master

cann-robotfix: linear ACLNN NZ runner storageShape对齐(非batch+batch)及950 linear golden修正(einsum改NPU计算转CPU; dequant去掉out_data_type保持fp32)

11aded7a创建于 8 天前历史提交

文件	最后提交记录	最后更新时间
linear_aclnn_runner.cpp	fix: linear ACLNN NZ runner storageShape对齐(非batch+batch)及950 linear golden修正(einsum改NPU计算转CPU; dequant去掉out_data_type保持fp32) Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2133 merge fix/linear-aclnn-nz-storageshape into master fix: linear ACLNN NZ runner storageShape对齐(非batch+batch)及950 linear golden修正(einsum改NPU计算转CPU; dequant去掉out_data_type保持fp32) Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 fix: linear ACLNN NZ runner storageShape对齐(非batch+batch)及950 linear golden修正(einsum改NPU计算转CPU; dequant去掉out_data_type保持fp32) ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2133	8 天前
linear_aclnn_runner.h	修改版权声明为标准声明格式 Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2073 merge lisensefix into master 修改版权声明为标准声明格式 Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 修改版权声明为标准声明格式<a href="https://gitcode.com/user-attachments/files/7623009/4ebdac8d0b1c46b3bd69d61234eb983f.xlsx" target="_blank">4ebdac8d0b1c46b3bd69d61234eb983f.xlsx</a> ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2073	1 个月前
linear_dequant_aclnn_runner.cpp	支持matmul反量化与叠加偏置 feat(linear): add LinearDequantAclnnRunner for 950 dequant support Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2103 merge a5mm_dequant into master 支持matmul反量化与叠加偏置 feat(linear): add LinearDequantAclnnRunner for 950 dequant support Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 feat(linear): add LinearDequantAclnnRunner for 950 dequant support ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2103	1 个月前
linear_dequant_aclnn_runner.h	支持matmul反量化与叠加偏置 feat(linear): add LinearDequantAclnnRunner for 950 dequant support Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2103 merge a5mm_dequant into master 支持matmul反量化与叠加偏置 feat(linear): add LinearDequantAclnnRunner for 950 dequant support Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 feat(linear): add LinearDequantAclnnRunner for 950 dequant support ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2103	1 个月前
linear_einsum_aclnn_runner.cpp	【A5】【ATB】支持MatmulEinSum算子接入ATB框架 Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2089 merge a5mm_einsum_nz into master 【A5】【ATB】支持MatmulEinSum算子接入ATB框架 Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 【A5】【ATB】支持MatmulEinSum算子接入ATB框架 ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2089	1 个月前
linear_einsum_aclnn_runner.h	【A5】【ATB】支持MatmulEinSum算子接入ATB框架 Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2089 merge a5mm_einsum_nz into master 【A5】【ATB】支持MatmulEinSum算子接入ATB框架 Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 【A5】【ATB】支持MatmulEinSum算子接入ATB框架 ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2089	1 个月前
linear_operation.cpp	支持matmul反量化与叠加偏置 feat(linear): add LinearDequantAclnnRunner for 950 dequant support Co-authored-by: zouyanlong<zouyanlong@h-partners.com> # message auto-generated for no-merge-commit merge: !2103 merge a5mm_dequant into master 支持matmul反量化与叠加偏置 feat(linear): add LinearDequantAclnnRunner for 950 dequant support Created-by: ninihihi Commit-by: zouyanlong Merged-by: cann-robot Description: ## 描述 feat(linear): add LinearDequantAclnnRunner for 950 dequant support ## checklist - [x] 是否通过本地IDE对代码进行静态检查 - [x] 是否通过本地IDE对代码进行格式化处理 - [x] 是否进行空指针校验 - [x] 是否进行返回值校验 - [x] 是否正确释放new/malloc申请的内存 - [x] 是否充分考虑接口的异常场景 - [x] 是否正确记录错误日志 See merge request: cann/ascend-transformer-boost!2103	1 个月前
linear_operation.h	update license Co-authored-by: qq_44359711<caobingjie@huawei.com>	7 个月前
linear_ops_runner.cpp	update license Co-authored-by: qq_44359711<caobingjie@huawei.com>	7 个月前
linear_ops_runner.h	update license Co-authored-by: qq_44359711<caobingjie@huawei.com>	7 个月前