This avoids the following error with GCC 15:
src/tim/transform/ops/../permute_vector.h:41:11: error: 'uint32_t'
does not name a type
41 | virtual uint32_t Rank() const = 0;
| ^~~~~~~~
src/tim/transform/ops/../permute_vector.h:32:1: note: 'uint32_t' is
defined in header '<cstdint>'; this is probably fixable by adding
'#include <cstdint>'
31 | #include <string>
+++ |+#include <cstdint>
32 |
Co-authored-by: Peter Kjellerstedt <pkj@axis.com>
* Refine platform code and samples
1. Support viplite v2 API
2. Unify the Lite and Native platform APIs so that the same code
can run on different platforms through different compilation options.
Type: Code Improvement
Signed-off-by: Kee <xuke537@hotmail.com>
* Fix build error if VSI device API is not supported
Signed-off-by: Kee <xuke537@hotmail.com>
---------
Signed-off-by: Kee <xuke537@hotmail.com>
* Fixed typing error in gather test
Deleted the external output tensor creation in scalar_1d test
Type: Bug Fix
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
* Added ifdef marco for some later added ops
Type: Code Improvement
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
---------
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
add a TIMEOUT for TIM-VX check. If this step fails, there is a high probability that the local board has problems.There is a TIMEOUT to find problems faster.
Backgroud: instance_norm models run crash after latest lay out code
refine
Reason: erase() function will delete the last element if the param is vector.end()
Solution: Check iterator validtion before erase.
Type: Bug Fix
Issue: 37449
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
Change type of shape from int to uint & add ut for broadcast
From github issue https://github.com/VeriSilicon/TIM-VX/issues/376
Define opversion to avoid incompatibility with upper level software
Type: Code Improvement
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
Added Tan Cos operaion
Added reduction param in scatternd_onnx_v16
Added R2X bias in bidirectional_lstm
Type: New Feature
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
* Remove unnecessary compiler flags
* Refactor CMakeLists.txt
* Tweak CMakeLists.txt for libtim_internal
* Tweak CMakeLists.txt for libtim-vx
* Make TIM_VX_ENABLE_TEST defaults to OFF
* Eliminate usage of include_directories
* Fix CI unit test
* Fix warnings relating to inheritance
* Keep graph output order in layout inference
Type: Code Improvement
* Fix typos in layout inference
Type: Code Improvement
---------
authored-by: Xiaoran Weng <Xiaoran.Weng@verisilicon.com>
graph compile will crash when shape is broadcast from 1 to 1,1,1,1
Type: Bug fix
Signed-off-by: Chen <jack.chen@verisilicon.com>
Co-authored-by: Chen <jack.chen@verisilicon.com>
Modify tolerance in some of these op unit tests for StreamProcessor
Type: Bug Fix
Issue: 37103
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
1.Skip elementwise/relational op unittest if input/output have INF which sp cannot
handle
2.Set different tolerance in Layernom/SoftMax/GRU unittest if SP supported
Type: Bug Fix
Issue: 37103
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
Added SetScalar api to support scalar input
Added 2 cases for scalar index Gather
Type: New Feature
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
Resolve the issue of underlying hardware not supporting float16 bias in fc
by converting bias type to float32
Type: Code Improvement
Signed-off-by: Feiyue Chen <Feiyue.Chen@verisilicon.com>
Only NNAPI instance norm spec have scalar gamma and beta, which can not
support by sp, rewrite it into tensor.
Type: Bug Fix
Co-authored-by: zhouheng.zheng <zhouheng.zheng@ouotlook.com>
Colin
Kee
Peter Kjellerstedt
Chen Feiyue
Kainan Cha
Antkillerfarm
zhongzhuonan
Yunshan
Zhouheng Zheng
chxin66
xie-oritek