-------------------------------------------------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  --------------------------------------------------------------------------------  ------------  
                                                   Name    Self CPU %      Self CPU   CPU total %     CPU total  CPU time avg     Self CUDA   Self CUDA %    CUDA total  CUDA time avg       CPU Mem  Self CPU Mem      CUDA Mem  Self CUDA Mem    # of Calls                                                                      Input Shapes   Total FLOPs  
-------------------------------------------------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  --------------------------------------------------------------------------------  ------------  
                                          ProfilerStep*         0.00%       0.000us         0.00%       0.000us       0.000us     588.955ms       166.72%     588.955ms       9.816ms           0 B           0 B           0 B           0 B            60                                                                                []            --  
                                          ProfilerStep*        50.88%     100.377ms        70.29%     138.660ms      13.866ms       0.000us         0.00%     353.270ms      35.327ms           0 B           0 B           0 B     -11.18 GB            10                                                                                []            --  
                                       aten::_cudnn_rnn         6.31%      12.450ms        21.40%      42.212ms       2.111ms     104.482ms        29.58%     227.892ms      11.395ms           0 B           0 B       4.80 GB     -15.94 GB            20  [[8192, 68, 200], [], [], [588672], [2, 8192, 224], [], [], [], [], [], [], [],             --  
                                         aten::_to_copy         0.03%      58.018us         0.28%     545.711us      27.286us       0.000us         0.00%     197.269ms       9.863ms           0 B           0 B       6.24 GB           0 B            20                                         [[8192, 68, 200], [], [], [], [], [], []]            --  
                                            aten::copy_         0.03%      65.437us         0.08%     152.338us       7.617us     197.269ms        55.84%     197.269ms       9.863ms           0 B           0 B           0 B           0 B            20                                            [[8192, 68, 200], [8192, 68, 200], []]            --  
                         Memcpy HtoD (Pinned -> Device)         0.00%       0.000us         0.00%       0.000us       0.000us     190.069ms        53.80%     190.069ms       6.336ms           0 B           0 B           0 B           0 B            30                                                                                []            --  
                                               aten::to         0.01%      11.277us         0.23%     449.686us      22.484us       0.000us         0.00%     190.004ms       9.500ms           0 B           0 B       4.16 GB           0 B            20                                     [[8192, 68, 200], [], [], [], [], [], [], []]            --  
                                              aten::gru         0.15%     295.008us        11.94%      23.550ms       2.355ms       0.000us         0.00%     117.699ms      11.770ms           0 B           0 B       2.39 GB     -11.23 MB            10                     [[8192, 68, 200], [2, 8192, 224], [], [], [], [], [], [], []]            --  
ampere_fp16_s1688gemm_fp16_128x128_ldg8_f2f_stages_3...         0.00%       0.000us         0.00%       0.000us       0.000us      37.236ms        10.54%      37.236ms      27.380us           0 B           0 B           0 B           0 B          1360                                                                                []            --  
void elemWiseRNNcell<__half, __half, float, (cudnnRN...         0.00%       0.000us         0.00%       0.000us       0.000us      35.180ms         9.96%      35.180ms      25.868us           0 B           0 B           0 B           0 B          1360                                                                                []            --  
ampere_fp16_s1688gemm_fp16_256x64_ldg8_f2f_stages_32...         0.00%       0.000us         0.00%       0.000us       0.000us      32.065ms         9.08%      32.065ms      94.309us           0 B           0 B           0 B           0 B           340                                                                                []            --  
                                             aten::sub_         0.05%      97.451us         0.08%     154.252us      15.425us       9.749ms         2.76%       9.749ms     974.900us           0 B           0 B           0 B           0 B            10                                                [[8192, 68, 200], [1, 1, 200], []]            --  
void at::native::elementwise_kernel<128, 2, at::nati...         0.00%       0.000us         0.00%       0.000us       0.000us       9.749ms         2.76%       9.749ms     974.900us           0 B           0 B           0 B           0 B            10                                                                                []            --  
                                             aten::mul_         0.03%      62.153us         0.05%     104.047us      10.405us       9.692ms         2.74%       9.692ms     969.213us           0 B           0 B           0 B           0 B            10                                                    [[8192, 68, 200], [1, 1, 200]]            --  
void at::native::elementwise_kernel<128, 2, at::nati...         0.00%       0.000us         0.00%       0.000us       0.000us       9.692ms         2.74%       9.692ms     969.213us           0 B           0 B           0 B           0 B            10                                                                                []            --  
void at::native::vectorized_elementwise_kernel<4, at...         0.00%       0.000us         0.00%       0.000us       0.000us       9.552ms         2.70%       9.552ms     238.802us           0 B           0 B           0 B           0 B            40                                                                                []            --  
                                      aten::nan_to_num_         0.02%      30.675us         0.11%     220.652us      22.065us       0.000us         0.00%       9.520ms     952.001us           0 B           0 B           0 B           0 B            10                                                     [[8192, 68, 200], [], [], []]            --  
                                       aten::nan_to_num         0.04%      74.544us         0.10%     189.977us      18.998us       9.520ms         2.69%       9.520ms     952.001us           0 B           0 B           0 B           0 B            10                                    [[8192, 68, 200], [], [], [], [8192, 68, 200]]            --  
                                              aten::mul         0.05%      97.133us         0.07%     142.192us      14.219us       9.368ms         2.65%       9.368ms     936.759us           0 B           0 B       4.16 GB       4.16 GB            10                                                 [[8192, 68, 200], [8192, 1, 200]]  1114112000.000  
void at::native::elementwise_kernel<128, 4, at::nati...         0.00%       0.000us         0.00%       0.000us       0.000us       9.368ms         2.65%       9.368ms     936.759us           0 B           0 B           0 B           0 B            10                                                                                []            --  
void at::native::vectorized_elementwise_kernel<4, at...         0.00%       0.000us         0.00%       0.000us       0.000us       7.492ms         2.12%       7.492ms      74.915us           0 B           0 B           0 B           0 B           100                                                                                []            --  
                                               aten::to         0.01%      15.920us         0.06%     123.222us       6.161us       0.000us         0.00%       7.265ms     363.256us           0 B           0 B       2.08 GB           0 B            20                                                 [[8192, 68, 200], [], [], [], []]            --  
                                       aten::contiguous         0.01%      14.450us         0.09%     169.665us      16.966us       0.000us         0.00%       5.711ms     571.112us           0 B           0 B       2.08 GB           0 B            10                                                             [[68, 8192, 200], []]            --  
                                            aten::clone         0.01%      24.932us         0.08%     155.215us      15.521us       0.000us         0.00%       5.711ms     571.112us           0 B           0 B       2.08 GB           0 B            10                                                             [[68, 8192, 200], []]            --  
                                            aten::copy_         0.02%      45.557us         0.04%      85.396us       8.540us       5.711ms         1.62%       5.711ms     571.112us           0 B           0 B           0 B           0 B            10                                            [[68, 8192, 200], [68, 8192, 200], []]            --  
void at::native::elementwise_kernel<128, 4, at::nati...         0.00%       0.000us         0.00%       0.000us       0.000us       5.711ms         1.62%       5.711ms     571.112us           0 B           0 B           0 B           0 B            10                                                                                []            --  
                                             aten::mean         0.08%     166.141us         0.12%     239.191us      23.919us       5.370ms         1.52%       5.370ms     537.032us           0 B           0 B      62.50 MB      62.50 MB            10                                                     [[8192, 68, 200], [], [], []]            --  
void at::native::reduce_kernel<128, 4, at::native::R...         0.00%       0.000us         0.00%       0.000us       0.000us       5.370ms         1.52%       5.370ms     537.032us           0 B           0 B           0 B           0 B            10                                                                                []            --  
                                              aten::sum         0.52%       1.030ms         0.77%       1.519ms      10.125us     365.598us         0.10%     365.598us       2.437us           0 B           0 B      75.00 KB      75.00 KB           150                                                                      [[8192], []]            --  
void at::native::reduce_kernel<512, 1, at::native::R...         0.00%       0.000us         0.00%       0.000us       0.000us     365.598us         0.10%     365.598us       2.437us           0 B           0 B           0 B           0 B           150                                                                                []            --  
                                       aten::layer_norm         0.03%      58.468us         0.38%     750.467us      37.523us       0.000us         0.00%     309.921us      15.496us           0 B           0 B     147.06 MB     -63.62 MB            20                                           [[8192, 224], [], [224], [224], [], []]            --  
void at::native::(anonymous namespace)::vectorized_l...         0.00%       0.000us         0.00%       0.000us       0.000us     258.560us         0.07%     258.560us      12.928us           0 B           0 B           0 B           0 B            20                                                                                []            --  
                                         aten::_to_copy         0.36%     719.363us         0.86%       1.693ms      24.190us       0.000us         0.00%     247.966us       3.542us           0 B           0 B       2.81 MB           0 B            70                                                  [[8192], [], [], [], [], [], []]            --  
                                            aten::copy_         0.12%     236.719us         0.36%     714.638us      10.209us     247.966us         0.07%     247.966us       3.542us           0 B           0 B           0 B           0 B            70                                                              [[8192], [8192], []]            --  
                                           aten::linear         0.06%     112.242us         0.62%       1.227ms      61.356us       0.000us         0.00%     195.901us       9.795us           0 B           0 B      55.02 MB     -28.00 MB            20                                                  [[8192, 224], [112, 224], [112]]            --  
                                               aten::to         0.16%     316.973us         0.83%       1.646ms      27.426us       0.000us         0.00%     190.686us       3.178us           0 B           0 B       1.88 MB           0 B            60                                              [[8192], [], [], [], [], [], [], []]            --  
                                              aten::mul         0.31%     617.910us         0.55%       1.085ms       6.783us     167.330us         0.05%     167.330us       1.046us           0 B           0 B       4.53 MB       4.53 MB           160                                                                  [[8192], [8192]]   1310720.000  
void at::native::vectorized_elementwise_kernel<4, at...         0.00%       0.000us         0.00%       0.000us       0.000us     163.164us         0.05%     163.164us       1.632us           0 B           0 B           0 B           0 B           100                                                                                []            --  
void at::native::vectorized_elementwise_kernel<4, at...         0.00%       0.000us         0.00%       0.000us       0.000us     147.297us         0.04%     147.297us       1.052us           0 B           0 B           0 B           0 B           140                                                                                []            --  
                                           aten::linear         0.59%       1.155ms         1.88%       3.703ms     185.154us       0.000us         0.00%     137.981us       6.899us           0 B           0 B      13.25 MB     -28.12 MB            20                                                    [[8192, 200], [32, 200], [32]]            --  
-------------------------------------------------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  ------------  --------------------------------------------------------------------------------  ------------  
Self CPU time total: 197.275ms
Self CUDA time total: 353.270ms
