You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.

chengduo 4140fe11a4 Open fuse optimization ops (#18741) ... * open fuse optimization ops test=develop		6 years ago
..
fuse_optimizer_ops_pass	Open fuse optimization ops (#18741)	6 years ago
memory_optimize_pass	Feature/mem opt pass refactor (#18735)	6 years ago
mkldnn	Fix Pooling output scale (#18186)	6 years ago
multi_devices_graph_pass	Open fuse optimization ops (#18741)	6 years ago
CMakeLists.txt	Make fuse_optimizer_op_pass also work when the model contains sparse gradients. (#18664)	6 years ago
attention_lstm_fuse_pass.cc	Add setting Scope function for the graph class (#17417)	6 years ago
attention_lstm_fuse_pass.h	…
coalesce_grad_tensor_pass.cc	Open fuse optimization ops (#18741)	6 years ago
coalesce_grad_tensor_pass.h	Make fuse_optimizer_op_pass also work when the model contains sparse gradients. (#18664)	6 years ago
conv_affine_channel_fuse_pass.cc	…
conv_affine_channel_fuse_pass.h	…
conv_bn_fuse_pass.cc	…
conv_bn_fuse_pass.h	…
conv_elementwise_add2_act_fuse.cc	…
conv_elementwise_add2_act_fuse_pass.cc	…
conv_elementwise_add2_act_fuse_pass.h	…
conv_elementwise_add_act_fuse_pass.cc	…
conv_elementwise_add_act_fuse_pass.h	…
conv_elementwise_add_fuse_pass.cc	…
conv_elementwise_add_fuse_pass.h	…
delete_quant_dequant_op_pass.cc	TRT: Support set dynamic range in int8 mode. (#17524)	6 years ago
delete_quant_dequant_op_pass.h	TRT: Support set dynamic range in int8 mode. (#17524)	6 years ago
embedding_fc_lstm_fuse_pass.cc	Add setting Scope function for the graph class (#17417)	6 years ago
embedding_fc_lstm_fuse_pass.h	…
fc_fuse_pass.cc	TRT: Support set dynamic range in int8 mode. (#17524)	6 years ago
fc_fuse_pass.h	…
fc_fuse_pass_tester.cc	…
fc_gru_fuse_pass.cc	Add setting Scope function for the graph class (#17417)	6 years ago
fc_gru_fuse_pass.h	…
fc_lstm_fuse_pass.cc	Add setting Scope function for the graph class (#17417)	6 years ago
fc_lstm_fuse_pass.h	…
fillconstant_elementwisemul_fuse.cc	…
fillconstant_elementwisemul_fuse.h	…
fuse_elewise_add_act_pass.cc	Enhance fused_elementwise_activation op and add python api in contrib.layers (#17236)	6 years ago
fuse_elewise_add_act_pass.h	…
fuse_pass_base.cc	Add setting Scope function for the graph class (#17417)	6 years ago
fuse_pass_base.h	…
fuse_relu_depthwise_conv_pass.cc	…
fuse_relu_depthwise_conv_pass.h	…
graph.cc	Add backward and optimizer operator dependency pass. (#17746)	6 years ago
graph.h	…
graph_helper.cc	Make fuse_optimizer_op_pass also work when the model contains sparse gradients. (#18664)	6 years ago
graph_helper.h	Make fuse_optimizer_op_pass also work when the model contains sparse gradients. (#18664)	6 years ago
graph_helper_test.cc	…
graph_pattern_detector.cc	Update trt5 for paddle-trt (#18645)	6 years ago
graph_pattern_detector.h	Inference: fix mask rcnn model diff, optim memory usage, memory leak. (#18532)	6 years ago
graph_pattern_detector_tester.cc	…
graph_printer.h	Open fuse optimization ops (#18741)	6 years ago
graph_test.cc	…
graph_to_program_pass.cc	…
graph_to_program_pass.h	…
graph_to_program_pass_test.cc	…
graph_traits.cc	…
graph_traits.h	…
graph_viz_pass.cc	Open fuse optimization ops (#18741)	6 years ago
graph_viz_pass.h	…
identity_scale_op_clean_pass.cc	…
identity_scale_op_clean_pass.h	…
infer_clean_graph_pass.cc	…
is_test_pass.cc	…
is_test_pass.h	…
is_test_pass_tester.cc	…
lock_free_optimize_pass.cc	…
lock_free_optimize_pass.h	Remove attribute in Allocator::Allocate (#17878)	6 years ago
multi_batch_merge_pass.cc	…
multi_batch_merge_pass.h	…
ngraph_subgraph_pass.cc	Capi for a ngraph engine (#17037)	6 years ago
ngraph_subgraph_pass.h	Capi for a ngraph engine (#17037)	6 years ago
node.cc	…
node.h	Feature/mem opt pass refactor (#18735)	6 years ago
node_test.cc	…
pass.cc	Open fuse optimization ops (#18741)	6 years ago
pass.h	Open fuse optimization ops (#18741)	6 years ago
pass_builder.cc	Open fuse optimization ops (#18741)	6 years ago
pass_builder.h	…
pass_test.cc	Feature/mem opt pass refactor (#18735)	6 years ago
quant_conv2d_dequant_fuse_pass.cc	TRT: Support set dynamic range in int8 mode. (#17524)	6 years ago
quant_conv2d_dequant_fuse_pass.h	…
repeated_fc_relu_fuse_pass.cc	…
repeated_fc_relu_fuse_pass.h	…
runtime_context_cache_pass.cc	…
runtime_context_cache_pass.h	…
seq_concat_fc_fuse_pass.cc	…
seq_concat_fc_fuse_pass.h	…
seqconv_eltadd_relu_fuse_pass.cc	Add setting Scope function for the graph class (#17417)	6 years ago
seqconv_eltadd_relu_fuse_pass.h	…
seqpool_concat_fuse_pass.cc	…
seqpool_concat_fuse_pass.h	…
seqpool_concat_fuse_pass_tester.cc	…
shuffle_channel_detect_pass.cc	Cherry-pick benchmark related changes from release/1.4 (#17156)	6 years ago
shuffle_channel_detect_pass.h	Cherry-pick benchmark related changes from release/1.4 (#17156)	6 years ago
simplify_anakin_priorbox_detection_out_pass.cc	…
simplify_anakin_priorbox_detection_out_pass.h	…
squared_mat_sub_fuse_pass.cc	…
squared_mat_sub_fuse_pass.h	…
sync_batch_norm_pass.cc	Open fuse optimization ops (#18741)	6 years ago
sync_batch_norm_pass_tester.cc	Code Clean: Move all pass to paddle::framework::ir (#17228)	6 years ago
transpose_flatten_concat_fuse_pass.cc	…
transpose_flatten_concat_fuse_pass.h	…