Gu, Jinghui c96c91da22 Improve optimizations for DNNLOWP support on MKL-DNN (#18843) ... Summary: In this PR, the fusion alogrithms are improved to support DNNLOWP. 1. Enabled conv fusions for DNNLOWP 2. Fused order switch op into following quantize op 3. Improve conv+sum fusion to parse larger scope/window 4. re-org fusion code to fix random crash issue due to changing graph Pull Request resolved: https://github.com/pytorch/pytorch/pull/18843 Differential Revision: D15021030 Pulled By: yinghai fbshipit-source-id: 88d2199d9fc69f392de9bfbe1f291e0ebf78ab08		2019-04-20 02:12:06 -07:00
..
__init__.py
adam_op_test.py
blobs_queue_db_test.py
channel_shuffle_op_test.py
concat_split_op_test.py
conv_op_test.py	implement operators for DNNLOWP (#18656)	2019-04-10 12:04:39 -07:00
conv_transpose_test.py	Implementation convolutionTranspose operator for mkl-dnn (#12866)	2019-02-20 17:26:10 -08:00
convfusion_op_test.py	Improve optimizations for DNNLOWP support on MKL-DNN (#18843)	2019-04-20 02:12:06 -07:00
copy_op_test.py
dropout_op_test.py
elementwise_sum_op_test.py	implement operators for DNNLOWP (#18656)	2019-04-10 12:04:39 -07:00
expanddims_squeeze_op_test.py	Impl ExpandDims op and fallback to CPU if needed (#15264)	2019-02-08 12:04:53 -08:00
fc_op_test.py	implement operators for DNNLOWP (#18656)	2019-04-10 12:04:39 -07:00
leaky_relu_op_test.py
LRN_op_test.py
moment_sgd_op_test.py
operator_fallback_op_test.py
order_switch_op_test.py	implement operators for DNNLOWP (#18656)	2019-04-10 12:04:39 -07:00
pool_op_test.py	implement operators for DNNLOWP (#18656)	2019-04-10 12:04:39 -07:00
pre_convert_test.py	support pre-convert filter format for mkldnn training mode and change 'OptimizeForIdeep' to 'OptimizeForMkldnn' (#15171)	2019-03-29 19:00:48 -07:00
relu_op_test.py	implement operators for DNNLOWP (#18656)	2019-04-10 12:04:39 -07:00
reshape_op_test.py	Support int64_t shape data for ideep reshape op	2019-01-30 09:00:09 -08:00
shape_op_test.py	support pre-convert filter format for mkldnn training mode and change 'OptimizeForIdeep' to 'OptimizeForMkldnn' (#15171)	2019-03-29 19:00:48 -07:00
sigmoid_op_test.py
softmax_op_test.py
spatial_bn_op_test.py
test_ideep_net.py
transform_ideep_net.py	support pre-convert filter format for mkldnn training mode and change 'OptimizeForIdeep' to 'OptimizeForMkldnn' (#15171)	2019-03-29 19:00:48 -07:00
weightedsum_op_test.py