h-partners.com> Co-authored-by: Wang_Xing001<wangxing73@huawei.com> Co-authored-by: lilongqianxi<lilongqianxi@h-partners.com> Co-authored-by: 田野<tianye82@huawei.com> Co-authored-by: 季骏<jijun1@huawei.com> Co-authored-by: luo-yongqing<luoyongqing4@h-partners.com> Co-authored-by: 赵珩<zhaoheng32@h-partners.com> # message auto-generated for no-merge-commit merge: !3431 merge pta_plugin_sync into master 【950】cast && quant类算子回合 Created-by: yuanbin_22 Commit-by: 赵珩;luo-yongqing;季骏;lilongqianxi;Wang_Xing001;田野;李根 Merged-by: ascend-robot Description: <!-- Thanks for sending a pull request! --> **What type of PR is this?** > Uncomment only one /kind <> line, hit enter to put that in a new line, and remove leading whitespaces from that line: > > /kind bug > /kind task > /kind feature **What does this PR do / why do we need it**: A5代码回合主线： 1）npu_dtype_cast支持float4和float8数据类型 2）新增npu_gelu_quant，npu_grouped_dynamic_mx_quant，npu_dynamic_mx_quant，npu_dynamic_block_quant接口 3）QuantUpdateScatter 图模式设置输出dtype，以适配hifp8 4）【Quantize】输出y设置 desc dtype **Special notes for your reviewers**: See merge request: Ascend/op-plugin!34316 个月前
CummaxKernelNpu.cpp
!2247 cleancode for ops. Merge pull request !2247 from 王夏夏/master 1 年前