You are viewing a plain text version of this content. The canonical link for it is here.
Posted to commits@singa.apache.org by wa...@apache.org on 2020/10/06 03:07:12 UTC
[singa] branch dev updated (fb9be0f -> 554507a)
This is an automated email from the ASF dual-hosted git repository.
wangwei pushed a change to branch dev
in repository https://gitbox.apache.org/repos/asf/singa.git.
from fb9be0f Adding Operators (#738)
new 85fc6a1 added half cpp backend, half tensor conversion, half tensor math operations, added half to linear, softmax layer, cnn layers, opt module. tested examples naive, module, train_cnn on mlp and cnn, refactored broadcast, added benchmark for fp16 vs fp32
new f20345c update
new b439aac update
new 554507a Merge pull request #797 from dcslin/hp6
The 2205 revisions listed above as "new" are entirely new to this
repository and will be described in separate emails. The revisions
listed as "add" were already present in the repository and have only
been added to this reference.
Summary of changes:
CMakeLists.txt | 2 +
LICENSE | 1 +
examples/cnn/train_cnn.py | 40 +-
examples/mlp/module.py | 47 +-
examples/mlp/native.py | 67 +-
include/half.hpp | 4575 ++++++++++++++++++++++++++++++++
include/singa/core/common.h | 8 +-
include/singa/core/tensor.h | 10 +-
python/singa/autograd.py | 25 +-
python/singa/layer.py | 66 +-
python/singa/opt.py | 24 +-
python/singa/tensor.py | 50 +-
src/api/core_tensor.i | 13 +
src/api/numpy.i | 2 +
src/core/tensor/math_kernel.cu | 498 +++-
src/core/tensor/math_kernel.h | 25 +
src/core/tensor/tensor.cc | 232 +-
src/core/tensor/tensor_math.h | 18 +-
src/core/tensor/tensor_math_cpp.h | 85 +-
src/core/tensor/tensor_math_cuda.h | 412 ++-
src/model/operation/convolution.cc | 21 +-
test/singa/test_operation_benchmark.cc | 147 +
22 files changed, 5947 insertions(+), 421 deletions(-)
create mode 100644 include/half.hpp
create mode 100644 test/singa/test_operation_benchmark.cc