digraph { node1 [label="Input0 [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node4 [label="Embedding(32320, 1024, padding_idx=0) [forward_compute_time=0.140,backward_compute_time=25.538,activation_size=13107200.0,parameter_size=132382720.0]"] node5 [label="EmuBidirLSTM( (bidir): LSTM(1024, 1024, bidirectional=True) (layer1): LSTM(1024, 1024) (layer2): LSTM(1024, 1024) ) [forward_compute_time=14.126,backward_compute_time=0.019,activation_size=26214400.0,parameter_size=67174400.0]"] node2 [label="Input1 [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node6 [label="Dropout(p=0.2) [forward_compute_time=0.186,backward_compute_time=0.285,activation_size=26214400.0,parameter_size=0.0]"] node7 [label="LSTM(2048, 1024) [forward_compute_time=10.298,backward_compute_time=18.777,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node8 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node9 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node10 [label="Dropout(p=0.2) [forward_compute_time=0.124,backward_compute_time=0.159,activation_size=13107200.0,parameter_size=0.0]"] node11 [label="LSTM(1024, 1024) [forward_compute_time=6.887,backward_compute_time=13.549,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=33587200.0]"] node12 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node13 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node14 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node15 [label="Dropout(p=0.2) [forward_compute_time=0.129,backward_compute_time=0.167,activation_size=13107200.0,parameter_size=0.0]"] node16 [label="LSTM(1024, 1024) [forward_compute_time=6.903,backward_compute_time=13.518,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=33587200.0]"] node17 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node18 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node19 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node20 [label="Dropout(p=0.2) [forward_compute_time=0.122,backward_compute_time=0.160,activation_size=13107200.0,parameter_size=0.0]"] node21 [label="LSTM(1024, 1024) [forward_compute_time=6.889,backward_compute_time=13.594,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=33587200.0]"] node22 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node23 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node24 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node25 [label="Dropout(p=0.2) [forward_compute_time=0.120,backward_compute_time=0.160,activation_size=13107200.0,parameter_size=0.0]"] node26 [label="LSTM(1024, 1024) [forward_compute_time=7.019,backward_compute_time=13.628,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=33587200.0]"] node27 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node28 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node29 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node30 [label="Dropout(p=0.2) [forward_compute_time=0.120,backward_compute_time=0.164,activation_size=13107200.0,parameter_size=0.0]"] node31 [label="LSTM(1024, 1024) [forward_compute_time=6.991,backward_compute_time=13.614,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=33587200.0]"] node32 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node33 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node34 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node35 [label="Dropout(p=0.2) [forward_compute_time=0.120,backward_compute_time=0.169,activation_size=13107200.0,parameter_size=0.0]"] node36 [label="LSTM(1024, 1024) [forward_compute_time=7.055,backward_compute_time=14.112,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=33587200.0]"] node37 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node38 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node39 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node3 [label="Input2 [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node41 [label="Embedding(32320, 1024, padding_idx=0) [forward_compute_time=0.128,backward_compute_time=0.460,activation_size=13107200.0,parameter_size=132382720.0]"] node40 [label="hidden [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node42 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node43 [label="RecurrentAttention( (rnn): LSTM(1024, 1024) (attn): BahdanauAttention( (linear_q): Linear(in_features=1024, out_features=1024, bias=False) (linear_k): Linear(in_features=1024, out_features=1024, bias=False) (dropout): Dropout(p=0) ) (dropout): Dropout(p=0) ) [forward_compute_time=15.210,backward_compute_time=23.086,activation_size=[13107200.0, 524288.0, 524288.0, 13107200.0, 320000.0],parameter_size=41979904.0]"] node44 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node45 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node46 [label="__getitem__(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node47 [label="__getitem__(3) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node48 [label="Dropout(p=0.2) [forward_compute_time=0.104,backward_compute_time=0.262,activation_size=13107200.0,parameter_size=0.0]"] node49 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node50 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node51 [label="LSTM(2048, 1024) [forward_compute_time=10.517,backward_compute_time=18.744,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node52 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node53 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node54 [label="Dropout(p=0.2) [forward_compute_time=0.109,backward_compute_time=0.277,activation_size=13107200.0,parameter_size=0.0]"] node55 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node56 [label="__getitem__(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node57 [label="LSTM(2048, 1024) [forward_compute_time=10.553,backward_compute_time=18.771,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node58 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node59 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node60 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node61 [label="Dropout(p=0.2) [forward_compute_time=0.120,backward_compute_time=0.256,activation_size=13107200.0,parameter_size=0.0]"] node62 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node63 [label="__getitem__(3) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node64 [label="LSTM(2048, 1024) [forward_compute_time=10.725,backward_compute_time=18.802,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node65 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node66 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node67 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node68 [label="Dropout(p=0.2) [forward_compute_time=0.121,backward_compute_time=0.263,activation_size=13107200.0,parameter_size=0.0]"] node69 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node70 [label="__getitem__(4) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node71 [label="LSTM(2048, 1024) [forward_compute_time=10.643,backward_compute_time=18.863,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node72 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node73 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node74 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node75 [label="Dropout(p=0.2) [forward_compute_time=0.131,backward_compute_time=0.263,activation_size=13107200.0,parameter_size=0.0]"] node76 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node77 [label="__getitem__(5) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node78 [label="LSTM(2048, 1024) [forward_compute_time=10.651,backward_compute_time=19.036,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node79 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node80 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node81 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node82 [label="Dropout(p=0.2) [forward_compute_time=0.126,backward_compute_time=0.261,activation_size=13107200.0,parameter_size=0.0]"] node83 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node84 [label="__getitem__(6) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node85 [label="LSTM(2048, 1024) [forward_compute_time=10.659,backward_compute_time=19.145,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node86 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node87 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node88 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node89 [label="Dropout(p=0.2) [forward_compute_time=0.144,backward_compute_time=0.201,activation_size=13107200.0,parameter_size=0.0]"] node90 [label="Concat(2) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node91 [label="__getitem__(7) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=0.0,parameter_size=0.0]"] node92 [label="LSTM(2048, 1024) [forward_compute_time=10.611,backward_compute_time=66.214,activation_size=[13107200.0, 524288.0, 524288.0],parameter_size=50364416.0]"] node93 [label="__getitem__(0) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node94 [label="__getitem__(1) [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=524288.0,parameter_size=0.0]"] node95 [label="Add [forward_compute_time=0.000,backward_compute_time=0.000,activation_size=13107200.0,parameter_size=0.0]"] node96 [label="Classifier( (classifier): Linear(in_features=1024, out_features=32320, bias=True) ) [forward_compute_time=24.782,backward_compute_time=5.373,activation_size=413696000.0,parameter_size=132512000.0]"] node1 -> node4 node4 -> node5 node5 -> node6 node2 -> node5 node2 -> node43 node6 -> node7 node7 -> node8 node7 -> node9 node8 -> node10 node8 -> node14 node10 -> node11 node11 -> node12 node11 -> node13 node12 -> node14 node14 -> node15 node14 -> node19 node15 -> node16 node16 -> node17 node16 -> node18 node17 -> node19 node19 -> node20 node19 -> node24 node20 -> node21 node21 -> node22 node21 -> node23 node22 -> node24 node24 -> node25 node24 -> node29 node25 -> node26 node26 -> node27 node26 -> node28 node27 -> node29 node29 -> node30 node29 -> node34 node30 -> node31 node31 -> node32 node31 -> node33 node32 -> node34 node34 -> node35 node34 -> node39 node35 -> node36 node36 -> node37 node36 -> node38 node37 -> node39 node39 -> node43 node3 -> node41 node41 -> node43 node40 -> node42 node40 -> node50 node40 -> node56 node40 -> node63 node40 -> node70 node40 -> node77 node40 -> node84 node40 -> node91 node42 -> node43 node43 -> node44 node43 -> node45 node43 -> node46 node43 -> node47 node44 -> node48 node46 -> node49 node46 -> node55 node46 -> node62 node46 -> node69 node46 -> node76 node46 -> node83 node46 -> node90 node48 -> node49 node49 -> node51 node50 -> node51 node51 -> node52 node51 -> node53 node52 -> node54 node52 -> node60 node54 -> node55 node55 -> node57 node56 -> node57 node57 -> node58 node57 -> node59 node58 -> node60 node60 -> node61 node60 -> node67 node61 -> node62 node62 -> node64 node63 -> node64 node64 -> node65 node64 -> node66 node65 -> node67 node67 -> node68 node67 -> node74 node68 -> node69 node69 -> node71 node70 -> node71 node71 -> node72 node71 -> node73 node72 -> node74 node74 -> node75 node74 -> node81 node75 -> node76 node76 -> node78 node77 -> node78 node78 -> node79 node78 -> node80 node79 -> node81 node81 -> node82 node81 -> node88 node82 -> node83 node83 -> node85 node84 -> node85 node85 -> node86 node85 -> node87 node86 -> node88 node88 -> node89 node88 -> node95 node89 -> node90 node90 -> node92 node91 -> node92 node92 -> node93 node92 -> node94 node93 -> node95 node95 -> node96 }