Advertisement
Guest User

Untitled

a guest
Aug 23rd, 2016
85
0
Never
Not a member of Pastebin yet? Sign Up, it unlocks many cool features!
Python 425.66 KB | None | 0 0
  1. Function profiling
  2. ==================
  3.   Message: sb/convnet/sb_resnet.py:238
  4.   Time in 1 calls to Function.__call__: 1.692772e-05s
  5.   Time in Function.fn.__call__: 5.006790e-06s (29.577%)
  6.   Time in thunks: 1.907349e-06s (11.268%)
  7.   Total compile time: 9.513402e-02s
  8.     Number of Apply nodes: 1
  9.     Theano Optimizer time: 6.597400e-02s
  10.        Theano validate time: 4.982948e-05s
  11.     Theano Linker time (includes C, CUDA code generation/compiling): 3.857851e-03s
  12.        Import time 2.822876e-03s
  13.        Node make_thunk time 3.688097e-03s
  14.  
  15. Time in all call to theano.grad() 2.656322e+00s
  16. Time since theano import 477.707s
  17. Class
  18. ---
  19. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  20.   100.0%   100.0%       0.000s       1.91e-06s     C        1       1   theano.compile.ops.Shape_i
  21.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  22.  
  23. Ops
  24. ---
  25. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  26.   100.0%   100.0%       0.000s       1.91e-06s     C        1        1   Shape_i{0}
  27.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  28.  
  29. Apply
  30. ------
  31. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  32.   100.0%   100.0%       0.000s       1.91e-06s      1     0   Shape_i{0}(<GpuArrayType<None>(float32, (False,))>)
  33.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  34.  
  35. Optimizer Profile
  36. -----------------
  37.  SeqOptimizer  OPT_FAST_RUN  time 0.066s for 2/1 nodes before/after optimization
  38.    0.001s for callback
  39.        0.000s for fgraph.validate()
  40.    time      - (name, class, index, nodes before, nodes after) - validate time
  41.    0.056725s - ('add_destroy_handler', 'AddDestroyHandler', 23, 1, 1) - 0.000s
  42.    0.001584s - ('canonicalize', 'EquilibriumOptimizer', 6, 2, 1) - 0.000s
  43.      EquilibriumOptimizer      canonicalize
  44.        time 0.001s for 3 passes
  45.        nb nodes (start, end,  max) 2 1 3
  46.        time io_toposort 0.000s
  47.        time in local optimizers 0.001s
  48.        time in global optimizers 0.000s
  49.        time in final optimizers 0.000s
  50.        time in cleanup optimizers 0.000s
  51.         0 - 0.001s 1 (0.000s in global opts, 0.000s io_toposort) - 2 nodes - ('local_shape_to_shape_i', 1)
  52.         1 - 0.000s 1 (0.000s in global opts, 0.000s io_toposort) - 3 nodes - ('local_subtensor_make_vector', 1)
  53.         2 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  54.        times - times applied - nb node created - name:
  55.        0.000s - 1 - 2 - local_shape_to_shape_i
  56.        0.000s - 1 - 0 - local_subtensor_make_vector
  57.        0.000s - in 87 optimization that were not used (display only those with a runtime > 0)
  58.          0.000s - MergeOptimizer
  59.          0.000s - topo_constant_folding
  60.          0.000s - local_useless_subtensor
  61.          0.000s - local_subtensor_remove_broadcastable_index
  62.          0.000s - local_track_shape_i
  63.          0.000s - local_useless_slice
  64.          0.000s - local_subtensor_merge
  65.          0.000s - local_subtensor_lift
  66.          0.000s - local_subtensor_of_dot
  67.          0.000s - local_subtensor_of_alloc
  68.  
  69.      Global, final and clean up optimizers
  70.      Iter 0
  71.              TopoOptimizer  topo_constant_folding
  72.                nb_node (start, end, changed) (3, 3, 0)
  73.                init io_toposort 4.10079956055e-05
  74.                loop time 8.10623168945e-06
  75.                callback_time 0.0
  76.              MergeOptimizer
  77.                nb fail=    0 merged=    0 constant=    0
  78.                time replace=0.00 validate=0.00 callback=0.00
  79.      Iter 1
  80.              TopoOptimizer  topo_constant_folding
  81.                nb_node (start, end, changed) (1, 1, 0)
  82.                init io_toposort 1.90734863281e-05
  83.                loop time 2.14576721191e-06
  84.                callback_time 0.0
  85.              MergeOptimizer
  86.                nb fail=    0 merged=    0 constant=    0
  87.                time replace=0.00 validate=0.00 callback=0.00
  88.      Iter 2
  89.              TopoOptimizer  topo_constant_folding
  90.                nb_node (start, end, changed) (1, 1, 0)
  91.                init io_toposort 1.50203704834e-05
  92.                loop time 1.90734863281e-06
  93.                callback_time 0.0
  94.              MergeOptimizer
  95.                nb fail=    0 merged=    0 constant=    0
  96.                time replace=0.00 validate=0.00 callback=0.00
  97.    0.000795s - ('gpuarray_opt', 'SeqOptimizer', 16, 1, 1) - 0.000s
  98.      SeqOptimizer      gpuarray_opt  time 0.001s for 1/1 nodes before/after optimization
  99.        0.000s for callback
  100.            0.000s for fgraph.validate()
  101.        0.000347s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 1, 1) - 0.000s
  102.          EquilibriumOptimizer          gpuarray_local_optimizations
  103.            time 0.000s for 1 passes
  104.            nb nodes (start, end,  max) 1 1 1
  105.            time io_toposort 0.000s
  106.            time in local optimizers 0.000s
  107.            time in global optimizers 0.000s
  108.            time in final optimizers 0.000s
  109.            time in cleanup optimizers 0.000s
  110.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  111.        0.000221s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 1, 1) - 0.000s
  112.          GraphToGPUOptimizer          gpuarray_graph_optimization
  113.            time io_toposort 0.000s
  114.          Total time taken by local optimizers 0.000s
  115.        0.000092s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 1, 1) - 0.000s
  116.          EquilibriumOptimizer          gpuarray_cut_transfers
  117.            time 0.000s for 1 passes
  118.            nb nodes (start, end,  max) 1 1 1
  119.            time io_toposort 0.000s
  120.            time in local optimizers 0.000s
  121.            time in global optimizers 0.000s
  122.            time in final optimizers 0.000s
  123.            time in cleanup optimizers 0.000s
  124.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  125.        0.000009s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 1, 1) - 0.000s
  126.  
  127.    0.000573s - ('elemwise_fusion', 'SeqOptimizer', 19, 1, 1) - 0.000s
  128.      SeqOptimizer      elemwise_fusion  time 0.000s for 1/1 nodes before/after optimization
  129.        0.000s for callback
  130.            0.000s for fgraph.validate()
  131.        0.000196s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 1, 1) - 0.000s
  132.          FusionOptimizer
  133.           nb_iter 1
  134.           nb_replacement 0
  135.           nb_inconsistency_replace 0
  136.           validate_time 0.0
  137.           callback_time 0.0
  138.           time_toposort 9.53674316406e-07
  139.        0.000190s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 1, 1) - 0.000s
  140.          FusionOptimizer
  141.           nb_iter 1
  142.           nb_replacement 0
  143.           nb_inconsistency_replace 0
  144.           validate_time 0.0
  145.           callback_time 0.0
  146.           time_toposort 1.19209289551e-06
  147.  
  148.    0.000548s - ('specialize', 'EquilibriumOptimizer', 13, 1, 1) - 0.000s
  149.      EquilibriumOptimizer      specialize
  150.        time 0.000s for 1 passes
  151.        nb nodes (start, end,  max) 1 1 1
  152.        time io_toposort 0.000s
  153.        time in local optimizers 0.000s
  154.        time in global optimizers 0.000s
  155.        time in final optimizers 0.000s
  156.        time in cleanup optimizers 0.000s
  157.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  158.      Global, final and clean up optimizers
  159.      Iter 0
  160.              TopoOptimizer  topo_constant_folding
  161.                nb_node (start, end, changed) (1, 1, 0)
  162.                init io_toposort 3.88622283936e-05
  163.                loop time 4.05311584473e-06
  164.                callback_time 0.0
  165.    0.000539s - ('BlasOpt', 'SeqOptimizer', 12, 1, 1) - 0.000s
  166.      SeqOptimizer      BlasOpt  time 0.000s for 1/1 nodes before/after optimization
  167.        0.000s for callback
  168.            0.000s for fgraph.validate()
  169.        0.000153s - ('gemm_optimizer', 'GemmOptimizer', 1, 1, 1) - 0.000s
  170.          GemmOptimizer
  171.           nb_iter 1
  172.           nb_replacement 0
  173.           nb_replacement_didn_t_remove 0
  174.           nb_inconsistency_make 0
  175.           nb_inconsistency_replace 0
  176.           time_canonicalize 0
  177.           time_factor_can 0
  178.           time_factor_list 0
  179.           time_toposort 1.50203704834e-05
  180.           validate_time 0.0
  181.           callback_time 0.0
  182.        0.000111s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 1, 1) - 0.000s
  183.          EquilibriumOptimizer          local_gemm_to_gemv
  184.            time 0.000s for 1 passes
  185.            nb nodes (start, end,  max) 1 1 1
  186.            time io_toposort 0.000s
  187.            time in local optimizers 0.000s
  188.            time in global optimizers 0.000s
  189.            time in final optimizers 0.000s
  190.            time in cleanup optimizers 0.000s
  191.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  192.        0.000052s - ('use_c_blas', 'TopoOptimizer', 4, 1, 1) - 0.000s
  193.          TopoOptimizer  use_c_blas
  194.            nb_node (start, end, changed) (1, 1, 0)
  195.            init io_toposort 1.47819519043e-05
  196.            loop time 1.09672546387e-05
  197.            callback_time 0.0
  198.  LocalOptGroup
  199.  ---------------------
  200.  --- The Optimizer wasn't successful ---
  201.  
  202.       0.000045s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 1, 1) - 0.000s
  203.         TopoOptimizer  local_dot_to_dot22
  204.           nb_node (start, end, changed) (1, 1, 0)
  205.           init io_toposort 1.50203704834e-05
  206.           loop time 5.96046447754e-06
  207.           callback_time 0.0
  208.       0.000041s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 1, 1) - 0.000s
  209.         TopoOptimizer  local_dot22_to_dot22scalar
  210.           nb_node (start, end, changed) (1, 1, 0)
  211.           init io_toposort 1.4066696167e-05
  212.           loop time 5.96046447754e-06
  213.           callback_time 0.0
  214.       0.000039s - ('use_scipy_ger', 'TopoOptimizer', 5, 1, 1) - 0.000s
  215.         TopoOptimizer  scipy_blas
  216.           nb_node (start, end, changed) (1, 1, 0)
  217.           init io_toposort 1.4066696167e-05
  218.           loop time 4.05311584473e-06
  219.           callback_time 0.0
  220.  
  221.   0.000471s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 2, 2) - 0.000s
  222.     EquilibriumOptimizer      scan_eqopt1
  223.       time 0.000s for 1 passes
  224.       nb nodes (start, end,  max) 2 2 2
  225.       time io_toposort 0.000s
  226.       time in local optimizers 0.000s
  227.       time in global optimizers 0.000s
  228.       time in final optimizers 0.000s
  229.       time in cleanup optimizers 0.000s
  230.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  231.     Global, final and clean up optimizers
  232.     Iter 0
  233.             SeqOptimizer              all_pushout_opt  time 0.000s for 2/2 nodes before/after optimization
  234.               0.000s for callback
  235.                   0.000s for fgraph.validate()
  236.               0.000078s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 2, 2) - 0.000s
  237.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  238.                   nb_node (start, end, changed) (2, 2, 0)
  239.                   init io_toposort 3.19480895996e-05
  240.                   loop time 1.28746032715e-05
  241.                   callback_time 0.0
  242.               0.000072s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 2, 2) - 0.000s
  243.               0.000057s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 2, 2) - 0.000s
  244.               0.000056s - ('scan_pushout_dot1', 'PushOutDot1', 3, 2, 2) - 0.000s
  245.               0.000053s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 2, 2) - 0.000s
  246.  
  247.   0.000316s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 1, 1) - 0.000s
  248.     EquilibriumOptimizer      scan_eqopt2
  249.       time 0.000s for 1 passes
  250.       nb nodes (start, end,  max) 1 1 1
  251.       time io_toposort 0.000s
  252.       time in local optimizers 0.000s
  253.       time in global optimizers 0.000s
  254.       time in final optimizers 0.000s
  255.       time in cleanup optimizers 0.000s
  256.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  257.     Global, final and clean up optimizers
  258.     Iter 0
  259.             TopoOptimizer  constant_folding_for_scan2
  260.               nb_node (start, end, changed) (1, 1, 0)
  261.               init io_toposort 1.50203704834e-05
  262.               loop time 2.14576721191e-06
  263.               callback_time 0.0
  264.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  265.               nb_node (start, end, changed) (1, 1, 0)
  266.               init io_toposort 1.59740447998e-05
  267.               loop time 1.90734863281e-06
  268.               callback_time 0.0
  269.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  270.               nb_node (start, end, changed) (1, 1, 0)
  271.               init io_toposort 1.4066696167e-05
  272.               loop time 1.90734863281e-06
  273.               callback_time 0.0
  274.             TopoOptimizer  scanOp_merge_inouts
  275.               nb_node (start, end, changed) (1, 1, 0)
  276.               init io_toposort 1.38282775879e-05
  277.               loop time 5.00679016113e-06
  278.               callback_time 0.0
  279.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  280.               nb_node (start, end, changed) (1, 1, 0)
  281.               init io_toposort 1.38282775879e-05
  282.               loop time 2.14576721191e-06
  283.               callback_time 0.0
  284.   0.000290s - ('stabilize', 'EquilibriumOptimizer', 8, 1, 1) - 0.000s
  285.     EquilibriumOptimizer      stabilize
  286.       time 0.000s for 1 passes
  287.       nb nodes (start, end,  max) 1 1 1
  288.       time io_toposort 0.000s
  289.       time in local optimizers 0.000s
  290.       time in global optimizers 0.000s
  291.       time in final optimizers 0.000s
  292.       time in cleanup optimizers 0.000s
  293.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  294.     Global, final and clean up optimizers
  295.     Iter 0
  296.             TopoOptimizer  topo_constant_folding
  297.               nb_node (start, end, changed) (1, 1, 0)
  298.               init io_toposort 1.50203704834e-05
  299.               loop time 1.90734863281e-06
  300.               callback_time 0.0
  301.   0.000277s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 1, 1) - 0.000s
  302.     FusionOptimizer
  303.      nb_iter 1
  304.      nb_replacement 0
  305.      nb_inconsistency_replace 0
  306.      validate_time 0.0
  307.      callback_time 0.0
  308.      time_toposort 9.53674316406e-07
  309.   0.000261s - ('merge2', 'MergeOptimizer', 22, 1, 1) - 0.000s
  310.     MergeOptimizer
  311.       nb fail=    0 merged=    1 constant=    0
  312.       time replace=0.00 validate=0.00 callback=0.00
  313.   0.000251s - ('ShapeOpt', 'ShapeOptimizer', 2, 2, 2) - 0.000s
  314.   0.000202s - ('merge3', 'MergeOptimizer', 51, 1, 1) - 0.000s
  315.     MergeOptimizer
  316.       nb fail=    0 merged=    0 constant=    0
  317.       time replace=0.00 validate=0.00 callback=0.00
  318.   0.000200s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 1, 1) - 0.000s
  319.     TopoOptimizer  local_dnn_conv_inplace
  320.       nb_node (start, end, changed) (1, 1, 0)
  321.       init io_toposort 2.98023223877e-05
  322.       loop time 4.79221343994e-05
  323.       callback_time 0.0
  324. LocalOptGroup
  325. ---------------------
  326. --- The Optimizer wasn't successful ---
  327.  
  328.    0.000197s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 1, 1) - 0.000s
  329.      FusionOptimizer
  330.       nb_iter 1
  331.       nb_replacement 0
  332.       nb_inconsistency_replace 0
  333.       validate_time 0.0
  334.       callback_time 0.0
  335.       time_toposort 9.53674316406e-07
  336.    0.000178s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 1, 1) - 0.000s
  337.      EquilibriumOptimizer      uncanonicalize
  338.        time 0.000s for 1 passes
  339.        nb nodes (start, end,  max) 1 1 1
  340.        time io_toposort 0.000s
  341.        time in local optimizers 0.000s
  342.        time in global optimizers 0.000s
  343.        time in final optimizers 0.000s
  344.        time in cleanup optimizers 0.000s
  345.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  346.      Global, final and clean up optimizers
  347.      Iter 0
  348.              TopoOptimizer  topo_constant_folding
  349.                nb_node (start, end, changed) (1, 1, 0)
  350.                init io_toposort 2.00271606445e-05
  351.                loop time 1.90734863281e-06
  352.                callback_time 0.0
  353.    0.000169s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 1, 1) - 0.000s
  354.      TopoOptimizer  local_dnna_conv_inplace
  355.        nb_node (start, end, changed) (1, 1, 0)
  356.        init io_toposort 2.00271606445e-05
  357.        loop time 1.28746032715e-05
  358.        callback_time 0.0
  359.  LocalOptGroup
  360.  ---------------------
  361.    time taken - times applied - times tried - name - node_created:
  362.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  363.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  364.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  365.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  366.  
  367.    0.000125s - ('merge1', 'MergeOptimizer', 0, 2, 2) - 0.000s
  368.      MergeOptimizer
  369.        nb fail=    0 merged=    0 constant=    0
  370.        time replace=0.00 validate=0.00 callback=0.00
  371.    0.000124s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 1, 1) - 0.000s
  372.      TopoOptimizer  local_gemm16_inplace
  373.        nb_node (start, end, changed) (1, 1, 0)
  374.        init io_toposort 4.10079956055e-05
  375.        loop time 5.00679016113e-06
  376.        callback_time 0.0
  377.    0.000112s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 1, 1) - 0.000s
  378.      TopoOptimizer  InplaceGpuBlasOpt
  379.        nb_node (start, end, changed) (1, 1, 0)
  380.        init io_toposort 1.59740447998e-05
  381.        loop time 1.00135803223e-05
  382.        callback_time 0.0
  383.  LocalOptGroup
  384.  ---------------------
  385.  --- The Optimizer wasn't successful ---
  386.  
  387.   0.000111s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 1, 1) - 0.000s
  388.     TopoOptimizer  InplaceGpuaBlasOpt
  389.       nb_node (start, end, changed) (1, 1, 0)
  390.       init io_toposort 1.50203704834e-05
  391.       loop time 1.00135803223e-05
  392.       callback_time 0.0
  393. LocalOptGroup
  394. ---------------------
  395.   time taken - times applied - times tried - name - node_created:
  396.   -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  397.   0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  398.  
  399.   0.000108s - ('useless', 'TopoOptimizer', 3, 2, 2) - 0.000s
  400.     TopoOptimizer  useless
  401.       nb_node (start, end, changed) (2, 2, 0)
  402.       init io_toposort 2.90870666504e-05
  403.       loop time 4.60147857666e-05
  404.       callback_time 0.0
  405. LocalOptGroup
  406. ---------------------
  407.   time taken - times applied - times tried - name - node_created:
  408.   -0.000s - 0 - 1 - local_subtensor_of_alloc - 0
  409.   -0.000s - 0 - 1 - local_subtensor_make_vector - 0
  410.   -0.000s - 0 - 1 - local_useless_slice - 0
  411.   0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  412.  
  413.   0.000103s - ('blas_opt_inplace', 'TopoOptimizer', 34, 1, 1) - 0.000s
  414.     TopoOptimizer  InplaceBlasOpt
  415.       nb_node (start, end, changed) (1, 1, 0)
  416.       init io_toposort 1.38282775879e-05
  417.       loop time 1.81198120117e-05
  418.       callback_time 0.0
  419. LocalOptGroup
  420. ---------------------
  421. --- The Optimizer wasn't successful ---
  422.  
  423.    0.000096s - ('specialize_device', 'EquilibriumOptimizer', 17, 1, 1) - 0.000s
  424.      EquilibriumOptimizer      specialize_device
  425.        time 0.000s for 1 passes
  426.        nb nodes (start, end,  max) 1 1 1
  427.        time io_toposort 0.000s
  428.        time in local optimizers 0.000s
  429.        time in global optimizers 0.000s
  430.        time in final optimizers 0.000s
  431.        time in cleanup optimizers 0.000s
  432.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  433.    0.000073s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 1, 1) - 0.000s
  434.      TopoOptimizer  dimshuffle_as_view
  435.        nb_node (start, end, changed) (1, 1, 0)
  436.        init io_toposort 2.8133392334e-05
  437.        loop time 4.05311584473e-06
  438.        callback_time 0.0
  439.    0.000073s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 1, 1) - 0.000s
  440.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  441.        nb_node (start, end, changed) (1, 1, 0)
  442.        init io_toposort 1.59740447998e-05
  443.        loop time 5.00679016113e-06
  444.        callback_time 0.0
  445.    0.000065s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 1, 1) - 0.000s
  446.    0.000061s - ('c_blas_destructive', 'TopoOptimizer', 37, 1, 1) - 0.000s
  447.      TopoOptimizer  c_blas_destructive
  448.        nb_node (start, end, changed) (1, 1, 0)
  449.        init io_toposort 1.59740447998e-05
  450.        loop time 8.10623168945e-06
  451.        callback_time 0.0
  452.  LocalOptGroup
  453.  ---------------------
  454.  --- The Optimizer wasn't successful ---
  455.  
  456.   0.000061s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 1, 1) - 0.000s
  457.     TopoOptimizer  local_inplace_sparseblockgemv
  458.       nb_node (start, end, changed) (1, 1, 0)
  459.       init io_toposort 1.4066696167e-05
  460.       loop time 5.96046447754e-06
  461.       callback_time 0.0
  462.   0.000060s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 1, 1) - 0.000s
  463.     TopoOptimizer  local_inplace_incsubtensor1
  464.       nb_node (start, end, changed) (1, 1, 0)
  465.       init io_toposort 1.50203704834e-05
  466.       loop time 5.00679016113e-06
  467.       callback_time 0.0
  468.   0.000060s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 1, 1) - 0.000s
  469.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  470.       nb_node (start, end, changed) (1, 1, 0)
  471.       init io_toposort 1.50203704834e-05
  472.       loop time 2.86102294922e-06
  473.       callback_time 0.0
  474.   0.000059s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 1, 1) - 0.000s
  475.     TopoOptimizer  local_inplace_sparse_block_outer
  476.       nb_node (start, end, changed) (1, 1, 0)
  477.       init io_toposort 1.4066696167e-05
  478.       loop time 2.14576721191e-06
  479.       callback_time 0.0
  480.   0.000059s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 1, 1) - 0.000s
  481.     TopoOptimizer  local_inplace_sparse_block_gemv
  482.       nb_node (start, end, changed) (1, 1, 0)
  483.       init io_toposort 1.4066696167e-05
  484.       loop time 5.00679016113e-06
  485.       callback_time 0.0
  486.   0.000057s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 2, 2) - 0.000s
  487.     TopoOptimizer  pre_local_IncSubtensor_serialize
  488.       nb_node (start, end, changed) (2, 2, 0)
  489.       init io_toposort 2.78949737549e-05
  490.       loop time 6.91413879395e-06
  491.       callback_time 0.0
  492.   0.000056s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 1, 1) - 0.000s
  493.     TopoOptimizer  local_inplace_sparseblockouter
  494.       nb_node (start, end, changed) (1, 1, 0)
  495.       init io_toposort 1.4066696167e-05
  496.       loop time 3.09944152832e-06
  497.       callback_time 0.0
  498.   0.000056s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 1, 1) - 0.000s
  499.     TopoOptimizer  local_inplace_setsubtensor
  500.       nb_node (start, end, changed) (1, 1, 0)
  501.       init io_toposort 1.4066696167e-05
  502.       loop time 1.90734863281e-06
  503.       callback_time 0.0
  504.   0.000048s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 1, 1) - 0.000s
  505.     TopoOptimizer  local_advincsub1_gpua_inplace
  506.       nb_node (start, end, changed) (1, 1, 0)
  507.       init io_toposort 1.59740447998e-05
  508.       loop time 8.10623168945e-06
  509.       callback_time 0.0
  510.   0.000047s - ('make_ger_destructive', 'TopoOptimizer', 41, 1, 1) - 0.000s
  511.     TopoOptimizer  make_scipy_blas_destructive
  512.       nb_node (start, end, changed) (1, 1, 0)
  513.       init io_toposort 1.50203704834e-05
  514.       loop time 9.05990600586e-06
  515.       callback_time 0.0
  516.   0.000044s - ('cond_make_inplace', 'TopoOptimizer', 47, 1, 1) - 0.000s
  517.     TopoOptimizer  cond_make_inplace
  518.       nb_node (start, end, changed) (1, 1, 0)
  519.       init io_toposort 1.38282775879e-05
  520.       loop time 4.05311584473e-06
  521.       callback_time 0.0
  522.   0.000044s - ('AbstractConvCheck', 'TopoOptimizer', 18, 1, 1) - 0.000s
  523.     TopoOptimizer  AbstractConvCheck
  524.       nb_node (start, end, changed) (1, 1, 0)
  525.       init io_toposort 1.50203704834e-05
  526.       loop time 7.15255737305e-06
  527.       callback_time 0.0
  528.   0.000044s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 1, 1) - 0.000s
  529.     TopoOptimizer  random_make_inplace_mrg
  530.       nb_node (start, end, changed) (1, 1, 0)
  531.       init io_toposort 1.59740447998e-05
  532.       loop time 3.09944152832e-06
  533.       callback_time 0.0
  534.   0.000043s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 1, 1) - 0.000s
  535.     TopoOptimizer  local_fill_to_alloc
  536.       nb_node (start, end, changed) (1, 1, 0)
  537.       init io_toposort 1.50203704834e-05
  538.       loop time 5.00679016113e-06
  539.       callback_time 0.0
  540.   0.000043s - ('local_destructive', 'TopoOptimizer', 48, 1, 1) - 0.000s
  541.     TopoOptimizer  CURAND_destructive
  542.       nb_node (start, end, changed) (1, 1, 0)
  543.       init io_toposort 1.4066696167e-05
  544.       loop time 5.00679016113e-06
  545.       callback_time 0.0
  546.   0.000041s - ('random_make_inplace', 'TopoOptimizer', 49, 1, 1) - 0.000s
  547.     TopoOptimizer  random_make_inplace
  548.       nb_node (start, end, changed) (1, 1, 0)
  549.       init io_toposort 1.4066696167e-05
  550.       loop time 2.86102294922e-06
  551.       callback_time 0.0
  552.   0.000040s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 1, 1) - 0.000s
  553.   0.000039s - ('merge1.2', 'MergeOptimizer', 7, 1, 1) - 0.000s
  554.     MergeOptimizer
  555.       nb fail=    0 merged=    0 constant=    0
  556.       time replace=0.00 validate=0.00 callback=0.00
  557.   0.000038s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 1, 1) - 0.000s
  558.   0.000038s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 1, 1) - 0.000s
  559.   0.000038s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 1, 1) - 0.000s
  560.     TopoOptimizer  local_elemwise_alloc
  561.       nb_node (start, end, changed) (1, 1, 0)
  562.       init io_toposort 1.31130218506e-05
  563.       loop time 3.09944152832e-06
  564.       callback_time 0.0
  565.   0.000025s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 1, 1) - 0.000s
  566.   0.000023s - ('merge1.1', 'MergeOptimizer', 4, 2, 2) - 0.000s
  567.     MergeOptimizer
  568.       nb fail=    0 merged=    0 constant=    0
  569.       time replace=0.00 validate=0.00 callback=0.00
  570.   0.000008s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 1, 1) - 0.000s
  571.  
  572. Here are tips to potentially make your code run faster
  573.                 (if you think of new ones, suggest them on the mailing list).
  574.                 Test them first, as they are not guaranteed to always provide a speedup.
  575.  Sorry, no tip for today.
  576. Function profiling
  577. ==================
  578.  Message: sb/convnet/sb_resnet.py:239
  579.  Time in 1 calls to Function.__call__: 1.287460e-05s
  580.  Time in Function.fn.__call__: 5.006790e-06s (38.889%)
  581.  Time in thunks: 2.145767e-06s (16.667%)
  582.  Total compile time: 2.890587e-02s
  583.    Number of Apply nodes: 1
  584.    Theano Optimizer time: 9.943962e-03s
  585.       Theano validate time: 5.078316e-05s
  586.    Theano Linker time (includes C, CUDA code generation/compiling): 5.960464e-04s
  587.       Import time 0.000000e+00s
  588.       Node make_thunk time 4.749298e-04s
  589.  
  590. Time in all call to theano.grad() 2.656322e+00s
  591. Time since theano import 477.713s
  592. Class
  593. ---
  594. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  595.  100.0%   100.0%       0.000s       2.15e-06s     C        1       1   theano.compile.ops.Shape_i
  596.   ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  597.  
  598. Ops
  599. ---
  600. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  601.  100.0%   100.0%       0.000s       2.15e-06s     C        1        1   Shape_i{0}
  602.   ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  603.  
  604. Apply
  605. ------
  606. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  607.  100.0%   100.0%       0.000s       2.15e-06s      1     0   Shape_i{0}(<GpuArrayType<None>(float32, (False,))>)
  608.   ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  609.  
  610. Optimizer Profile
  611. -----------------
  612. SeqOptimizer  OPT_FAST_RUN  time 0.010s for 2/1 nodes before/after optimization
  613.   0.000s for callback
  614.       0.000s for fgraph.validate()
  615.   time      - (name, class, index, nodes before, nodes after) - validate time
  616.   0.001802s - ('canonicalize', 'EquilibriumOptimizer', 6, 2, 1) - 0.000s
  617.     EquilibriumOptimizer      canonicalize
  618.       time 0.001s for 3 passes
  619.       nb nodes (start, end,  max) 2 1 3
  620.       time io_toposort 0.000s
  621.       time in local optimizers 0.001s
  622.       time in global optimizers 0.000s
  623.       time in final optimizers 0.000s
  624.       time in cleanup optimizers 0.000s
  625.        0 - 0.001s 1 (0.000s in global opts, 0.000s io_toposort) - 2 nodes - ('local_shape_to_shape_i', 1)
  626.        1 - 0.000s 1 (0.000s in global opts, 0.000s io_toposort) - 3 nodes - ('local_subtensor_make_vector', 1)
  627.        2 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  628.       times - times applied - nb node created - name:
  629.       0.000s - 1 - 2 - local_shape_to_shape_i
  630.       0.000s - 1 - 0 - local_subtensor_make_vector
  631.       0.001s - in 87 optimization that were not used (display only those with a runtime > 0)
  632.         0.000s - MergeOptimizer
  633.         0.000s - topo_constant_folding
  634.         0.000s - local_useless_subtensor
  635.         0.000s - local_subtensor_remove_broadcastable_index
  636.         0.000s - local_track_shape_i
  637.         0.000s - local_useless_slice
  638.         0.000s - local_subtensor_lift
  639.         0.000s - local_subtensor_of_dot
  640.         0.000s - local_subtensor_merge
  641.         0.000s - local_subtensor_of_alloc
  642.  
  643.     Global, final and clean up optimizers
  644.     Iter 0
  645.             TopoOptimizer  topo_constant_folding
  646.               nb_node (start, end, changed) (3, 3, 0)
  647.               init io_toposort 4.88758087158e-05
  648.               loop time 5.96046447754e-06
  649.               callback_time 0.0
  650.             MergeOptimizer
  651.               nb fail=    0 merged=    0 constant=    0
  652.               time replace=0.00 validate=0.00 callback=0.00
  653.     Iter 1
  654.             TopoOptimizer  topo_constant_folding
  655.               nb_node (start, end, changed) (1, 1, 0)
  656.               init io_toposort 2.19345092773e-05
  657.               loop time 2.14576721191e-06
  658.               callback_time 0.0
  659.             MergeOptimizer
  660.               nb fail=    0 merged=    0 constant=    0
  661.               time replace=0.00 validate=0.00 callback=0.00
  662.     Iter 2
  663.             TopoOptimizer  topo_constant_folding
  664.               nb_node (start, end, changed) (1, 1, 0)
  665.               init io_toposort 1.69277191162e-05
  666.               loop time 1.90734863281e-06
  667.               callback_time 0.0
  668.             MergeOptimizer
  669.               nb fail=    0 merged=    0 constant=    0
  670.               time replace=0.00 validate=0.00 callback=0.00
  671.   0.000850s - ('gpuarray_opt', 'SeqOptimizer', 16, 1, 1) - 0.000s
  672.     SeqOptimizer      gpuarray_opt  time 0.001s for 1/1 nodes before/after optimization
  673.       0.000s for callback
  674.           0.000s for fgraph.validate()
  675.       0.000375s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 1, 1) - 0.000s
  676.         EquilibriumOptimizer          gpuarray_local_optimizations
  677.           time 0.000s for 1 passes
  678.           nb nodes (start, end,  max) 1 1 1
  679.           time io_toposort 0.000s
  680.           time in local optimizers 0.000s
  681.           time in global optimizers 0.000s
  682.           time in final optimizers 0.000s
  683.           time in cleanup optimizers 0.000s
  684.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  685.       0.000224s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 1, 1) - 0.000s
  686.         GraphToGPUOptimizer          gpuarray_graph_optimization
  687.           time io_toposort 0.000s
  688.         Total time taken by local optimizers 0.000s
  689.       0.000106s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 1, 1) - 0.000s
  690.         EquilibriumOptimizer          gpuarray_cut_transfers
  691.           time 0.000s for 1 passes
  692.           nb nodes (start, end,  max) 1 1 1
  693.           time io_toposort 0.000s
  694.           time in local optimizers 0.000s
  695.           time in global optimizers 0.000s
  696.           time in final optimizers 0.000s
  697.           time in cleanup optimizers 0.000s
  698.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  699.       0.000010s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 1, 1) - 0.000s
  700.  
  701.   0.000619s - ('BlasOpt', 'SeqOptimizer', 12, 1, 1) - 0.000s
  702.     SeqOptimizer      BlasOpt  time 0.000s for 1/1 nodes before/after optimization
  703.       0.000s for callback
  704.           0.000s for fgraph.validate()
  705.       0.000181s - ('gemm_optimizer', 'GemmOptimizer', 1, 1, 1) - 0.000s
  706.         GemmOptimizer
  707.          nb_iter 1
  708.          nb_replacement 0
  709.          nb_replacement_didn_t_remove 0
  710.          nb_inconsistency_make 0
  711.          nb_inconsistency_replace 0
  712.          time_canonicalize 0
  713.          time_factor_can 0
  714.          time_factor_list 0
  715.          time_toposort 1.69277191162e-05
  716.          validate_time 0.0
  717.          callback_time 0.0
  718.       0.000119s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 1, 1) - 0.000s
  719.         EquilibriumOptimizer          local_gemm_to_gemv
  720.           time 0.000s for 1 passes
  721.           nb nodes (start, end,  max) 1 1 1
  722.           time io_toposort 0.000s
  723.           time in local optimizers 0.000s
  724.           time in global optimizers 0.000s
  725.           time in final optimizers 0.000s
  726.           time in cleanup optimizers 0.000s
  727.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  728.       0.000059s - ('use_c_blas', 'TopoOptimizer', 4, 1, 1) - 0.000s
  729.         TopoOptimizer  use_c_blas
  730.           nb_node (start, end, changed) (1, 1, 0)
  731.           init io_toposort 1.69277191162e-05
  732.           loop time 1.21593475342e-05
  733.           callback_time 0.0
  734. LocalOptGroup
  735. ---------------------
  736. --- The Optimizer wasn't successful ---
  737.  
  738.        0.000047s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 1, 1) - 0.000s
  739.          TopoOptimizer  local_dot22_to_dot22scalar
  740.            nb_node (start, end, changed) (1, 1, 0)
  741.            init io_toposort 1.69277191162e-05
  742.            loop time 4.05311584473e-06
  743.            callback_time 0.0
  744.        0.000045s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 1, 1) - 0.000s
  745.          TopoOptimizer  local_dot_to_dot22
  746.            nb_node (start, end, changed) (1, 1, 0)
  747.            init io_toposort 1.71661376953e-05
  748.            loop time 4.05311584473e-06
  749.            callback_time 0.0
  750.        0.000044s - ('use_scipy_ger', 'TopoOptimizer', 5, 1, 1) - 0.000s
  751.          TopoOptimizer  scipy_blas
  752.            nb_node (start, end, changed) (1, 1, 0)
  753.            init io_toposort 1.59740447998e-05
  754.            loop time 3.09944152832e-06
  755.            callback_time 0.0
  756.  
  757.    0.000602s - ('elemwise_fusion', 'SeqOptimizer', 19, 1, 1) - 0.000s
  758.      SeqOptimizer      elemwise_fusion  time 0.000s for 1/1 nodes before/after optimization
  759.        0.000s for callback
  760.            0.000s for fgraph.validate()
  761.        0.000282s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 1, 1) - 0.000s
  762.          FusionOptimizer
  763.           nb_iter 1
  764.           nb_replacement 0
  765.           nb_inconsistency_replace 0
  766.           validate_time 0.0
  767.           callback_time 0.0
  768.           time_toposort 1.90734863281e-06
  769.        0.000192s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 1, 1) - 0.000s
  770.          FusionOptimizer
  771.           nb_iter 1
  772.           nb_replacement 0
  773.           nb_inconsistency_replace 0
  774.           validate_time 0.0
  775.           callback_time 0.0
  776.           time_toposort 2.14576721191e-06
  777.  
  778.    0.000530s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 2, 2) - 0.000s
  779.      EquilibriumOptimizer      scan_eqopt1
  780.        time 0.000s for 1 passes
  781.        nb nodes (start, end,  max) 2 2 2
  782.        time io_toposort 0.000s
  783.        time in local optimizers 0.000s
  784.        time in global optimizers 0.000s
  785.        time in final optimizers 0.000s
  786.        time in cleanup optimizers 0.000s
  787.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  788.      Global, final and clean up optimizers
  789.      Iter 0
  790.              SeqOptimizer              all_pushout_opt  time 0.000s for 2/2 nodes before/after optimization
  791.                0.000s for callback
  792.                    0.000s for fgraph.validate()
  793.                0.000086s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 2, 2) - 0.000s
  794.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  795.                    nb_node (start, end, changed) (2, 2, 0)
  796.                    init io_toposort 3.81469726562e-05
  797.                    loop time 8.10623168945e-06
  798.                    callback_time 0.0
  799.                0.000079s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 2, 2) - 0.000s
  800.                0.000073s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 2, 2) - 0.000s
  801.                0.000067s - ('scan_pushout_dot1', 'PushOutDot1', 3, 2, 2) - 0.000s
  802.                0.000064s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 2, 2) - 0.000s
  803.  
  804.    0.000508s - ('specialize', 'EquilibriumOptimizer', 13, 1, 1) - 0.000s
  805.      EquilibriumOptimizer      specialize
  806.        time 0.000s for 1 passes
  807.        nb nodes (start, end,  max) 1 1 1
  808.        time io_toposort 0.000s
  809.        time in local optimizers 0.000s
  810.        time in global optimizers 0.000s
  811.        time in final optimizers 0.000s
  812.        time in cleanup optimizers 0.000s
  813.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  814.      Global, final and clean up optimizers
  815.      Iter 0
  816.              TopoOptimizer  topo_constant_folding
  817.                nb_node (start, end, changed) (1, 1, 0)
  818.                init io_toposort 1.81198120117e-05
  819.                loop time 1.90734863281e-06
  820.                callback_time 0.0
  821.    0.000492s - ('add_destroy_handler', 'AddDestroyHandler', 23, 1, 1) - 0.000s
  822.    0.000369s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 1, 1) - 0.000s
  823.      EquilibriumOptimizer      scan_eqopt2
  824.        time 0.000s for 1 passes
  825.        nb nodes (start, end,  max) 1 1 1
  826.        time io_toposort 0.000s
  827.        time in local optimizers 0.000s
  828.        time in global optimizers 0.000s
  829.        time in final optimizers 0.000s
  830.        time in cleanup optimizers 0.000s
  831.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  832.      Global, final and clean up optimizers
  833.      Iter 0
  834.              TopoOptimizer  constant_folding_for_scan2
  835.                nb_node (start, end, changed) (1, 1, 0)
  836.                init io_toposort 1.81198120117e-05
  837.                loop time 1.90734863281e-06
  838.                callback_time 0.0
  839.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  840.                nb_node (start, end, changed) (1, 1, 0)
  841.                init io_toposort 1.50203704834e-05
  842.                loop time 2.86102294922e-06
  843.                callback_time 0.0
  844.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  845.                nb_node (start, end, changed) (1, 1, 0)
  846.                init io_toposort 1.59740447998e-05
  847.                loop time 2.14576721191e-06
  848.                callback_time 0.0
  849.              TopoOptimizer  scanOp_merge_inouts
  850.                nb_node (start, end, changed) (1, 1, 0)
  851.                init io_toposort 1.62124633789e-05
  852.                loop time 3.09944152832e-06
  853.                callback_time 0.0
  854.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  855.                nb_node (start, end, changed) (1, 1, 0)
  856.                init io_toposort 1.50203704834e-05
  857.                loop time 1.90734863281e-06
  858.                callback_time 0.0
  859.    0.000337s - ('stabilize', 'EquilibriumOptimizer', 8, 1, 1) - 0.000s
  860.      EquilibriumOptimizer      stabilize
  861.        time 0.000s for 1 passes
  862.        nb nodes (start, end,  max) 1 1 1
  863.        time io_toposort 0.000s
  864.        time in local optimizers 0.000s
  865.        time in global optimizers 0.000s
  866.        time in final optimizers 0.000s
  867.        time in cleanup optimizers 0.000s
  868.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  869.      Global, final and clean up optimizers
  870.      Iter 0
  871.              TopoOptimizer  topo_constant_folding
  872.                nb_node (start, end, changed) (1, 1, 0)
  873.                init io_toposort 1.69277191162e-05
  874.                loop time 2.86102294922e-06
  875.                callback_time 0.0
  876.    0.000257s - ('ShapeOpt', 'ShapeOptimizer', 2, 2, 2) - 0.000s
  877.    0.000251s - ('merge2', 'MergeOptimizer', 22, 1, 1) - 0.000s
  878.      MergeOptimizer
  879.        nb fail=    0 merged=    1 constant=    0
  880.        time replace=0.00 validate=0.00 callback=0.00
  881.    0.000226s - ('blas_opt_inplace', 'TopoOptimizer', 34, 1, 1) - 0.000s
  882.      TopoOptimizer  InplaceBlasOpt
  883.        nb_node (start, end, changed) (1, 1, 0)
  884.        init io_toposort 0.000138998031616
  885.        loop time 1.31130218506e-05
  886.        callback_time 0.0
  887.  LocalOptGroup
  888.  ---------------------
  889.  --- The Optimizer wasn't successful ---
  890.  
  891.   0.000210s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 1, 1) - 0.000s
  892.     FusionOptimizer
  893.      nb_iter 1
  894.      nb_replacement 0
  895.      nb_inconsistency_replace 0
  896.      validate_time 0.0
  897.      callback_time 0.0
  898.      time_toposort 9.53674316406e-07
  899.   0.000188s - ('merge3', 'MergeOptimizer', 51, 1, 1) - 0.000s
  900.     MergeOptimizer
  901.       nb fail=    0 merged=    0 constant=    0
  902.       time replace=0.00 validate=0.00 callback=0.00
  903.   0.000187s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 1, 1) - 0.000s
  904.     FusionOptimizer
  905.      nb_iter 1
  906.      nb_replacement 0
  907.      nb_inconsistency_replace 0
  908.      validate_time 0.0
  909.      callback_time 0.0
  910.      time_toposort 9.53674316406e-07
  911.   0.000160s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 1, 1) - 0.000s
  912.     EquilibriumOptimizer      uncanonicalize
  913.       time 0.000s for 1 passes
  914.       nb nodes (start, end,  max) 1 1 1
  915.       time io_toposort 0.000s
  916.       time in local optimizers 0.000s
  917.       time in global optimizers 0.000s
  918.       time in final optimizers 0.000s
  919.       time in cleanup optimizers 0.000s
  920.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  921.     Global, final and clean up optimizers
  922.     Iter 0
  923.             TopoOptimizer  topo_constant_folding
  924.               nb_node (start, end, changed) (1, 1, 0)
  925.               init io_toposort 1.50203704834e-05
  926.               loop time 1.90734863281e-06
  927.               callback_time 0.0
  928.   0.000138s - ('merge1', 'MergeOptimizer', 0, 2, 2) - 0.000s
  929.     MergeOptimizer
  930.       nb fail=    0 merged=    0 constant=    0
  931.       time replace=0.00 validate=0.00 callback=0.00
  932.   0.000115s - ('useless', 'TopoOptimizer', 3, 2, 2) - 0.000s
  933.     TopoOptimizer  useless
  934.       nb_node (start, end, changed) (2, 2, 0)
  935.       init io_toposort 3.2901763916e-05
  936.       loop time 4.6968460083e-05
  937.       callback_time 0.0
  938. LocalOptGroup
  939. ---------------------
  940.   time taken - times applied - times tried - name - node_created:
  941.   -0.000s - 0 - 1 - local_subtensor_of_alloc - 0
  942.   -0.000s - 0 - 1 - local_subtensor_make_vector - 0
  943.   -0.000s - 0 - 1 - local_useless_slice - 0
  944.   0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  945.  
  946.   0.000107s - ('specialize_device', 'EquilibriumOptimizer', 17, 1, 1) - 0.000s
  947.     EquilibriumOptimizer      specialize_device
  948.       time 0.000s for 1 passes
  949.       nb nodes (start, end,  max) 1 1 1
  950.       time io_toposort 0.000s
  951.       time in local optimizers 0.000s
  952.       time in global optimizers 0.000s
  953.       time in final optimizers 0.000s
  954.       time in cleanup optimizers 0.000s
  955.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  956.   0.000104s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 1, 1) - 0.000s
  957.     TopoOptimizer  InplaceGpuBlasOpt
  958.       nb_node (start, end, changed) (1, 1, 0)
  959.       init io_toposort 1.78813934326e-05
  960.       loop time 1.09672546387e-05
  961.       callback_time 0.0
  962. LocalOptGroup
  963. ---------------------
  964. --- The Optimizer wasn't successful ---
  965.  
  966.    0.000103s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 1, 1) - 0.000s
  967.      TopoOptimizer  local_dnn_conv_inplace
  968.        nb_node (start, end, changed) (1, 1, 0)
  969.        init io_toposort 1.50203704834e-05
  970.        loop time 1.00135803223e-05
  971.        callback_time 0.0
  972.  LocalOptGroup
  973.  ---------------------
  974.  --- The Optimizer wasn't successful ---
  975.  
  976.   0.000099s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 1, 1) - 0.000s
  977.     TopoOptimizer  InplaceGpuaBlasOpt
  978.       nb_node (start, end, changed) (1, 1, 0)
  979.       init io_toposort 1.59740447998e-05
  980.       loop time 1.00135803223e-05
  981.       callback_time 0.0
  982. LocalOptGroup
  983. ---------------------
  984.   time taken - times applied - times tried - name - node_created:
  985.   -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  986.   0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  987.  
  988.   0.000095s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 1, 1) - 0.000s
  989.     TopoOptimizer  local_dnna_conv_inplace
  990.       nb_node (start, end, changed) (1, 1, 0)
  991.       init io_toposort 1.50203704834e-05
  992.       loop time 1.00135803223e-05
  993.       callback_time 0.0
  994. LocalOptGroup
  995. ---------------------
  996.   time taken - times applied - times tried - name - node_created:
  997.   -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  998.   -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  999.   -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  1000.   0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  1001.  
  1002.   0.000065s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 1, 1) - 0.000s
  1003.     TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  1004.       nb_node (start, end, changed) (1, 1, 0)
  1005.       init io_toposort 1.59740447998e-05
  1006.       loop time 3.09944152832e-06
  1007.       callback_time 0.0
  1008.   0.000065s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 2, 2) - 0.000s
  1009.     TopoOptimizer  pre_local_IncSubtensor_serialize
  1010.       nb_node (start, end, changed) (2, 2, 0)
  1011.       init io_toposort 3.21865081787e-05
  1012.       loop time 6.91413879395e-06
  1013.       callback_time 0.0
  1014.   0.000058s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 1, 1) - 0.000s
  1015.     TopoOptimizer  local_inplace_sparse_block_outer
  1016.       nb_node (start, end, changed) (1, 1, 0)
  1017.       init io_toposort 1.4066696167e-05
  1018.       loop time 1.90734863281e-06
  1019.       callback_time 0.0
  1020.   0.000057s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 1, 1) - 0.000s
  1021.     TopoOptimizer  local_gemm16_inplace
  1022.       nb_node (start, end, changed) (1, 1, 0)
  1023.       init io_toposort 1.4066696167e-05
  1024.       loop time 2.86102294922e-06
  1025.       callback_time 0.0
  1026.   0.000057s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 1, 1) - 0.000s
  1027.     TopoOptimizer  local_inplace_incsubtensor1
  1028.       nb_node (start, end, changed) (1, 1, 0)
  1029.       init io_toposort 1.50203704834e-05
  1030.       loop time 1.90734863281e-06
  1031.       callback_time 0.0
  1032.   0.000057s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 1, 1) - 0.000s
  1033.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  1034.       nb_node (start, end, changed) (1, 1, 0)
  1035.       init io_toposort 1.47819519043e-05
  1036.       loop time 1.90734863281e-06
  1037.       callback_time 0.0
  1038.   0.000056s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 1, 1) - 0.000s
  1039.     TopoOptimizer  local_inplace_setsubtensor
  1040.       nb_node (start, end, changed) (1, 1, 0)
  1041.       init io_toposort 1.4066696167e-05
  1042.       loop time 2.86102294922e-06
  1043.       callback_time 0.0
  1044.   0.000055s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 1, 1) - 0.000s
  1045.     TopoOptimizer  local_inplace_sparseblockouter
  1046.       nb_node (start, end, changed) (1, 1, 0)
  1047.       init io_toposort 1.4066696167e-05
  1048.       loop time 1.90734863281e-06
  1049.       callback_time 0.0
  1050.   0.000055s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 1, 1) - 0.000s
  1051.     TopoOptimizer  local_inplace_sparseblockgemv
  1052.       nb_node (start, end, changed) (1, 1, 0)
  1053.       init io_toposort 1.4066696167e-05
  1054.       loop time 1.90734863281e-06
  1055.       callback_time 0.0
  1056.   0.000054s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 1, 1) - 0.000s
  1057.     TopoOptimizer  local_inplace_sparse_block_gemv
  1058.       nb_node (start, end, changed) (1, 1, 0)
  1059.       init io_toposort 1.4066696167e-05
  1060.       loop time 1.90734863281e-06
  1061.       callback_time 0.0
  1062.   0.000050s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 1, 1) - 0.000s
  1063.     TopoOptimizer  dimshuffle_as_view
  1064.       nb_node (start, end, changed) (1, 1, 0)
  1065.       init io_toposort 1.90734863281e-05
  1066.       loop time 2.86102294922e-06
  1067.       callback_time 0.0
  1068.   0.000050s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 1, 1) - 0.000s
  1069.     TopoOptimizer  local_fill_to_alloc
  1070.       nb_node (start, end, changed) (1, 1, 0)
  1071.       init io_toposort 1.69277191162e-05
  1072.       loop time 3.09944152832e-06
  1073.       callback_time 0.0
  1074.   0.000048s - ('c_blas_destructive', 'TopoOptimizer', 37, 1, 1) - 0.000s
  1075.     TopoOptimizer  c_blas_destructive
  1076.       nb_node (start, end, changed) (1, 1, 0)
  1077.       init io_toposort 1.4066696167e-05
  1078.       loop time 9.05990600586e-06
  1079.       callback_time 0.0
  1080. LocalOptGroup
  1081. ---------------------
  1082. --- The Optimizer wasn't successful ---
  1083.  
  1084.    0.000047s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 1, 1) - 0.000s
  1085.    0.000046s - ('merge1.2', 'MergeOptimizer', 7, 1, 1) - 0.000s
  1086.      MergeOptimizer
  1087.        nb fail=    0 merged=    0 constant=    0
  1088.        time replace=0.00 validate=0.00 callback=0.00
  1089.    0.000045s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 1, 1) - 0.000s
  1090.      TopoOptimizer  local_elemwise_alloc
  1091.        nb_node (start, end, changed) (1, 1, 0)
  1092.        init io_toposort 1.69277191162e-05
  1093.        loop time 3.09944152832e-06
  1094.        callback_time 0.0
  1095.    0.000044s - ('AbstractConvCheck', 'TopoOptimizer', 18, 1, 1) - 0.000s
  1096.      TopoOptimizer  AbstractConvCheck
  1097.        nb_node (start, end, changed) (1, 1, 0)
  1098.        init io_toposort 1.59740447998e-05
  1099.        loop time 4.05311584473e-06
  1100.        callback_time 0.0
  1101.    0.000042s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 1, 1) - 0.000s
  1102.      TopoOptimizer  local_advincsub1_gpua_inplace
  1103.        nb_node (start, end, changed) (1, 1, 0)
  1104.        init io_toposort 1.50203704834e-05
  1105.        loop time 3.09944152832e-06
  1106.        callback_time 0.0
  1107.    0.000041s - ('cond_make_inplace', 'TopoOptimizer', 47, 1, 1) - 0.000s
  1108.      TopoOptimizer  cond_make_inplace
  1109.        nb_node (start, end, changed) (1, 1, 0)
  1110.        init io_toposort 1.4066696167e-05
  1111.        loop time 2.86102294922e-06
  1112.        callback_time 0.0
  1113.    0.000040s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 1, 1) - 0.000s
  1114.    0.000039s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 1, 1) - 0.000s
  1115.      TopoOptimizer  random_make_inplace_mrg
  1116.        nb_node (start, end, changed) (1, 1, 0)
  1117.        init io_toposort 1.19209289551e-05
  1118.        loop time 4.05311584473e-06
  1119.        callback_time 0.0
  1120.    0.000039s - ('local_destructive', 'TopoOptimizer', 48, 1, 1) - 0.000s
  1121.      TopoOptimizer  CURAND_destructive
  1122.        nb_node (start, end, changed) (1, 1, 0)
  1123.        init io_toposort 1.28746032715e-05
  1124.        loop time 3.09944152832e-06
  1125.        callback_time 0.0
  1126.    0.000038s - ('make_ger_destructive', 'TopoOptimizer', 41, 1, 1) - 0.000s
  1127.      TopoOptimizer  make_scipy_blas_destructive
  1128.        nb_node (start, end, changed) (1, 1, 0)
  1129.        init io_toposort 1.38282775879e-05
  1130.        loop time 2.14576721191e-06
  1131.        callback_time 0.0
  1132.    0.000038s - ('random_make_inplace', 'TopoOptimizer', 49, 1, 1) - 0.000s
  1133.      TopoOptimizer  random_make_inplace
  1134.        nb_node (start, end, changed) (1, 1, 0)
  1135.        init io_toposort 1.31130218506e-05
  1136.        loop time 2.86102294922e-06
  1137.        callback_time 0.0
  1138.    0.000037s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 1, 1) - 0.000s
  1139.    0.000030s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 1, 1) - 0.000s
  1140.    0.000025s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 1, 1) - 0.000s
  1141.    0.000024s - ('merge1.1', 'MergeOptimizer', 4, 2, 2) - 0.000s
  1142.      MergeOptimizer
  1143.        nb fail=    0 merged=    0 constant=    0
  1144.        time replace=0.00 validate=0.00 callback=0.00
  1145.    0.000006s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 1, 1) - 0.000s
  1146.  
  1147. Here are tips to potentially make your code run faster
  1148.                  (if you think of new ones, suggest them on the mailing list).
  1149.                  Test them first, as they are not guaranteed to always provide a speedup.
  1150.   Sorry, no tip for today.
  1151. Function profiling
  1152. ==================
  1153.   Message: sb/convnet/sb_resnet.py:240
  1154.   Time in 1 calls to Function.__call__: 1.502037e-05s
  1155.   Time in Function.fn.__call__: 3.814697e-06s (25.397%)
  1156.   Time in thunks: 1.907349e-06s (12.698%)
  1157.   Total compile time: 2.578807e-02s
  1158.     Number of Apply nodes: 1
  1159.     Theano Optimizer time: 8.777857e-03s
  1160.        Theano validate time: 4.410744e-05s
  1161.     Theano Linker time (includes C, CUDA code generation/compiling): 6.120205e-04s
  1162.        Import time 0.000000e+00s
  1163.        Node make_thunk time 4.870892e-04s
  1164.  
  1165. Time in all call to theano.grad() 2.656322e+00s
  1166. Time since theano import 477.721s
  1167. Class
  1168. ---
  1169. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  1170.   100.0%   100.0%       0.000s       1.91e-06s     C        1       1   theano.compile.ops.Shape_i
  1171.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  1172.  
  1173. Ops
  1174. ---
  1175. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  1176.   100.0%   100.0%       0.000s       1.91e-06s     C        1        1   Shape_i{0}
  1177.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  1178.  
  1179. Apply
  1180. ------
  1181. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  1182.   100.0%   100.0%       0.000s       1.91e-06s      1     0   Shape_i{0}(<GpuArrayType<None>(float32, (False,))>)
  1183.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  1184.  
  1185. Optimizer Profile
  1186. -----------------
  1187.  SeqOptimizer  OPT_FAST_RUN  time 0.009s for 2/1 nodes before/after optimization
  1188.    0.000s for callback
  1189.        0.000s for fgraph.validate()
  1190.    time      - (name, class, index, nodes before, nodes after) - validate time
  1191.    0.001467s - ('canonicalize', 'EquilibriumOptimizer', 6, 2, 1) - 0.000s
  1192.      EquilibriumOptimizer      canonicalize
  1193.        time 0.001s for 3 passes
  1194.        nb nodes (start, end,  max) 2 1 3
  1195.        time io_toposort 0.000s
  1196.        time in local optimizers 0.000s
  1197.        time in global optimizers 0.000s
  1198.        time in final optimizers 0.000s
  1199.        time in cleanup optimizers 0.000s
  1200.         0 - 0.001s 1 (0.000s in global opts, 0.000s io_toposort) - 2 nodes - ('local_shape_to_shape_i', 1)
  1201.         1 - 0.000s 1 (0.000s in global opts, 0.000s io_toposort) - 3 nodes - ('local_subtensor_make_vector', 1)
  1202.         2 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1203.        times - times applied - nb node created - name:
  1204.        0.000s - 1 - 2 - local_shape_to_shape_i
  1205.        0.000s - 1 - 0 - local_subtensor_make_vector
  1206.        0.000s - in 87 optimization that were not used (display only those with a runtime > 0)
  1207.          0.000s - MergeOptimizer
  1208.          0.000s - topo_constant_folding
  1209.          0.000s - local_useless_subtensor
  1210.          0.000s - local_subtensor_remove_broadcastable_index
  1211.          0.000s - local_track_shape_i
  1212.          0.000s - local_useless_slice
  1213.          0.000s - local_subtensor_lift
  1214.          0.000s - local_subtensor_of_dot
  1215.          0.000s - local_subtensor_merge
  1216.          0.000s - local_subtensor_of_alloc
  1217.  
  1218.      Global, final and clean up optimizers
  1219.      Iter 0
  1220.              TopoOptimizer  topo_constant_folding
  1221.                nb_node (start, end, changed) (3, 3, 0)
  1222.                init io_toposort 3.71932983398e-05
  1223.                loop time 4.05311584473e-06
  1224.                callback_time 0.0
  1225.              MergeOptimizer
  1226.                nb fail=    0 merged=    0 constant=    0
  1227.                time replace=0.00 validate=0.00 callback=0.00
  1228.      Iter 1
  1229.              TopoOptimizer  topo_constant_folding
  1230.                nb_node (start, end, changed) (1, 1, 0)
  1231.                init io_toposort 1.81198120117e-05
  1232.                loop time 2.14576721191e-06
  1233.                callback_time 0.0
  1234.              MergeOptimizer
  1235.                nb fail=    0 merged=    0 constant=    0
  1236.                time replace=0.00 validate=0.00 callback=0.00
  1237.      Iter 2
  1238.              TopoOptimizer  topo_constant_folding
  1239.                nb_node (start, end, changed) (1, 1, 0)
  1240.                init io_toposort 1.69277191162e-05
  1241.                loop time 9.53674316406e-07
  1242.                callback_time 0.0
  1243.              MergeOptimizer
  1244.                nb fail=    0 merged=    0 constant=    0
  1245.                time replace=0.00 validate=0.00 callback=0.00
  1246.    0.000796s - ('gpuarray_opt', 'SeqOptimizer', 16, 1, 1) - 0.000s
  1247.      SeqOptimizer      gpuarray_opt  time 0.001s for 1/1 nodes before/after optimization
  1248.        0.000s for callback
  1249.            0.000s for fgraph.validate()
  1250.        0.000393s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 1, 1) - 0.000s
  1251.          EquilibriumOptimizer          gpuarray_local_optimizations
  1252.            time 0.000s for 1 passes
  1253.            nb nodes (start, end,  max) 1 1 1
  1254.            time io_toposort 0.000s
  1255.            time in local optimizers 0.000s
  1256.            time in global optimizers 0.000s
  1257.            time in final optimizers 0.000s
  1258.            time in cleanup optimizers 0.000s
  1259.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1260.        0.000184s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 1, 1) - 0.000s
  1261.          GraphToGPUOptimizer          gpuarray_graph_optimization
  1262.            time io_toposort 0.000s
  1263.          Total time taken by local optimizers 0.000s
  1264.        0.000086s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 1, 1) - 0.000s
  1265.          EquilibriumOptimizer          gpuarray_cut_transfers
  1266.            time 0.000s for 1 passes
  1267.            nb nodes (start, end,  max) 1 1 1
  1268.            time io_toposort 0.000s
  1269.            time in local optimizers 0.000s
  1270.            time in global optimizers 0.000s
  1271.            time in final optimizers 0.000s
  1272.            time in cleanup optimizers 0.000s
  1273.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1274.        0.000008s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 1, 1) - 0.000s
  1275.  
  1276.    0.000520s - ('elemwise_fusion', 'SeqOptimizer', 19, 1, 1) - 0.000s
  1277.      SeqOptimizer      elemwise_fusion  time 0.000s for 1/1 nodes before/after optimization
  1278.        0.000s for callback
  1279.            0.000s for fgraph.validate()
  1280.        0.000197s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 1, 1) - 0.000s
  1281.          FusionOptimizer
  1282.           nb_iter 1
  1283.           nb_replacement 0
  1284.           nb_inconsistency_replace 0
  1285.           validate_time 0.0
  1286.           callback_time 0.0
  1287.           time_toposort 2.14576721191e-06
  1288.        0.000193s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 1, 1) - 0.000s
  1289.          FusionOptimizer
  1290.           nb_iter 1
  1291.           nb_replacement 0
  1292.           nb_inconsistency_replace 0
  1293.           validate_time 0.0
  1294.           callback_time 0.0
  1295.           time_toposort 1.90734863281e-06
  1296.  
  1297.    0.000513s - ('BlasOpt', 'SeqOptimizer', 12, 1, 1) - 0.000s
  1298.      SeqOptimizer      BlasOpt  time 0.000s for 1/1 nodes before/after optimization
  1299.        0.000s for callback
  1300.            0.000s for fgraph.validate()
  1301.        0.000150s - ('gemm_optimizer', 'GemmOptimizer', 1, 1, 1) - 0.000s
  1302.          GemmOptimizer
  1303.           nb_iter 1
  1304.           nb_replacement 0
  1305.           nb_replacement_didn_t_remove 0
  1306.           nb_inconsistency_make 0
  1307.           nb_inconsistency_replace 0
  1308.           time_canonicalize 0
  1309.           time_factor_can 0
  1310.           time_factor_list 0
  1311.           time_toposort 1.4066696167e-05
  1312.           validate_time 0.0
  1313.           callback_time 0.0
  1314.        0.000105s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 1, 1) - 0.000s
  1315.          EquilibriumOptimizer          local_gemm_to_gemv
  1316.            time 0.000s for 1 passes
  1317.            nb nodes (start, end,  max) 1 1 1
  1318.            time io_toposort 0.000s
  1319.            time in local optimizers 0.000s
  1320.            time in global optimizers 0.000s
  1321.            time in final optimizers 0.000s
  1322.            time in cleanup optimizers 0.000s
  1323.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1324.        0.000047s - ('use_c_blas', 'TopoOptimizer', 4, 1, 1) - 0.000s
  1325.          TopoOptimizer  use_c_blas
  1326.            nb_node (start, end, changed) (1, 1, 0)
  1327.            init io_toposort 1.4066696167e-05
  1328.            loop time 1.00135803223e-05
  1329.            callback_time 0.0
  1330.  LocalOptGroup
  1331.  ---------------------
  1332.  --- The Optimizer wasn't successful ---
  1333.  
  1334.       0.000039s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 1, 1) - 0.000s
  1335.         TopoOptimizer  local_dot_to_dot22
  1336.           nb_node (start, end, changed) (1, 1, 0)
  1337.           init io_toposort 1.50203704834e-05
  1338.           loop time 2.86102294922e-06
  1339.           callback_time 0.0
  1340.       0.000038s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 1, 1) - 0.000s
  1341.         TopoOptimizer  local_dot22_to_dot22scalar
  1342.           nb_node (start, end, changed) (1, 1, 0)
  1343.           init io_toposort 1.4066696167e-05
  1344.           loop time 3.09944152832e-06
  1345.           callback_time 0.0
  1346.       0.000037s - ('use_scipy_ger', 'TopoOptimizer', 5, 1, 1) - 0.000s
  1347.         TopoOptimizer  scipy_blas
  1348.           nb_node (start, end, changed) (1, 1, 0)
  1349.           init io_toposort 1.28746032715e-05
  1350.           loop time 3.09944152832e-06
  1351.           callback_time 0.0
  1352.  
  1353.   0.000504s - ('add_destroy_handler', 'AddDestroyHandler', 23, 1, 1) - 0.000s
  1354.   0.000416s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 1, 1) - 0.000s
  1355.     EquilibriumOptimizer      scan_eqopt2
  1356.       time 0.000s for 1 passes
  1357.       nb nodes (start, end,  max) 1 1 1
  1358.       time io_toposort 0.000s
  1359.       time in local optimizers 0.000s
  1360.       time in global optimizers 0.000s
  1361.       time in final optimizers 0.000s
  1362.       time in cleanup optimizers 0.000s
  1363.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1364.     Global, final and clean up optimizers
  1365.     Iter 0
  1366.             TopoOptimizer  constant_folding_for_scan2
  1367.               nb_node (start, end, changed) (1, 1, 0)
  1368.               init io_toposort 1.8835067749e-05
  1369.               loop time 1.90734863281e-06
  1370.               callback_time 0.0
  1371.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  1372.               nb_node (start, end, changed) (1, 1, 0)
  1373.               init io_toposort 1.4066696167e-05
  1374.               loop time 3.09944152832e-06
  1375.               callback_time 0.0
  1376.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  1377.               nb_node (start, end, changed) (1, 1, 0)
  1378.               init io_toposort 1.38282775879e-05
  1379.               loop time 1.90734863281e-06
  1380.               callback_time 0.0
  1381.             TopoOptimizer  scanOp_merge_inouts
  1382.               nb_node (start, end, changed) (1, 1, 0)
  1383.               init io_toposort 1.31130218506e-05
  1384.               loop time 1.90734863281e-06
  1385.               callback_time 0.0
  1386.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  1387.               nb_node (start, end, changed) (1, 1, 0)
  1388.               init io_toposort 1.31130218506e-05
  1389.               loop time 1.19209289551e-06
  1390.               callback_time 0.0
  1391.   0.000406s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 2, 2) - 0.000s
  1392.     EquilibriumOptimizer      scan_eqopt1
  1393.       time 0.000s for 1 passes
  1394.       nb nodes (start, end,  max) 2 2 2
  1395.       time io_toposort 0.000s
  1396.       time in local optimizers 0.000s
  1397.       time in global optimizers 0.000s
  1398.       time in final optimizers 0.000s
  1399.       time in cleanup optimizers 0.000s
  1400.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1401.     Global, final and clean up optimizers
  1402.     Iter 0
  1403.             SeqOptimizer              all_pushout_opt  time 0.000s for 2/2 nodes before/after optimization
  1404.               0.000s for callback
  1405.                   0.000s for fgraph.validate()
  1406.               0.000066s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 2, 2) - 0.000s
  1407.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  1408.                   nb_node (start, end, changed) (2, 2, 0)
  1409.                   init io_toposort 2.90870666504e-05
  1410.                   loop time 7.15255737305e-06
  1411.                   callback_time 0.0
  1412.               0.000063s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 2, 2) - 0.000s
  1413.               0.000053s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 2, 2) - 0.000s
  1414.               0.000051s - ('scan_pushout_dot1', 'PushOutDot1', 3, 2, 2) - 0.000s
  1415.               0.000050s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 2, 2) - 0.000s
  1416.  
  1417.   0.000383s - ('specialize', 'EquilibriumOptimizer', 13, 1, 1) - 0.000s
  1418.     EquilibriumOptimizer      specialize
  1419.       time 0.000s for 1 passes
  1420.       nb nodes (start, end,  max) 1 1 1
  1421.       time io_toposort 0.000s
  1422.       time in local optimizers 0.000s
  1423.       time in global optimizers 0.000s
  1424.       time in final optimizers 0.000s
  1425.       time in cleanup optimizers 0.000s
  1426.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1427.     Global, final and clean up optimizers
  1428.     Iter 0
  1429.             TopoOptimizer  topo_constant_folding
  1430.               nb_node (start, end, changed) (1, 1, 0)
  1431.               init io_toposort 1.69277191162e-05
  1432.               loop time 1.90734863281e-06
  1433.               callback_time 0.0
  1434.   0.000276s - ('stabilize', 'EquilibriumOptimizer', 8, 1, 1) - 0.000s
  1435.     EquilibriumOptimizer      stabilize
  1436.       time 0.000s for 1 passes
  1437.       nb nodes (start, end,  max) 1 1 1
  1438.       time io_toposort 0.000s
  1439.       time in local optimizers 0.000s
  1440.       time in global optimizers 0.000s
  1441.       time in final optimizers 0.000s
  1442.       time in cleanup optimizers 0.000s
  1443.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1444.     Global, final and clean up optimizers
  1445.     Iter 0
  1446.             TopoOptimizer  topo_constant_folding
  1447.               nb_node (start, end, changed) (1, 1, 0)
  1448.               init io_toposort 1.62124633789e-05
  1449.               loop time 1.90734863281e-06
  1450.               callback_time 0.0
  1451.   0.000249s - ('merge2', 'MergeOptimizer', 22, 1, 1) - 0.000s
  1452.     MergeOptimizer
  1453.       nb fail=    0 merged=    1 constant=    0
  1454.       time replace=0.00 validate=0.00 callback=0.00
  1455.   0.000219s - ('ShapeOpt', 'ShapeOptimizer', 2, 2, 2) - 0.000s
  1456.   0.000191s - ('merge3', 'MergeOptimizer', 51, 1, 1) - 0.000s
  1457.     MergeOptimizer
  1458.       nb fail=    0 merged=    0 constant=    0
  1459.       time replace=0.00 validate=0.00 callback=0.00
  1460.   0.000188s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 1, 1) - 0.000s
  1461.     FusionOptimizer
  1462.      nb_iter 1
  1463.      nb_replacement 0
  1464.      nb_inconsistency_replace 0
  1465.      validate_time 0.0
  1466.      callback_time 0.0
  1467.      time_toposort 9.53674316406e-07
  1468.   0.000187s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 1, 1) - 0.000s
  1469.     FusionOptimizer
  1470.      nb_iter 1
  1471.      nb_replacement 0
  1472.      nb_inconsistency_replace 0
  1473.      validate_time 0.0
  1474.      callback_time 0.0
  1475.      time_toposort 1.90734863281e-06
  1476.   0.000144s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 1, 1) - 0.000s
  1477.     EquilibriumOptimizer      uncanonicalize
  1478.       time 0.000s for 1 passes
  1479.       nb nodes (start, end,  max) 1 1 1
  1480.       time io_toposort 0.000s
  1481.       time in local optimizers 0.000s
  1482.       time in global optimizers 0.000s
  1483.       time in final optimizers 0.000s
  1484.       time in cleanup optimizers 0.000s
  1485.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1486.     Global, final and clean up optimizers
  1487.     Iter 0
  1488.             TopoOptimizer  topo_constant_folding
  1489.               nb_node (start, end, changed) (1, 1, 0)
  1490.               init io_toposort 1.59740447998e-05
  1491.               loop time 1.90734863281e-06
  1492.               callback_time 0.0
  1493.   0.000102s - ('useless', 'TopoOptimizer', 3, 2, 2) - 0.000s
  1494.     TopoOptimizer  useless
  1495.       nb_node (start, end, changed) (2, 2, 0)
  1496.       init io_toposort 3.09944152832e-05
  1497.       loop time 4.00543212891e-05
  1498.       callback_time 0.0
  1499. LocalOptGroup
  1500. ---------------------
  1501.   time taken - times applied - times tried - name - node_created:
  1502.   -0.000s - 0 - 1 - local_subtensor_of_alloc - 0
  1503.   -0.000s - 0 - 1 - local_subtensor_make_vector - 0
  1504.   -0.000s - 0 - 1 - local_useless_slice - 0
  1505.   0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  1506.  
  1507.   0.000101s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 1, 1) - 0.000s
  1508.     TopoOptimizer  InplaceGpuBlasOpt
  1509.       nb_node (start, end, changed) (1, 1, 0)
  1510.       init io_toposort 1.4066696167e-05
  1511.       loop time 1.59740447998e-05
  1512.       callback_time 0.0
  1513. LocalOptGroup
  1514. ---------------------
  1515. --- The Optimizer wasn't successful ---
  1516.  
  1517.    0.000101s - ('blas_opt_inplace', 'TopoOptimizer', 34, 1, 1) - 0.000s
  1518.      TopoOptimizer  InplaceBlasOpt
  1519.        nb_node (start, end, changed) (1, 1, 0)
  1520.        init io_toposort 1.4066696167e-05
  1521.        loop time 1.31130218506e-05
  1522.        callback_time 0.0
  1523.  LocalOptGroup
  1524.  ---------------------
  1525.  --- The Optimizer wasn't successful ---
  1526.  
  1527.   0.000100s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 1, 1) - 0.000s
  1528.     TopoOptimizer  local_dnn_conv_inplace
  1529.       nb_node (start, end, changed) (1, 1, 0)
  1530.       init io_toposort 1.50203704834e-05
  1531.       loop time 1.00135803223e-05
  1532.       callback_time 0.0
  1533. LocalOptGroup
  1534. ---------------------
  1535. --- The Optimizer wasn't successful ---
  1536.  
  1537.    0.000100s - ('merge1', 'MergeOptimizer', 0, 2, 2) - 0.000s
  1538.      MergeOptimizer
  1539.        nb fail=    0 merged=    0 constant=    0
  1540.        time replace=0.00 validate=0.00 callback=0.00
  1541.    0.000098s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 1, 1) - 0.000s
  1542.      TopoOptimizer  local_dnna_conv_inplace
  1543.        nb_node (start, end, changed) (1, 1, 0)
  1544.        init io_toposort 1.50203704834e-05
  1545.        loop time 1.09672546387e-05
  1546.        callback_time 0.0
  1547.  LocalOptGroup
  1548.  ---------------------
  1549.    time taken - times applied - times tried - name - node_created:
  1550.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  1551.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  1552.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  1553.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  1554.  
  1555.    0.000096s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 1, 1) - 0.000s
  1556.      TopoOptimizer  InplaceGpuaBlasOpt
  1557.        nb_node (start, end, changed) (1, 1, 0)
  1558.        init io_toposort 1.50203704834e-05
  1559.        loop time 1.00135803223e-05
  1560.        callback_time 0.0
  1561.  LocalOptGroup
  1562.  ---------------------
  1563.    time taken - times applied - times tried - name - node_created:
  1564.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  1565.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  1566.  
  1567.    0.000095s - ('specialize_device', 'EquilibriumOptimizer', 17, 1, 1) - 0.000s
  1568.      EquilibriumOptimizer      specialize_device
  1569.        time 0.000s for 1 passes
  1570.        nb nodes (start, end,  max) 1 1 1
  1571.        time io_toposort 0.000s
  1572.        time in local optimizers 0.000s
  1573.        time in global optimizers 0.000s
  1574.        time in final optimizers 0.000s
  1575.        time in cleanup optimizers 0.000s
  1576.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  1577.    0.000062s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 1, 1) - 0.000s
  1578.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  1579.        nb_node (start, end, changed) (1, 1, 0)
  1580.        init io_toposort 1.47819519043e-05
  1581.        loop time 2.86102294922e-06
  1582.        callback_time 0.0
  1583.    0.000057s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 1, 1) - 0.000s
  1584.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  1585.        nb_node (start, end, changed) (1, 1, 0)
  1586.        init io_toposort 1.4066696167e-05
  1587.        loop time 3.09944152832e-06
  1588.        callback_time 0.0
  1589.    0.000056s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 1, 1) - 0.000s
  1590.      TopoOptimizer  local_gemm16_inplace
  1591.        nb_node (start, end, changed) (1, 1, 0)
  1592.        init io_toposort 1.50203704834e-05
  1593.        loop time 1.90734863281e-06
  1594.        callback_time 0.0
  1595.    0.000055s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 1, 1) - 0.000s
  1596.      TopoOptimizer  local_inplace_sparse_block_outer
  1597.        nb_node (start, end, changed) (1, 1, 0)
  1598.        init io_toposort 1.28746032715e-05
  1599.        loop time 2.86102294922e-06
  1600.        callback_time 0.0
  1601.    0.000055s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 1, 1) - 0.000s
  1602.      TopoOptimizer  local_inplace_sparse_block_gemv
  1603.        nb_node (start, end, changed) (1, 1, 0)
  1604.        init io_toposort 1.4066696167e-05
  1605.        loop time 1.90734863281e-06
  1606.        callback_time 0.0
  1607.    0.000055s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 2, 2) - 0.000s
  1608.      TopoOptimizer  pre_local_IncSubtensor_serialize
  1609.        nb_node (start, end, changed) (2, 2, 0)
  1610.        init io_toposort 2.69412994385e-05
  1611.        loop time 6.19888305664e-06
  1612.        callback_time 0.0
  1613.    0.000055s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 1, 1) - 0.000s
  1614.      TopoOptimizer  local_inplace_incsubtensor1
  1615.        nb_node (start, end, changed) (1, 1, 0)
  1616.        init io_toposort 1.4066696167e-05
  1617.        loop time 2.14576721191e-06
  1618.        callback_time 0.0
  1619.    0.000054s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 1, 1) - 0.000s
  1620.      TopoOptimizer  local_inplace_sparseblockgemv
  1621.        nb_node (start, end, changed) (1, 1, 0)
  1622.        init io_toposort 1.38282775879e-05
  1623.        loop time 2.14576721191e-06
  1624.        callback_time 0.0
  1625.    0.000054s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 1, 1) - 0.000s
  1626.      TopoOptimizer  local_inplace_setsubtensor
  1627.        nb_node (start, end, changed) (1, 1, 0)
  1628.        init io_toposort 1.4066696167e-05
  1629.        loop time 1.90734863281e-06
  1630.        callback_time 0.0
  1631.    0.000053s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 1, 1) - 0.000s
  1632.      TopoOptimizer  local_inplace_sparseblockouter
  1633.        nb_node (start, end, changed) (1, 1, 0)
  1634.        init io_toposort 1.31130218506e-05
  1635.        loop time 1.90734863281e-06
  1636.        callback_time 0.0
  1637.    0.000050s - ('cond_make_inplace', 'TopoOptimizer', 47, 1, 1) - 0.000s
  1638.      TopoOptimizer  cond_make_inplace
  1639.        nb_node (start, end, changed) (1, 1, 0)
  1640.        init io_toposort 2.28881835938e-05
  1641.        loop time 2.86102294922e-06
  1642.        callback_time 0.0
  1643.    0.000050s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 1, 1) - 0.000s
  1644.      TopoOptimizer  dimshuffle_as_view
  1645.        nb_node (start, end, changed) (1, 1, 0)
  1646.        init io_toposort 1.90734863281e-05
  1647.        loop time 4.05311584473e-06
  1648.        callback_time 0.0
  1649.    0.000049s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 1, 1) - 0.000s
  1650.    0.000047s - ('c_blas_destructive', 'TopoOptimizer', 37, 1, 1) - 0.000s
  1651.      TopoOptimizer  c_blas_destructive
  1652.        nb_node (start, end, changed) (1, 1, 0)
  1653.        init io_toposort 1.4066696167e-05
  1654.        loop time 1.00135803223e-05
  1655.        callback_time 0.0
  1656.  LocalOptGroup
  1657.  ---------------------
  1658.  --- The Optimizer wasn't successful ---
  1659.  
  1660.   0.000046s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 1, 1) - 0.000s
  1661.   0.000043s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 1, 1) - 0.000s
  1662.     TopoOptimizer  local_advincsub1_gpua_inplace
  1663.       nb_node (start, end, changed) (1, 1, 0)
  1664.       init io_toposort 1.59740447998e-05
  1665.       loop time 3.09944152832e-06
  1666.       callback_time 0.0
  1667.   0.000041s - ('AbstractConvCheck', 'TopoOptimizer', 18, 1, 1) - 0.000s
  1668.     TopoOptimizer  AbstractConvCheck
  1669.       nb_node (start, end, changed) (1, 1, 0)
  1670.       init io_toposort 1.4066696167e-05
  1671.       loop time 4.05311584473e-06
  1672.       callback_time 0.0
  1673.   0.000040s - ('local_destructive', 'TopoOptimizer', 48, 1, 1) - 0.000s
  1674.     TopoOptimizer  CURAND_destructive
  1675.       nb_node (start, end, changed) (1, 1, 0)
  1676.       init io_toposort 1.4066696167e-05
  1677.       loop time 2.86102294922e-06
  1678.       callback_time 0.0
  1679.   0.000039s - ('make_ger_destructive', 'TopoOptimizer', 41, 1, 1) - 0.000s
  1680.     TopoOptimizer  make_scipy_blas_destructive
  1681.       nb_node (start, end, changed) (1, 1, 0)
  1682.       init io_toposort 1.31130218506e-05
  1683.       loop time 2.86102294922e-06
  1684.       callback_time 0.0
  1685.   0.000039s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 1, 1) - 0.000s
  1686.     TopoOptimizer  local_fill_to_alloc
  1687.       nb_node (start, end, changed) (1, 1, 0)
  1688.       init io_toposort 1.4066696167e-05
  1689.       loop time 3.81469726562e-06
  1690.       callback_time 0.0
  1691.   0.000039s - ('merge1.2', 'MergeOptimizer', 7, 1, 1) - 0.000s
  1692.     MergeOptimizer
  1693.       nb fail=    0 merged=    0 constant=    0
  1694.       time replace=0.00 validate=0.00 callback=0.00
  1695.   0.000038s - ('random_make_inplace', 'TopoOptimizer', 49, 1, 1) - 0.000s
  1696.     TopoOptimizer  random_make_inplace
  1697.       nb_node (start, end, changed) (1, 1, 0)
  1698.       init io_toposort 1.28746032715e-05
  1699.       loop time 3.09944152832e-06
  1700.       callback_time 0.0
  1701.   0.000038s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 1, 1) - 0.000s
  1702.   0.000038s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 1, 1) - 0.000s
  1703.     TopoOptimizer  random_make_inplace_mrg
  1704.       nb_node (start, end, changed) (1, 1, 0)
  1705.       init io_toposort 1.31130218506e-05
  1706.       loop time 1.90734863281e-06
  1707.       callback_time 0.0
  1708.   0.000036s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 1, 1) - 0.000s
  1709.     TopoOptimizer  local_elemwise_alloc
  1710.       nb_node (start, end, changed) (1, 1, 0)
  1711.       init io_toposort 1.38282775879e-05
  1712.       loop time 2.14576721191e-06
  1713.       callback_time 0.0
  1714.   0.000034s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 1, 1) - 0.000s
  1715.   0.000029s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 1, 1) - 0.000s
  1716.   0.000020s - ('merge1.1', 'MergeOptimizer', 4, 2, 2) - 0.000s
  1717.     MergeOptimizer
  1718.       nb fail=    0 merged=    0 constant=    0
  1719.       time replace=0.00 validate=0.00 callback=0.00
  1720.   0.000006s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 1, 1) - 0.000s
  1721.  
  1722. Here are tips to potentially make your code run faster
  1723.                 (if you think of new ones, suggest them on the mailing list).
  1724.                 Test them first, as they are not guaranteed to always provide a speedup.
  1725.  Sorry, no tip for today.
  1726. Function profiling
  1727. ==================
  1728.  Message: sb/convnet/sb_resnet.py:260
  1729.  Time in 1 calls to Function.__call__: 8.409023e-04s
  1730.  Time in Function.fn.__call__: 8.080006e-04s (96.087%)
  1731.  Time in thunks: 7.898808e-04s (93.933%)
  1732.  Total compile time: 3.525209e-02s
  1733.    Number of Apply nodes: 2
  1734.    Theano Optimizer time: 1.099181e-02s
  1735.       Theano validate time: 1.215935e-05s
  1736.    Theano Linker time (includes C, CUDA code generation/compiling): 6.864071e-03s
  1737.       Import time 5.328894e-03s
  1738.       Node make_thunk time 6.590128e-03s
  1739.  
  1740. Time in all call to theano.grad() 2.656322e+00s
  1741. Time since theano import 477.762s
  1742. Class
  1743. ---
  1744. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  1745.  98.0%    98.0%       0.001s       7.74e-04s     C        1       1   theano.gpuarray.basic_ops.HostFromGpu
  1746.   2.0%   100.0%       0.000s       1.60e-05s     C        1       1   theano.gpuarray.subtensor.GpuSubtensor
  1747.   ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  1748.  
  1749. Ops
  1750. ---
  1751. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  1752.  98.0%    98.0%       0.001s       7.74e-04s     C        1        1   HostFromGpu(gpuarray)
  1753.   2.0%   100.0%       0.000s       1.60e-05s     C        1        1   GpuSubtensor{:int64:}
  1754.   ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  1755.  
  1756. Apply
  1757. ------
  1758. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  1759.  98.0%    98.0%       0.001s       7.74e-04s      1     1   HostFromGpu(gpuarray)(GpuSubtensor{:int64:}.0)
  1760.   2.0%   100.0%       0.000s       1.60e-05s      1     0   GpuSubtensor{:int64:}(<GpuArrayType<None>(float32, (False, False, False, False))>, Constant{128})
  1761.   ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  1762.  
  1763. Optimizer Profile
  1764. -----------------
  1765. SeqOptimizer  OPT_FAST_RUN  time 0.011s for 2/2 nodes before/after optimization
  1766.   0.001s for callback
  1767.       0.000s for fgraph.validate()
  1768.   time      - (name, class, index, nodes before, nodes after) - validate time
  1769.   0.001547s - ('gpuarray_opt', 'SeqOptimizer', 16, 2, 2) - 0.000s
  1770.     SeqOptimizer      gpuarray_opt  time 0.001s for 2/2 nodes before/after optimization
  1771.       0.001s for callback
  1772.           0.000s for fgraph.validate()
  1773.       0.000959s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 2, 2) - 0.000s
  1774.         GraphToGPUOptimizer          gpuarray_graph_optimization
  1775.           time io_toposort 0.000s
  1776.         Total time taken by local optimizers 0.000s
  1777.           times - times applied - Node created - name:
  1778.           0.000s - 1 - 1 - local_gpua_subtensor_graph
  1779.           0.000s - in 0 optimization that were not used (display only those with a runtime > 0)
  1780.  
  1781.       0.000353s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 2, 2) - 0.000s
  1782.         EquilibriumOptimizer          gpuarray_local_optimizations
  1783.           time 0.000s for 1 passes
  1784.           nb nodes (start, end,  max) 2 2 2
  1785.           time io_toposort 0.000s
  1786.           time in local optimizers 0.000s
  1787.           time in global optimizers 0.000s
  1788.           time in final optimizers 0.000s
  1789.           time in cleanup optimizers 0.000s
  1790.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1791.       0.000113s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 2, 2) - 0.000s
  1792.         EquilibriumOptimizer          gpuarray_cut_transfers
  1793.           time 0.000s for 1 passes
  1794.           nb nodes (start, end,  max) 2 2 2
  1795.           time io_toposort 0.000s
  1796.           time in local optimizers 0.000s
  1797.           time in global optimizers 0.000s
  1798.           time in final optimizers 0.000s
  1799.           time in cleanup optimizers 0.000s
  1800.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1801.       0.000010s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 2, 2) - 0.000s
  1802.  
  1803.   0.000818s - ('ShapeOpt', 'ShapeOptimizer', 2, 2, 2) - 0.000s
  1804.   0.000686s - ('canonicalize', 'EquilibriumOptimizer', 6, 2, 2) - 0.000s
  1805.     EquilibriumOptimizer      canonicalize
  1806.       time 0.000s for 1 passes
  1807.       nb nodes (start, end,  max) 2 2 2
  1808.       time io_toposort 0.000s
  1809.       time in local optimizers 0.000s
  1810.       time in global optimizers 0.000s
  1811.       time in final optimizers 0.000s
  1812.       time in cleanup optimizers 0.000s
  1813.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1814.     Global, final and clean up optimizers
  1815.     Iter 0
  1816.             TopoOptimizer  topo_constant_folding
  1817.               nb_node (start, end, changed) (2, 2, 0)
  1818.               init io_toposort 2.90870666504e-05
  1819.               loop time 3.81469726562e-06
  1820.               callback_time 0.0
  1821.             MergeOptimizer
  1822.               nb fail=    0 merged=    0 constant=    0
  1823.               time replace=0.00 validate=0.00 callback=0.00
  1824.   0.000629s - ('specialize', 'EquilibriumOptimizer', 13, 2, 2) - 0.000s
  1825.     EquilibriumOptimizer      specialize
  1826.       time 0.000s for 1 passes
  1827.       nb nodes (start, end,  max) 2 2 2
  1828.       time io_toposort 0.000s
  1829.       time in local optimizers 0.000s
  1830.       time in global optimizers 0.000s
  1831.       time in final optimizers 0.000s
  1832.       time in cleanup optimizers 0.000s
  1833.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1834.     Global, final and clean up optimizers
  1835.     Iter 0
  1836.             TopoOptimizer  topo_constant_folding
  1837.               nb_node (start, end, changed) (2, 2, 0)
  1838.               init io_toposort 2.78949737549e-05
  1839.               loop time 2.86102294922e-06
  1840.               callback_time 0.0
  1841.   0.000599s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 2, 2) - 0.000s
  1842.     EquilibriumOptimizer      scan_eqopt2
  1843.       time 0.001s for 1 passes
  1844.       nb nodes (start, end,  max) 2 2 2
  1845.       time io_toposort 0.000s
  1846.       time in local optimizers 0.000s
  1847.       time in global optimizers 0.000s
  1848.       time in final optimizers 0.000s
  1849.       time in cleanup optimizers 0.000s
  1850.        0 - 0.001s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1851.     Global, final and clean up optimizers
  1852.     Iter 0
  1853.             TopoOptimizer  constant_folding_for_scan2
  1854.               nb_node (start, end, changed) (2, 2, 0)
  1855.               init io_toposort 2.69412994385e-05
  1856.               loop time 3.09944152832e-06
  1857.               callback_time 0.0
  1858.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  1859.               nb_node (start, end, changed) (2, 2, 0)
  1860.               init io_toposort 2.50339508057e-05
  1861.               loop time 2.86102294922e-06
  1862.               callback_time 0.0
  1863.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  1864.               nb_node (start, end, changed) (2, 2, 0)
  1865.               init io_toposort 0.000126123428345
  1866.               loop time 3.81469726562e-06
  1867.               callback_time 0.0
  1868.             TopoOptimizer  scanOp_merge_inouts
  1869.               nb_node (start, end, changed) (2, 2, 0)
  1870.               init io_toposort 2.69412994385e-05
  1871.               loop time 3.81469726562e-06
  1872.               callback_time 0.0
  1873.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  1874.               nb_node (start, end, changed) (2, 2, 0)
  1875.               init io_toposort 2.40802764893e-05
  1876.               loop time 3.09944152832e-06
  1877.               callback_time 0.0
  1878.   0.000584s - ('BlasOpt', 'SeqOptimizer', 12, 2, 2) - 0.000s
  1879.     SeqOptimizer      BlasOpt  time 0.000s for 2/2 nodes before/after optimization
  1880.       0.000s for callback
  1881.           0.000s for fgraph.validate()
  1882.       0.000149s - ('gemm_optimizer', 'GemmOptimizer', 1, 2, 2) - 0.000s
  1883.         GemmOptimizer
  1884.          nb_iter 1
  1885.          nb_replacement 0
  1886.          nb_replacement_didn_t_remove 0
  1887.          nb_inconsistency_make 0
  1888.          nb_inconsistency_replace 0
  1889.          time_canonicalize 0
  1890.          time_factor_can 0
  1891.          time_factor_list 0
  1892.          time_toposort 2.50339508057e-05
  1893.          validate_time 0.0
  1894.          callback_time 0.0
  1895.       0.000120s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 2, 2) - 0.000s
  1896.         EquilibriumOptimizer          local_gemm_to_gemv
  1897.           time 0.000s for 1 passes
  1898.           nb nodes (start, end,  max) 2 2 2
  1899.           time io_toposort 0.000s
  1900.           time in local optimizers 0.000s
  1901.           time in global optimizers 0.000s
  1902.           time in final optimizers 0.000s
  1903.           time in cleanup optimizers 0.000s
  1904.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1905.       0.000067s - ('use_c_blas', 'TopoOptimizer', 4, 2, 2) - 0.000s
  1906.         TopoOptimizer  use_c_blas
  1907.           nb_node (start, end, changed) (2, 2, 0)
  1908.           init io_toposort 2.69412994385e-05
  1909.           loop time 1.81198120117e-05
  1910.           callback_time 0.0
  1911. LocalOptGroup
  1912. ---------------------
  1913. --- The Optimizer wasn't successful ---
  1914.  
  1915.        0.000053s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 2, 2) - 0.000s
  1916.          TopoOptimizer  local_dot22_to_dot22scalar
  1917.            nb_node (start, end, changed) (2, 2, 0)
  1918.            init io_toposort 2.50339508057e-05
  1919.            loop time 6.19888305664e-06
  1920.            callback_time 0.0
  1921.        0.000053s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 2, 2) - 0.000s
  1922.          TopoOptimizer  local_dot_to_dot22
  1923.            nb_node (start, end, changed) (2, 2, 0)
  1924.            init io_toposort 2.59876251221e-05
  1925.            loop time 4.05311584473e-06
  1926.            callback_time 0.0
  1927.        0.000051s - ('use_scipy_ger', 'TopoOptimizer', 5, 2, 2) - 0.000s
  1928.          TopoOptimizer  scipy_blas
  1929.            nb_node (start, end, changed) (2, 2, 0)
  1930.            init io_toposort 2.59876251221e-05
  1931.            loop time 4.05311584473e-06
  1932.            callback_time 0.0
  1933.  
  1934.    0.000577s - ('add_destroy_handler', 'AddDestroyHandler', 23, 2, 2) - 0.000s
  1935.    0.000558s - ('elemwise_fusion', 'SeqOptimizer', 19, 2, 2) - 0.000s
  1936.      SeqOptimizer      elemwise_fusion  time 0.000s for 2/2 nodes before/after optimization
  1937.        0.000s for callback
  1938.            0.000s for fgraph.validate()
  1939.        0.000228s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 2, 2) - 0.000s
  1940.          FusionOptimizer
  1941.           nb_iter 1
  1942.           nb_replacement 0
  1943.           nb_inconsistency_replace 0
  1944.           validate_time 0.0
  1945.           callback_time 0.0
  1946.           time_toposort 5.41210174561e-05
  1947.        0.000216s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 2, 2) - 0.000s
  1948.          FusionOptimizer
  1949.           nb_iter 1
  1950.           nb_replacement 0
  1951.           nb_inconsistency_replace 0
  1952.           validate_time 0.0
  1953.           callback_time 0.0
  1954.           time_toposort 4.60147857666e-05
  1955.  
  1956.    0.000421s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 2, 2) - 0.000s
  1957.      EquilibriumOptimizer      scan_eqopt1
  1958.        time 0.000s for 1 passes
  1959.        nb nodes (start, end,  max) 2 2 2
  1960.        time io_toposort 0.000s
  1961.        time in local optimizers 0.000s
  1962.        time in global optimizers 0.000s
  1963.        time in final optimizers 0.000s
  1964.        time in cleanup optimizers 0.000s
  1965.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1966.      Global, final and clean up optimizers
  1967.      Iter 0
  1968.              SeqOptimizer              all_pushout_opt  time 0.000s for 2/2 nodes before/after optimization
  1969.                0.000s for callback
  1970.                    0.000s for fgraph.validate()
  1971.                0.000068s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 2, 2) - 0.000s
  1972.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  1973.                    nb_node (start, end, changed) (2, 2, 0)
  1974.                    init io_toposort 3.00407409668e-05
  1975.                    loop time 5.96046447754e-06
  1976.                    callback_time 0.0
  1977.                0.000064s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 2, 2) - 0.000s
  1978.                0.000055s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 2, 2) - 0.000s
  1979.                0.000052s - ('scan_pushout_dot1', 'PushOutDot1', 3, 2, 2) - 0.000s
  1980.                0.000051s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 2, 2) - 0.000s
  1981.  
  1982.    0.000368s - ('stabilize', 'EquilibriumOptimizer', 8, 2, 2) - 0.000s
  1983.      EquilibriumOptimizer      stabilize
  1984.        time 0.000s for 1 passes
  1985.        nb nodes (start, end,  max) 2 2 2
  1986.        time io_toposort 0.000s
  1987.        time in local optimizers 0.000s
  1988.        time in global optimizers 0.000s
  1989.        time in final optimizers 0.000s
  1990.        time in cleanup optimizers 0.000s
  1991.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  1992.      Global, final and clean up optimizers
  1993.      Iter 0
  1994.              TopoOptimizer  topo_constant_folding
  1995.                nb_node (start, end, changed) (2, 2, 0)
  1996.                init io_toposort 2.69412994385e-05
  1997.                loop time 2.86102294922e-06
  1998.                callback_time 0.0
  1999.    0.000252s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 2, 2) - 0.000s
  2000.      TopoOptimizer  local_gemm16_inplace
  2001.        nb_node (start, end, changed) (2, 2, 0)
  2002.        init io_toposort 0.000180006027222
  2003.        loop time 4.05311584473e-06
  2004.        callback_time 0.0
  2005.    0.000216s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 2, 2) - 0.000s
  2006.      FusionOptimizer
  2007.       nb_iter 1
  2008.       nb_replacement 0
  2009.       nb_inconsistency_replace 0
  2010.       validate_time 0.0
  2011.       callback_time 0.0
  2012.       time_toposort 4.6968460083e-05
  2013.    0.000212s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 2, 2) - 0.000s
  2014.      FusionOptimizer
  2015.       nb_iter 1
  2016.       nb_replacement 0
  2017.       nb_inconsistency_replace 0
  2018.       validate_time 0.0
  2019.       callback_time 0.0
  2020.       time_toposort 4.38690185547e-05
  2021.    0.000200s - ('merge3', 'MergeOptimizer', 51, 2, 2) - 0.000s
  2022.      MergeOptimizer
  2023.        nb fail=    0 merged=    0 constant=    0
  2024.        time replace=0.00 validate=0.00 callback=0.00
  2025.    0.000199s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 2, 2) - 0.000s
  2026.      TopoOptimizer  local_dnna_conv_inplace
  2027.        nb_node (start, end, changed) (2, 2, 0)
  2028.        init io_toposort 5.00679016113e-05
  2029.        loop time 5.48362731934e-05
  2030.        callback_time 0.0
  2031.  LocalOptGroup
  2032.  ---------------------
  2033.    time taken - times applied - times tried - name - node_created:
  2034.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  2035.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  2036.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  2037.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  2038.  
  2039.    0.000177s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 2, 2) - 0.000s
  2040.      EquilibriumOptimizer      uncanonicalize
  2041.        time 0.000s for 1 passes
  2042.        nb nodes (start, end,  max) 2 2 2
  2043.        time io_toposort 0.000s
  2044.        time in local optimizers 0.000s
  2045.        time in global optimizers 0.000s
  2046.        time in final optimizers 0.000s
  2047.        time in cleanup optimizers 0.000s
  2048.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2049.      Global, final and clean up optimizers
  2050.      Iter 0
  2051.              TopoOptimizer  topo_constant_folding
  2052.                nb_node (start, end, changed) (2, 2, 0)
  2053.                init io_toposort 2.50339508057e-05
  2054.                loop time 3.09944152832e-06
  2055.                callback_time 0.0
  2056.    0.000119s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 2, 2) - 0.000s
  2057.      TopoOptimizer  InplaceGpuaBlasOpt
  2058.        nb_node (start, end, changed) (2, 2, 0)
  2059.        init io_toposort 2.69412994385e-05
  2060.        loop time 1.59740447998e-05
  2061.        callback_time 0.0
  2062.  LocalOptGroup
  2063.  ---------------------
  2064.    time taken - times applied - times tried - name - node_created:
  2065.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  2066.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  2067.  
  2068.    0.000116s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 2, 2) - 0.000s
  2069.      TopoOptimizer  local_dnn_conv_inplace
  2070.        nb_node (start, end, changed) (2, 2, 0)
  2071.        init io_toposort 2.59876251221e-05
  2072.        loop time 1.59740447998e-05
  2073.        callback_time 0.0
  2074.  LocalOptGroup
  2075.  ---------------------
  2076.  --- The Optimizer wasn't successful ---
  2077.  
  2078.   0.000116s - ('blas_opt_inplace', 'TopoOptimizer', 34, 2, 2) - 0.000s
  2079.     TopoOptimizer  InplaceBlasOpt
  2080.       nb_node (start, end, changed) (2, 2, 0)
  2081.       init io_toposort 2.50339508057e-05
  2082.       loop time 2.09808349609e-05
  2083.       callback_time 0.0
  2084. LocalOptGroup
  2085. ---------------------
  2086. --- The Optimizer wasn't successful ---
  2087.  
  2088.    0.000115s - ('useless', 'TopoOptimizer', 3, 2, 2) - 0.000s
  2089.      TopoOptimizer  useless
  2090.        nb_node (start, end, changed) (2, 2, 0)
  2091.        init io_toposort 3.31401824951e-05
  2092.        loop time 4.6968460083e-05
  2093.        callback_time 0.0
  2094.  LocalOptGroup
  2095.  ---------------------
  2096.    time taken - times applied - times tried - name - node_created:
  2097.    -0.000s - 0 - 1 - local_subtensor_of_alloc - 0
  2098.    -0.000s - 0 - 1 - local_subtensor_make_vector - 0
  2099.    -0.000s - 0 - 1 - local_useless_slice - 0
  2100.    0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  2101.  
  2102.    0.000113s - ('specialize_device', 'EquilibriumOptimizer', 17, 2, 2) - 0.000s
  2103.      EquilibriumOptimizer      specialize_device
  2104.        time 0.000s for 1 passes
  2105.        nb nodes (start, end,  max) 2 2 2
  2106.        time io_toposort 0.000s
  2107.        time in local optimizers 0.000s
  2108.        time in global optimizers 0.000s
  2109.        time in final optimizers 0.000s
  2110.        time in cleanup optimizers 0.000s
  2111.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2112.    0.000113s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 2, 2) - 0.000s
  2113.      TopoOptimizer  InplaceGpuBlasOpt
  2114.        nb_node (start, end, changed) (2, 2, 0)
  2115.        init io_toposort 2.59876251221e-05
  2116.        loop time 1.59740447998e-05
  2117.        callback_time 0.0
  2118.  LocalOptGroup
  2119.  ---------------------
  2120.  --- The Optimizer wasn't successful ---
  2121.  
  2122.   0.000111s - ('merge2', 'MergeOptimizer', 22, 2, 2) - 0.000s
  2123.     MergeOptimizer
  2124.       nb fail=    0 merged=    0 constant=    0
  2125.       time replace=0.00 validate=0.00 callback=0.00
  2126.   0.000109s - ('merge1', 'MergeOptimizer', 0, 2, 2) - 0.000s
  2127.     MergeOptimizer
  2128.       nb fail=    0 merged=    0 constant=    0
  2129.       time replace=0.00 validate=0.00 callback=0.00
  2130.   0.000090s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 2, 2) - 0.000s
  2131.   0.000090s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 2, 2) - 0.000s
  2132.     TopoOptimizer  local_inplace_incsubtensor1
  2133.       nb_node (start, end, changed) (2, 2, 0)
  2134.       init io_toposort 3.60012054443e-05
  2135.       loop time 3.09944152832e-06
  2136.       callback_time 0.0
  2137.   0.000086s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 2, 2) - 0.000s
  2138.     TopoOptimizer  pre_local_IncSubtensor_serialize
  2139.       nb_node (start, end, changed) (2, 2, 0)
  2140.       init io_toposort 3.60012054443e-05
  2141.       loop time 5.96046447754e-06
  2142.       callback_time 0.0
  2143.   0.000081s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 2, 2) - 0.000s
  2144.     TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  2145.       nb_node (start, end, changed) (2, 2, 0)
  2146.       init io_toposort 3.09944152832e-05
  2147.       loop time 4.05311584473e-06
  2148.       callback_time 0.0
  2149.   0.000079s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 2, 2) - 0.000s
  2150.   0.000071s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 2, 2) - 0.000s
  2151.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  2152.       nb_node (start, end, changed) (2, 2, 0)
  2153.       init io_toposort 2.59876251221e-05
  2154.       loop time 3.09944152832e-06
  2155.       callback_time 0.0
  2156.   0.000070s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 2, 2) - 0.000s
  2157.   0.000069s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 2, 2) - 0.000s
  2158.     TopoOptimizer  local_inplace_sparse_block_outer
  2159.       nb_node (start, end, changed) (2, 2, 0)
  2160.       init io_toposort 2.59876251221e-05
  2161.       loop time 4.05311584473e-06
  2162.       callback_time 0.0
  2163.   0.000069s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 2, 2) - 0.000s
  2164.     TopoOptimizer  local_inplace_setsubtensor
  2165.       nb_node (start, end, changed) (2, 2, 0)
  2166.       init io_toposort 2.50339508057e-05
  2167.       loop time 2.86102294922e-06
  2168.       callback_time 0.0
  2169.   0.000068s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 2, 2) - 0.000s
  2170.     TopoOptimizer  local_inplace_sparse_block_gemv
  2171.       nb_node (start, end, changed) (2, 2, 0)
  2172.       init io_toposort 2.50339508057e-05
  2173.       loop time 3.09944152832e-06
  2174.       callback_time 0.0
  2175.   0.000067s - ('c_blas_destructive', 'TopoOptimizer', 37, 2, 2) - 0.000s
  2176.     TopoOptimizer  c_blas_destructive
  2177.       nb_node (start, end, changed) (2, 2, 0)
  2178.       init io_toposort 2.50339508057e-05
  2179.       loop time 1.59740447998e-05
  2180.       callback_time 0.0
  2181. LocalOptGroup
  2182. ---------------------
  2183. --- The Optimizer wasn't successful ---
  2184.  
  2185.    0.000067s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 2, 2) - 0.000s
  2186.      TopoOptimizer  local_inplace_sparseblockouter
  2187.        nb_node (start, end, changed) (2, 2, 0)
  2188.        init io_toposort 2.50339508057e-05
  2189.        loop time 3.09944152832e-06
  2190.        callback_time 0.0
  2191.    0.000067s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 2, 2) - 0.000s
  2192.      TopoOptimizer  local_inplace_sparseblockgemv
  2193.        nb_node (start, end, changed) (2, 2, 0)
  2194.        init io_toposort 2.50339508057e-05
  2195.        loop time 2.86102294922e-06
  2196.        callback_time 0.0
  2197.    0.000066s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 2, 2) - 0.000s
  2198.      TopoOptimizer  dimshuffle_as_view
  2199.        nb_node (start, end, changed) (2, 2, 0)
  2200.        init io_toposort 3.2901763916e-05
  2201.        loop time 4.05311584473e-06
  2202.        callback_time 0.0
  2203.    0.000063s - ('make_ger_destructive', 'TopoOptimizer', 41, 2, 2) - 0.000s
  2204.      TopoOptimizer  make_scipy_blas_destructive
  2205.        nb_node (start, end, changed) (2, 2, 0)
  2206.        init io_toposort 3.09944152832e-05
  2207.        loop time 5.00679016113e-06
  2208.        callback_time 0.0
  2209.    0.000060s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 2, 2) - 0.000s
  2210.      TopoOptimizer  local_advincsub1_gpua_inplace
  2211.        nb_node (start, end, changed) (2, 2, 0)
  2212.        init io_toposort 2.8133392334e-05
  2213.        loop time 4.05311584473e-06
  2214.        callback_time 0.0
  2215.    0.000059s - ('cond_make_inplace', 'TopoOptimizer', 47, 2, 2) - 0.000s
  2216.      TopoOptimizer  cond_make_inplace
  2217.        nb_node (start, end, changed) (2, 2, 0)
  2218.        init io_toposort 2.59876251221e-05
  2219.        loop time 5.00679016113e-06
  2220.        callback_time 0.0
  2221.    0.000057s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 2, 2) - 0.000s
  2222.    0.000056s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 2, 2) - 0.000s
  2223.      TopoOptimizer  random_make_inplace_mrg
  2224.        nb_node (start, end, changed) (2, 2, 0)
  2225.        init io_toposort 2.59876251221e-05
  2226.        loop time 5.00679016113e-06
  2227.        callback_time 0.0
  2228.    0.000055s - ('local_destructive', 'TopoOptimizer', 48, 2, 2) - 0.000s
  2229.      TopoOptimizer  CURAND_destructive
  2230.        nb_node (start, end, changed) (2, 2, 0)
  2231.        init io_toposort 2.59876251221e-05
  2232.        loop time 5.00679016113e-06
  2233.        callback_time 0.0
  2234.    0.000055s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 2, 2) - 0.000s
  2235.    0.000054s - ('random_make_inplace', 'TopoOptimizer', 49, 2, 2) - 0.000s
  2236.      TopoOptimizer  random_make_inplace
  2237.        nb_node (start, end, changed) (2, 2, 0)
  2238.        init io_toposort 2.59876251221e-05
  2239.        loop time 4.05311584473e-06
  2240.        callback_time 0.0
  2241.    0.000054s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 2, 2) - 0.000s
  2242.    0.000054s - ('AbstractConvCheck', 'TopoOptimizer', 18, 2, 2) - 0.000s
  2243.      TopoOptimizer  AbstractConvCheck
  2244.        nb_node (start, end, changed) (2, 2, 0)
  2245.        init io_toposort 2.59876251221e-05
  2246.        loop time 4.76837158203e-06
  2247.        callback_time 0.0
  2248.    0.000054s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 2, 2) - 0.000s
  2249.      TopoOptimizer  local_fill_to_alloc
  2250.        nb_node (start, end, changed) (2, 2, 0)
  2251.        init io_toposort 2.59876251221e-05
  2252.        loop time 4.05311584473e-06
  2253.        callback_time 0.0
  2254.    0.000053s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 2, 2) - 0.000s
  2255.      TopoOptimizer  local_elemwise_alloc
  2256.        nb_node (start, end, changed) (2, 2, 0)
  2257.        init io_toposort 2.69412994385e-05
  2258.        loop time 3.09944152832e-06
  2259.        callback_time 0.0
  2260.    0.000030s - ('merge1.2', 'MergeOptimizer', 7, 2, 2) - 0.000s
  2261.      MergeOptimizer
  2262.        nb fail=    0 merged=    0 constant=    0
  2263.        time replace=0.00 validate=0.00 callback=0.00
  2264.    0.000021s - ('merge1.1', 'MergeOptimizer', 4, 2, 2) - 0.000s
  2265.      MergeOptimizer
  2266.        nb fail=    0 merged=    0 constant=    0
  2267.        time replace=0.00 validate=0.00 callback=0.00
  2268.  
  2269. Here are tips to potentially make your code run faster
  2270.                  (if you think of new ones, suggest them on the mailing list).
  2271.                  Test them first, as they are not guaranteed to always provide a speedup.
  2272.   Sorry, no tip for today.
  2273. Function profiling
  2274. ==================
  2275.   Message: sb/convnet/sb_resnet.py:261
  2276.   Time in 1 calls to Function.__call__: 8.106232e-05s
  2277.   Time in Function.fn.__call__: 6.794930e-05s (83.824%)
  2278.   Time in thunks: 6.198883e-05s (76.471%)
  2279.   Total compile time: 3.346705e-02s
  2280.     Number of Apply nodes: 2
  2281.     Theano Optimizer time: 1.266003e-02s
  2282.        Theano validate time: 1.287460e-05s
  2283.     Theano Linker time (includes C, CUDA code generation/compiling): 4.466057e-03s
  2284.        Import time 3.094912e-03s
  2285.        Node make_thunk time 4.228830e-03s
  2286.  
  2287. Time in all call to theano.grad() 2.656322e+00s
  2288. Time since theano import 477.768s
  2289. Class
  2290. ---
  2291. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  2292.   93.8%    93.8%       0.000s       5.82e-05s     C        1       1   theano.gpuarray.basic_ops.HostFromGpu
  2293.    6.2%   100.0%       0.000s       3.81e-06s     C        1       1   theano.gpuarray.subtensor.GpuSubtensor
  2294.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  2295.  
  2296. Ops
  2297. ---
  2298. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  2299.   93.8%    93.8%       0.000s       5.82e-05s     C        1        1   HostFromGpu(gpuarray)
  2300.    6.2%   100.0%       0.000s       3.81e-06s     C        1        1   GpuSubtensor{:int64:}
  2301.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  2302.  
  2303. Apply
  2304. ------
  2305. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  2306.   93.8%    93.8%       0.000s       5.82e-05s      1     1   HostFromGpu(gpuarray)(GpuSubtensor{:int64:}.0)
  2307.    6.2%   100.0%       0.000s       3.81e-06s      1     0   GpuSubtensor{:int64:}(<GpuArrayType<None>(float32, (False,))>, Constant{128})
  2308.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  2309.  
  2310. Optimizer Profile
  2311. -----------------
  2312.  SeqOptimizer  OPT_FAST_RUN  time 0.012s for 2/2 nodes before/after optimization
  2313.    0.001s for callback
  2314.        0.000s for fgraph.validate()
  2315.    time      - (name, class, index, nodes before, nodes after) - validate time
  2316.    0.002204s - ('canonicalize', 'EquilibriumOptimizer', 6, 2, 2) - 0.000s
  2317.      EquilibriumOptimizer      canonicalize
  2318.        time 0.002s for 1 passes
  2319.        nb nodes (start, end,  max) 2 2 2
  2320.        time io_toposort 0.000s
  2321.        time in local optimizers 0.000s
  2322.        time in global optimizers 0.000s
  2323.        time in final optimizers 0.001s
  2324.        time in cleanup optimizers 0.000s
  2325.         0 - 0.002s 0 (0.002s in global opts, 0.000s io_toposort) - 2 nodes -
  2326.      Global, final and clean up optimizers
  2327.      Iter 0
  2328.              TopoOptimizer  topo_constant_folding
  2329.                nb_node (start, end, changed) (2, 2, 0)
  2330.                init io_toposort 0.00145483016968
  2331.                loop time 5.96046447754e-06
  2332.                callback_time 0.0
  2333.              MergeOptimizer
  2334.                nb fail=    0 merged=    0 constant=    0
  2335.                time replace=0.00 validate=0.00 callback=0.00
  2336.    0.001610s - ('gpuarray_opt', 'SeqOptimizer', 16, 2, 2) - 0.000s
  2337.      SeqOptimizer      gpuarray_opt  time 0.001s for 2/2 nodes before/after optimization
  2338.        0.001s for callback
  2339.            0.000s for fgraph.validate()
  2340.        0.000982s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 2, 2) - 0.000s
  2341.          GraphToGPUOptimizer          gpuarray_graph_optimization
  2342.            time io_toposort 0.000s
  2343.          Total time taken by local optimizers 0.000s
  2344.            times - times applied - Node created - name:
  2345.            0.000s - 1 - 1 - local_gpua_subtensor_graph
  2346.            0.000s - in 0 optimization that were not used (display only those with a runtime > 0)
  2347.  
  2348.        0.000393s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 2, 2) - 0.000s
  2349.          EquilibriumOptimizer          gpuarray_local_optimizations
  2350.            time 0.000s for 1 passes
  2351.            nb nodes (start, end,  max) 2 2 2
  2352.            time io_toposort 0.000s
  2353.            time in local optimizers 0.000s
  2354.            time in global optimizers 0.000s
  2355.            time in final optimizers 0.000s
  2356.            time in cleanup optimizers 0.000s
  2357.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2358.        0.000111s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 2, 2) - 0.000s
  2359.          EquilibriumOptimizer          gpuarray_cut_transfers
  2360.            time 0.000s for 1 passes
  2361.            nb nodes (start, end,  max) 2 2 2
  2362.            time io_toposort 0.000s
  2363.            time in local optimizers 0.000s
  2364.            time in global optimizers 0.000s
  2365.            time in final optimizers 0.000s
  2366.            time in cleanup optimizers 0.000s
  2367.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2368.        0.000010s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 2, 2) - 0.000s
  2369.  
  2370.    0.000769s - ('ShapeOpt', 'ShapeOptimizer', 2, 2, 2) - 0.000s
  2371.    0.000677s - ('BlasOpt', 'SeqOptimizer', 12, 2, 2) - 0.000s
  2372.      SeqOptimizer      BlasOpt  time 0.001s for 2/2 nodes before/after optimization
  2373.        0.000s for callback
  2374.            0.000s for fgraph.validate()
  2375.        0.000175s - ('gemm_optimizer', 'GemmOptimizer', 1, 2, 2) - 0.000s
  2376.          GemmOptimizer
  2377.           nb_iter 1
  2378.           nb_replacement 0
  2379.           nb_replacement_didn_t_remove 0
  2380.           nb_inconsistency_make 0
  2381.           nb_inconsistency_replace 0
  2382.           time_canonicalize 0
  2383.           time_factor_can 0
  2384.           time_factor_list 0
  2385.           time_toposort 2.90870666504e-05
  2386.           validate_time 0.0
  2387.           callback_time 0.0
  2388.        0.000139s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 2, 2) - 0.000s
  2389.          EquilibriumOptimizer          local_gemm_to_gemv
  2390.            time 0.000s for 1 passes
  2391.            nb nodes (start, end,  max) 2 2 2
  2392.            time io_toposort 0.000s
  2393.            time in local optimizers 0.000s
  2394.            time in global optimizers 0.000s
  2395.            time in final optimizers 0.000s
  2396.            time in cleanup optimizers 0.000s
  2397.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2398.        0.000077s - ('use_c_blas', 'TopoOptimizer', 4, 2, 2) - 0.000s
  2399.          TopoOptimizer  use_c_blas
  2400.            nb_node (start, end, changed) (2, 2, 0)
  2401.            init io_toposort 3.00407409668e-05
  2402.            loop time 1.90734863281e-05
  2403.            callback_time 0.0
  2404.  LocalOptGroup
  2405.  ---------------------
  2406.  --- The Optimizer wasn't successful ---
  2407.  
  2408.       0.000062s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 2, 2) - 0.000s
  2409.         TopoOptimizer  local_dot22_to_dot22scalar
  2410.           nb_node (start, end, changed) (2, 2, 0)
  2411.           init io_toposort 3.00407409668e-05
  2412.           loop time 5.96046447754e-06
  2413.           callback_time 0.0
  2414.       0.000060s - ('use_scipy_ger', 'TopoOptimizer', 5, 2, 2) - 0.000s
  2415.         TopoOptimizer  scipy_blas
  2416.           nb_node (start, end, changed) (2, 2, 0)
  2417.           init io_toposort 3.09944152832e-05
  2418.           loop time 4.05311584473e-06
  2419.           callback_time 0.0
  2420.       0.000060s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 2, 2) - 0.000s
  2421.         TopoOptimizer  local_dot_to_dot22
  2422.           nb_node (start, end, changed) (2, 2, 0)
  2423.           init io_toposort 2.88486480713e-05
  2424.           loop time 4.05311584473e-06
  2425.           callback_time 0.0
  2426.  
  2427.   0.000647s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 2, 2) - 0.000s
  2428.     EquilibriumOptimizer      scan_eqopt1
  2429.       time 0.001s for 1 passes
  2430.       nb nodes (start, end,  max) 2 2 2
  2431.       time io_toposort 0.000s
  2432.       time in local optimizers 0.000s
  2433.       time in global optimizers 0.000s
  2434.       time in final optimizers 0.000s
  2435.       time in cleanup optimizers 0.000s
  2436.        0 - 0.001s 0 (0.001s in global opts, 0.000s io_toposort) - 2 nodes -
  2437.     Global, final and clean up optimizers
  2438.     Iter 0
  2439.             SeqOptimizer              all_pushout_opt  time 0.000s for 2/2 nodes before/after optimization
  2440.               0.000s for callback
  2441.                   0.000s for fgraph.validate()
  2442.               0.000113s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 2, 2) - 0.000s
  2443.               0.000105s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 2, 2) - 0.000s
  2444.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  2445.                   nb_node (start, end, changed) (2, 2, 0)
  2446.                   init io_toposort 3.00407409668e-05
  2447.                   loop time 6.91413879395e-06
  2448.                   callback_time 0.0
  2449.               0.000084s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 2, 2) - 0.000s
  2450.               0.000084s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 2, 2) - 0.000s
  2451.               0.000077s - ('scan_pushout_dot1', 'PushOutDot1', 3, 2, 2) - 0.000s
  2452.  
  2453.   0.000638s - ('specialize', 'EquilibriumOptimizer', 13, 2, 2) - 0.000s
  2454.     EquilibriumOptimizer      specialize
  2455.       time 0.000s for 1 passes
  2456.       nb nodes (start, end,  max) 2 2 2
  2457.       time io_toposort 0.000s
  2458.       time in local optimizers 0.000s
  2459.       time in global optimizers 0.000s
  2460.       time in final optimizers 0.000s
  2461.       time in cleanup optimizers 0.000s
  2462.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2463.     Global, final and clean up optimizers
  2464.     Iter 0
  2465.             TopoOptimizer  topo_constant_folding
  2466.               nb_node (start, end, changed) (2, 2, 0)
  2467.               init io_toposort 4.41074371338e-05
  2468.               loop time 3.81469726562e-06
  2469.               callback_time 0.0
  2470.   0.000580s - ('add_destroy_handler', 'AddDestroyHandler', 23, 2, 2) - 0.000s
  2471.   0.000564s - ('elemwise_fusion', 'SeqOptimizer', 19, 2, 2) - 0.000s
  2472.     SeqOptimizer      elemwise_fusion  time 0.000s for 2/2 nodes before/after optimization
  2473.       0.000s for callback
  2474.           0.000s for fgraph.validate()
  2475.       0.000233s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 2, 2) - 0.000s
  2476.         FusionOptimizer
  2477.          nb_iter 1
  2478.          nb_replacement 0
  2479.          nb_inconsistency_replace 0
  2480.          validate_time 0.0
  2481.          callback_time 0.0
  2482.          time_toposort 5.29289245605e-05
  2483.       0.000217s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 2, 2) - 0.000s
  2484.         FusionOptimizer
  2485.          nb_iter 1
  2486.          nb_replacement 0
  2487.          nb_inconsistency_replace 0
  2488.          validate_time 0.0
  2489.          callback_time 0.0
  2490.          time_toposort 4.60147857666e-05
  2491.  
  2492.   0.000563s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 2, 2) - 0.000s
  2493.     EquilibriumOptimizer      scan_eqopt2
  2494.       time 0.000s for 1 passes
  2495.       nb nodes (start, end,  max) 2 2 2
  2496.       time io_toposort 0.000s
  2497.       time in local optimizers 0.000s
  2498.       time in global optimizers 0.000s
  2499.       time in final optimizers 0.000s
  2500.       time in cleanup optimizers 0.000s
  2501.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2502.     Global, final and clean up optimizers
  2503.     Iter 0
  2504.             TopoOptimizer  constant_folding_for_scan2
  2505.               nb_node (start, end, changed) (2, 2, 0)
  2506.               init io_toposort 3.09944152832e-05
  2507.               loop time 3.09944152832e-06
  2508.               callback_time 0.0
  2509.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  2510.               nb_node (start, end, changed) (2, 2, 0)
  2511.               init io_toposort 2.88486480713e-05
  2512.               loop time 4.05311584473e-06
  2513.               callback_time 0.0
  2514.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  2515.               nb_node (start, end, changed) (2, 2, 0)
  2516.               init io_toposort 2.8133392334e-05
  2517.               loop time 3.09944152832e-06
  2518.               callback_time 0.0
  2519.             TopoOptimizer  scanOp_merge_inouts
  2520.               nb_node (start, end, changed) (2, 2, 0)
  2521.               init io_toposort 2.90870666504e-05
  2522.               loop time 3.81469726562e-06
  2523.               callback_time 0.0
  2524.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  2525.               nb_node (start, end, changed) (2, 2, 0)
  2526.               init io_toposort 2.8133392334e-05
  2527.               loop time 2.86102294922e-06
  2528.               callback_time 0.0
  2529.   0.000438s - ('stabilize', 'EquilibriumOptimizer', 8, 2, 2) - 0.000s
  2530.     EquilibriumOptimizer      stabilize
  2531.       time 0.000s for 1 passes
  2532.       nb nodes (start, end,  max) 2 2 2
  2533.       time io_toposort 0.000s
  2534.       time in local optimizers 0.000s
  2535.       time in global optimizers 0.000s
  2536.       time in final optimizers 0.000s
  2537.       time in cleanup optimizers 0.000s
  2538.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2539.     Global, final and clean up optimizers
  2540.     Iter 0
  2541.             TopoOptimizer  topo_constant_folding
  2542.               nb_node (start, end, changed) (2, 2, 0)
  2543.               init io_toposort 3.31401824951e-05
  2544.               loop time 2.86102294922e-06
  2545.               callback_time 0.0
  2546.   0.000216s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 2, 2) - 0.000s
  2547.     FusionOptimizer
  2548.      nb_iter 1
  2549.      nb_replacement 0
  2550.      nb_inconsistency_replace 0
  2551.      validate_time 0.0
  2552.      callback_time 0.0
  2553.      time_toposort 4.60147857666e-05
  2554.   0.000215s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 2, 2) - 0.000s
  2555.     FusionOptimizer
  2556.      nb_iter 1
  2557.      nb_replacement 0
  2558.      nb_inconsistency_replace 0
  2559.      validate_time 0.0
  2560.      callback_time 0.0
  2561.      time_toposort 4.60147857666e-05
  2562.   0.000190s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 2, 2) - 0.000s
  2563.     EquilibriumOptimizer      uncanonicalize
  2564.       time 0.000s for 1 passes
  2565.       nb nodes (start, end,  max) 2 2 2
  2566.       time io_toposort 0.000s
  2567.       time in local optimizers 0.000s
  2568.       time in global optimizers 0.000s
  2569.       time in final optimizers 0.000s
  2570.       time in cleanup optimizers 0.000s
  2571.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2572.     Global, final and clean up optimizers
  2573.     Iter 0
  2574.             TopoOptimizer  topo_constant_folding
  2575.               nb_node (start, end, changed) (2, 2, 0)
  2576.               init io_toposort 2.69412994385e-05
  2577.               loop time 2.86102294922e-06
  2578.               callback_time 0.0
  2579.   0.000189s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 2, 2) - 0.000s
  2580.     TopoOptimizer  local_gemm16_inplace
  2581.       nb_node (start, end, changed) (2, 2, 0)
  2582.       init io_toposort 3.09944152832e-05
  2583.       loop time 4.05311584473e-06
  2584.       callback_time 0.0
  2585.   0.000180s - ('merge3', 'MergeOptimizer', 51, 2, 2) - 0.000s
  2586.     MergeOptimizer
  2587.       nb fail=    0 merged=    0 constant=    0
  2588.       time replace=0.00 validate=0.00 callback=0.00
  2589.   0.000135s - ('useless', 'TopoOptimizer', 3, 2, 2) - 0.000s
  2590.     TopoOptimizer  useless
  2591.       nb_node (start, end, changed) (2, 2, 0)
  2592.       init io_toposort 4.10079956055e-05
  2593.       loop time 5.3882598877e-05
  2594.       callback_time 0.0
  2595. LocalOptGroup
  2596. ---------------------
  2597.   time taken - times applied - times tried - name - node_created:
  2598.   -0.000s - 0 - 1 - local_subtensor_of_alloc - 0
  2599.   -0.000s - 0 - 1 - local_subtensor_make_vector - 0
  2600.   -0.000s - 0 - 1 - local_useless_slice - 0
  2601.   0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  2602.  
  2603.   0.000117s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 2, 2) - 0.000s
  2604.     TopoOptimizer  InplaceGpuBlasOpt
  2605.       nb_node (start, end, changed) (2, 2, 0)
  2606.       init io_toposort 2.90870666504e-05
  2607.       loop time 1.59740447998e-05
  2608.       callback_time 0.0
  2609. LocalOptGroup
  2610. ---------------------
  2611. --- The Optimizer wasn't successful ---
  2612.  
  2613.    0.000115s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 2, 2) - 0.000s
  2614.      TopoOptimizer  InplaceGpuaBlasOpt
  2615.        nb_node (start, end, changed) (2, 2, 0)
  2616.        init io_toposort 2.8133392334e-05
  2617.        loop time 1.59740447998e-05
  2618.        callback_time 0.0
  2619.  LocalOptGroup
  2620.  ---------------------
  2621.    time taken - times applied - times tried - name - node_created:
  2622.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  2623.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  2624.  
  2625.    0.000114s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 2, 2) - 0.000s
  2626.      TopoOptimizer  local_dnna_conv_inplace
  2627.        nb_node (start, end, changed) (2, 2, 0)
  2628.        init io_toposort 2.50339508057e-05
  2629.        loop time 1.81198120117e-05
  2630.        callback_time 0.0
  2631.  LocalOptGroup
  2632.  ---------------------
  2633.    time taken - times applied - times tried - name - node_created:
  2634.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  2635.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  2636.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  2637.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  2638.  
  2639.    0.000114s - ('blas_opt_inplace', 'TopoOptimizer', 34, 2, 2) - 0.000s
  2640.      TopoOptimizer  InplaceBlasOpt
  2641.        nb_node (start, end, changed) (2, 2, 0)
  2642.        init io_toposort 2.50339508057e-05
  2643.        loop time 2.00271606445e-05
  2644.        callback_time 0.0
  2645.  LocalOptGroup
  2646.  ---------------------
  2647.  --- The Optimizer wasn't successful ---
  2648.  
  2649.   0.000112s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 2, 2) - 0.000s
  2650.     TopoOptimizer  local_dnn_conv_inplace
  2651.       nb_node (start, end, changed) (2, 2, 0)
  2652.       init io_toposort 2.59876251221e-05
  2653.       loop time 1.47819519043e-05
  2654.       callback_time 0.0
  2655. LocalOptGroup
  2656. ---------------------
  2657. --- The Optimizer wasn't successful ---
  2658.  
  2659.    0.000112s - ('specialize_device', 'EquilibriumOptimizer', 17, 2, 2) - 0.000s
  2660.      EquilibriumOptimizer      specialize_device
  2661.        time 0.000s for 1 passes
  2662.        nb nodes (start, end,  max) 2 2 2
  2663.        time io_toposort 0.000s
  2664.        time in local optimizers 0.000s
  2665.        time in global optimizers 0.000s
  2666.        time in final optimizers 0.000s
  2667.        time in cleanup optimizers 0.000s
  2668.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 2 nodes -
  2669.    0.000111s - ('merge2', 'MergeOptimizer', 22, 2, 2) - 0.000s
  2670.      MergeOptimizer
  2671.        nb fail=    0 merged=    0 constant=    0
  2672.        time replace=0.00 validate=0.00 callback=0.00
  2673.    0.000109s - ('merge1', 'MergeOptimizer', 0, 2, 2) - 0.000s
  2674.      MergeOptimizer
  2675.        nb fail=    0 merged=    0 constant=    0
  2676.        time replace=0.00 validate=0.00 callback=0.00
  2677.    0.000079s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 2, 2) - 0.000s
  2678.    0.000077s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 2, 2) - 0.000s
  2679.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  2680.        nb_node (start, end, changed) (2, 2, 0)
  2681.        init io_toposort 2.90870666504e-05
  2682.        loop time 3.81469726562e-06
  2683.        callback_time 0.0
  2684.    0.000071s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 2, 2) - 0.000s
  2685.    0.000070s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 2, 2) - 0.000s
  2686.      TopoOptimizer  dimshuffle_as_view
  2687.        nb_node (start, end, changed) (2, 2, 0)
  2688.        init io_toposort 3.40938568115e-05
  2689.        loop time 4.05311584473e-06
  2690.        callback_time 0.0
  2691.    0.000070s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 2, 2) - 0.000s
  2692.      TopoOptimizer  pre_local_IncSubtensor_serialize
  2693.        nb_node (start, end, changed) (2, 2, 0)
  2694.        init io_toposort 3.50475311279e-05
  2695.        loop time 6.91413879395e-06
  2696.        callback_time 0.0
  2697.    0.000070s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 2, 2) - 0.000s
  2698.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  2699.        nb_node (start, end, changed) (2, 2, 0)
  2700.        init io_toposort 2.59876251221e-05
  2701.        loop time 2.86102294922e-06
  2702.        callback_time 0.0
  2703.    0.000068s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 2, 2) - 0.000s
  2704.      TopoOptimizer  local_inplace_sparse_block_outer
  2705.        nb_node (start, end, changed) (2, 2, 0)
  2706.        init io_toposort 2.59876251221e-05
  2707.        loop time 4.05311584473e-06
  2708.        callback_time 0.0
  2709.    0.000068s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 2, 2) - 0.000s
  2710.      TopoOptimizer  local_inplace_setsubtensor
  2711.        nb_node (start, end, changed) (2, 2, 0)
  2712.        init io_toposort 2.50339508057e-05
  2713.        loop time 3.09944152832e-06
  2714.        callback_time 0.0
  2715.    0.000067s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 2, 2) - 0.000s
  2716.      TopoOptimizer  local_inplace_sparseblockgemv
  2717.        nb_node (start, end, changed) (2, 2, 0)
  2718.        init io_toposort 2.50339508057e-05
  2719.        loop time 3.09944152832e-06
  2720.        callback_time 0.0
  2721.    0.000067s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 2, 2) - 0.000s
  2722.      TopoOptimizer  local_inplace_sparse_block_gemv
  2723.        nb_node (start, end, changed) (2, 2, 0)
  2724.        init io_toposort 2.50339508057e-05
  2725.        loop time 2.86102294922e-06
  2726.        callback_time 0.0
  2727.    0.000067s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 2, 2) - 0.000s
  2728.      TopoOptimizer  local_inplace_incsubtensor1
  2729.        nb_node (start, end, changed) (2, 2, 0)
  2730.        init io_toposort 2.50339508057e-05
  2731.        loop time 4.05311584473e-06
  2732.        callback_time 0.0
  2733.    0.000066s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 2, 2) - 0.000s
  2734.      TopoOptimizer  local_inplace_sparseblockouter
  2735.        nb_node (start, end, changed) (2, 2, 0)
  2736.        init io_toposort 2.40802764893e-05
  2737.        loop time 3.09944152832e-06
  2738.        callback_time 0.0
  2739.    0.000064s - ('c_blas_destructive', 'TopoOptimizer', 37, 2, 2) - 0.000s
  2740.      TopoOptimizer  c_blas_destructive
  2741.        nb_node (start, end, changed) (2, 2, 0)
  2742.        init io_toposort 2.50339508057e-05
  2743.        loop time 1.4066696167e-05
  2744.        callback_time 0.0
  2745.  LocalOptGroup
  2746.  ---------------------
  2747.  --- The Optimizer wasn't successful ---
  2748.  
  2749.   0.000063s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 2, 2) - 0.000s
  2750.     TopoOptimizer  local_fill_to_alloc
  2751.       nb_node (start, end, changed) (2, 2, 0)
  2752.       init io_toposort 3.09944152832e-05
  2753.       loop time 5.00679016113e-06
  2754.       callback_time 0.0
  2755.   0.000061s - ('local_destructive', 'TopoOptimizer', 48, 2, 2) - 0.000s
  2756.     TopoOptimizer  CURAND_destructive
  2757.       nb_node (start, end, changed) (2, 2, 0)
  2758.       init io_toposort 3.50475311279e-05
  2759.       loop time 2.86102294922e-06
  2760.       callback_time 0.0
  2761.   0.000061s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 2, 2) - 0.000s
  2762.   0.000059s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 2, 2) - 0.000s
  2763.   0.000058s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 2, 2) - 0.000s
  2764.     TopoOptimizer  local_elemwise_alloc
  2765.       nb_node (start, end, changed) (2, 2, 0)
  2766.       init io_toposort 3.00407409668e-05
  2767.       loop time 3.81469726562e-06
  2768.       callback_time 0.0
  2769.   0.000056s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 2, 2) - 0.000s
  2770.     TopoOptimizer  local_advincsub1_gpua_inplace
  2771.       nb_node (start, end, changed) (2, 2, 0)
  2772.       init io_toposort 2.78949737549e-05
  2773.       loop time 4.05311584473e-06
  2774.       callback_time 0.0
  2775.   0.000055s - ('make_ger_destructive', 'TopoOptimizer', 41, 2, 2) - 0.000s
  2776.     TopoOptimizer  make_scipy_blas_destructive
  2777.       nb_node (start, end, changed) (2, 2, 0)
  2778.       init io_toposort 2.59876251221e-05
  2779.       loop time 5.00679016113e-06
  2780.       callback_time 0.0
  2781.   0.000054s - ('random_make_inplace', 'TopoOptimizer', 49, 2, 2) - 0.000s
  2782.     TopoOptimizer  random_make_inplace
  2783.       nb_node (start, end, changed) (2, 2, 0)
  2784.       init io_toposort 2.38418579102e-05
  2785.       loop time 4.05311584473e-06
  2786.       callback_time 0.0
  2787.   0.000054s - ('AbstractConvCheck', 'TopoOptimizer', 18, 2, 2) - 0.000s
  2788.     TopoOptimizer  AbstractConvCheck
  2789.       nb_node (start, end, changed) (2, 2, 0)
  2790.       init io_toposort 2.71797180176e-05
  2791.       loop time 5.00679016113e-06
  2792.       callback_time 0.0
  2793.   0.000053s - ('cond_make_inplace', 'TopoOptimizer', 47, 2, 2) - 0.000s
  2794.     TopoOptimizer  cond_make_inplace
  2795.       nb_node (start, end, changed) (2, 2, 0)
  2796.       init io_toposort 2.40802764893e-05
  2797.       loop time 4.05311584473e-06
  2798.       callback_time 0.0
  2799.   0.000052s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 2, 2) - 0.000s
  2800.     TopoOptimizer  random_make_inplace_mrg
  2801.       nb_node (start, end, changed) (2, 2, 0)
  2802.       init io_toposort 2.38418579102e-05
  2803.       loop time 5.00679016113e-06
  2804.       callback_time 0.0
  2805.   0.000050s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 2, 2) - 0.000s
  2806.   0.000050s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 2, 2) - 0.000s
  2807.   0.000035s - ('merge1.2', 'MergeOptimizer', 7, 2, 2) - 0.000s
  2808.     MergeOptimizer
  2809.       nb fail=    0 merged=    0 constant=    0
  2810.       time replace=0.00 validate=0.00 callback=0.00
  2811.   0.000025s - ('merge1.1', 'MergeOptimizer', 4, 2, 2) - 0.000s
  2812.     MergeOptimizer
  2813.       nb fail=    0 merged=    0 constant=    0
  2814.       time replace=0.00 validate=0.00 callback=0.00
  2815.  
  2816. Here are tips to potentially make your code run faster
  2817.                 (if you think of new ones, suggest them on the mailing list).
  2818.                 Test them first, as they are not guaranteed to always provide a speedup.
  2819.  Sorry, no tip for today.
  2820. Function profiling
  2821. ==================
  2822.  Message: /Users/Ramana/projects/SBRNN/sb/utils.py:56
  2823.  Time in 1 calls to Function.__call__: 2.908707e-05s
  2824.  Time in Function.fn.__call__: 1.001358e-05s (34.426%)
  2825.  Time in thunks: 5.006790e-06s (17.213%)
  2826.  Total compile time: 3.316283e-02s
  2827.    Number of Apply nodes: 1
  2828.    Theano Optimizer time: 1.118994e-02s
  2829.       Theano validate time: 7.295609e-05s
  2830.    Theano Linker time (includes C, CUDA code generation/compiling): 5.037069e-03s
  2831.       Import time 4.065990e-03s
  2832.       Node make_thunk time 4.762173e-03s
  2833.  
  2834. Time in all call to theano.grad() 2.656322e+00s
  2835. Time since theano import 477.808s
  2836. Class
  2837. ---
  2838. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  2839.  100.0%   100.0%       0.000s       5.01e-06s     C        1       1   theano.compile.ops.DeepCopyOp
  2840.   ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  2841.  
  2842. Ops
  2843. ---
  2844. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  2845.  100.0%   100.0%       0.000s       5.01e-06s     C        1        1   DeepCopyOp
  2846.   ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  2847.  
  2848. Apply
  2849. ------
  2850. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  2851.  100.0%   100.0%       0.000s       5.01e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  2852.   ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  2853.  
  2854. Optimizer Profile
  2855. -----------------
  2856. SeqOptimizer  OPT_FAST_RUN  time 0.011s for 1/0 nodes before/after optimization
  2857.   0.001s for callback
  2858.       0.000s for fgraph.validate()
  2859.   time      - (name, class, index, nodes before, nodes after) - validate time
  2860.   0.003724s - ('canonicalize', 'EquilibriumOptimizer', 6, 1, 0) - 0.000s
  2861.     EquilibriumOptimizer      canonicalize
  2862.       time 0.003s for 2 passes
  2863.       nb nodes (start, end,  max) 1 0 1
  2864.       time io_toposort 0.000s
  2865.       time in local optimizers 0.001s
  2866.       time in global optimizers 0.000s
  2867.       time in final optimizers 0.002s
  2868.       time in cleanup optimizers 0.000s
  2869.        0 - 0.003s 4 (0.002s in global opts, 0.000s io_toposort) - 1 nodes - ('topo_constant_folding', 1) ('local_upcast_elemwise_constant_inputs', 1) ('local_dimshuffle_lift', 1) ('MergeOptimizer', 1)
  2870.        1 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  2871.       times - times applied - nb node created - name:
  2872.       0.002s - 1 - 0 - topo_constant_folding
  2873.       0.001s - 1 - 3 - local_upcast_elemwise_constant_inputs
  2874.       0.000s - 1 - 1 - local_dimshuffle_lift
  2875.       0.000s - 1 - 1 - MergeOptimizer
  2876.       0.000s - in 85 optimization that were not used (display only those with a runtime > 0)
  2877.         0.000s - local_func_inv
  2878.         0.000s - local_useless_elemwise
  2879.         0.000s - local_fill_sink
  2880.         0.000s - local_expm1
  2881.         0.000s - local_track_shape_i
  2882.         0.000s - local_merge_switch_same_cond
  2883.         0.000s - local_cast_cast
  2884.         0.000s - local_fill_cut
  2885.         0.000s - local_useless_switch
  2886.         0.000s - local_useless_elemwise_comparison
  2887.         0.000s - local_lift_transpose_through_dot
  2888.  
  2889.     Global, final and clean up optimizers
  2890.     Iter 0
  2891.             TopoOptimizer  topo_constant_folding
  2892.               nb_node (start, end, changed) (2, 0, 2)
  2893.               init io_toposort 2.59876251221e-05
  2894.               loop time 0.00179696083069
  2895.               callback_time 0.00019383430481
  2896.             MergeOptimizer
  2897.               nb fail=    0 merged=    1 constant=    0
  2898.               time replace=0.00 validate=0.00 callback=0.00
  2899.     Iter 1
  2900.             TopoOptimizer  topo_constant_folding
  2901.               nb_node (start, end, changed) (0, 0, 0)
  2902.               init io_toposort 8.10623168945e-06
  2903.               loop time 0.0
  2904.               callback_time 0.0
  2905.             MergeOptimizer
  2906.               nb fail=    0 merged=    0 constant=    0
  2907.               time replace=0.00 validate=0.00 callback=0.00
  2908.   0.001101s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 0, 0) - 0.000s
  2909.   0.000595s - ('add_destroy_handler', 'AddDestroyHandler', 23, 0, 0) - 0.000s
  2910.   0.000571s - ('elemwise_fusion', 'SeqOptimizer', 19, 0, 0) - 0.000s
  2911.     SeqOptimizer      elemwise_fusion  time 0.000s for 0/0 nodes before/after optimization
  2912.       0.000s for callback
  2913.           0.000s for fgraph.validate()
  2914.       0.000229s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 0, 0) - 0.000s
  2915.         FusionOptimizer
  2916.          nb_iter 1
  2917.          nb_replacement 0
  2918.          nb_inconsistency_replace 0
  2919.          validate_time 0.0
  2920.          callback_time 0.0
  2921.          time_toposort 9.53674316406e-07
  2922.       0.000203s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 0, 0) - 0.000s
  2923.         FusionOptimizer
  2924.          nb_iter 1
  2925.          nb_replacement 0
  2926.          nb_inconsistency_replace 0
  2927.          validate_time 0.0
  2928.          callback_time 0.0
  2929.          time_toposort 2.14576721191e-06
  2930.  
  2931.   0.000536s - ('gpuarray_opt', 'SeqOptimizer', 16, 0, 0) - 0.000s
  2932.     SeqOptimizer      gpuarray_opt  time 0.000s for 0/0 nodes before/after optimization
  2933.       0.000s for callback
  2934.           0.000s for fgraph.validate()
  2935.       0.000300s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 0, 0) - 0.000s
  2936.         EquilibriumOptimizer          gpuarray_local_optimizations
  2937.           time 0.000s for 1 passes
  2938.           nb nodes (start, end,  max) 0 0 0
  2939.           time io_toposort 0.000s
  2940.           time in local optimizers 0.000s
  2941.           time in global optimizers 0.000s
  2942.           time in final optimizers 0.000s
  2943.           time in cleanup optimizers 0.000s
  2944.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  2945.       0.000068s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  2946.         EquilibriumOptimizer          gpuarray_cut_transfers
  2947.           time 0.000s for 1 passes
  2948.           nb nodes (start, end,  max) 0 0 0
  2949.           time io_toposort 0.000s
  2950.           time in local optimizers 0.000s
  2951.           time in global optimizers 0.000s
  2952.           time in final optimizers 0.000s
  2953.           time in cleanup optimizers 0.000s
  2954.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  2955.       0.000045s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 0, 0) - 0.000s
  2956.         GraphToGPUOptimizer          gpuarray_graph_optimization
  2957.           time io_toposort 0.000s
  2958.         Total time taken by local optimizers 0.000s
  2959.       0.000007s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 0, 0) - 0.000s
  2960.  
  2961.   0.000424s - ('BlasOpt', 'SeqOptimizer', 12, 0, 0) - 0.000s
  2962.     SeqOptimizer      BlasOpt  time 0.000s for 0/0 nodes before/after optimization
  2963.       0.000s for callback
  2964.           0.000s for fgraph.validate()
  2965.       0.000135s - ('gemm_optimizer', 'GemmOptimizer', 1, 0, 0) - 0.000s
  2966.         GemmOptimizer
  2967.          nb_iter 1
  2968.          nb_replacement 0
  2969.          nb_replacement_didn_t_remove 0
  2970.          nb_inconsistency_make 0
  2971.          nb_inconsistency_replace 0
  2972.          time_canonicalize 0
  2973.          time_factor_can 0
  2974.          time_factor_list 0
  2975.          time_toposort 6.91413879395e-06
  2976.          validate_time 0.0
  2977.          callback_time 0.0
  2978.       0.000089s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  2979.         EquilibriumOptimizer          local_gemm_to_gemv
  2980.           time 0.000s for 1 passes
  2981.           nb nodes (start, end,  max) 0 0 0
  2982.           time io_toposort 0.000s
  2983.           time in local optimizers 0.000s
  2984.           time in global optimizers 0.000s
  2985.           time in final optimizers 0.000s
  2986.           time in cleanup optimizers 0.000s
  2987.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  2988.       0.000029s - ('use_c_blas', 'TopoOptimizer', 4, 0, 0) - 0.000s
  2989.         TopoOptimizer  use_c_blas
  2990.           nb_node (start, end, changed) (0, 0, 0)
  2991.           init io_toposort 5.96046447754e-06
  2992.           loop time 0.0
  2993.           callback_time 0.0
  2994. LocalOptGroup
  2995. ---------------------
  2996. --- The Optimizer wasn't successful ---
  2997.  
  2998.        0.000027s - ('use_scipy_ger', 'TopoOptimizer', 5, 0, 0) - 0.000s
  2999.          TopoOptimizer  scipy_blas
  3000.            nb_node (start, end, changed) (0, 0, 0)
  3001.            init io_toposort 5.96046447754e-06
  3002.            loop time 0.0
  3003.            callback_time 0.0
  3004.        0.000027s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 0, 0) - 0.000s
  3005.          TopoOptimizer  local_dot22_to_dot22scalar
  3006.            nb_node (start, end, changed) (0, 0, 0)
  3007.            init io_toposort 6.19888305664e-06
  3008.            loop time 0.0
  3009.            callback_time 0.0
  3010.        0.000027s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 0, 0) - 0.000s
  3011.          TopoOptimizer  local_dot_to_dot22
  3012.            nb_node (start, end, changed) (0, 0, 0)
  3013.            init io_toposort 7.15255737305e-06
  3014.            loop time 0.0
  3015.            callback_time 0.0
  3016.  
  3017.    0.000361s - ('specialize', 'EquilibriumOptimizer', 13, 0, 0) - 0.000s
  3018.      EquilibriumOptimizer      specialize
  3019.        time 0.000s for 1 passes
  3020.        nb nodes (start, end,  max) 0 0 0
  3021.        time io_toposort 0.000s
  3022.        time in local optimizers 0.000s
  3023.        time in global optimizers 0.000s
  3024.        time in final optimizers 0.000s
  3025.        time in cleanup optimizers 0.000s
  3026.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3027.      Global, final and clean up optimizers
  3028.      Iter 0
  3029.              TopoOptimizer  topo_constant_folding
  3030.                nb_node (start, end, changed) (0, 0, 0)
  3031.                init io_toposort 6.91413879395e-06
  3032.                loop time 0.0
  3033.                callback_time 0.0
  3034.    0.000274s - ('stabilize', 'EquilibriumOptimizer', 8, 0, 0) - 0.000s
  3035.      EquilibriumOptimizer      stabilize
  3036.        time 0.000s for 1 passes
  3037.        nb nodes (start, end,  max) 0 0 0
  3038.        time io_toposort 0.000s
  3039.        time in local optimizers 0.000s
  3040.        time in global optimizers 0.000s
  3041.        time in final optimizers 0.000s
  3042.        time in cleanup optimizers 0.000s
  3043.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3044.      Global, final and clean up optimizers
  3045.      Iter 0
  3046.              TopoOptimizer  topo_constant_folding
  3047.                nb_node (start, end, changed) (0, 0, 0)
  3048.                init io_toposort 6.91413879395e-06
  3049.                loop time 0.0
  3050.                callback_time 0.0
  3051.    0.000241s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 0, 0) - 0.000s
  3052.      EquilibriumOptimizer      scan_eqopt2
  3053.        time 0.000s for 1 passes
  3054.        nb nodes (start, end,  max) 0 0 0
  3055.        time io_toposort 0.000s
  3056.        time in local optimizers 0.000s
  3057.        time in global optimizers 0.000s
  3058.        time in final optimizers 0.000s
  3059.        time in cleanup optimizers 0.000s
  3060.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3061.      Global, final and clean up optimizers
  3062.      Iter 0
  3063.              TopoOptimizer  constant_folding_for_scan2
  3064.                nb_node (start, end, changed) (0, 0, 0)
  3065.                init io_toposort 7.86781311035e-06
  3066.                loop time 0.0
  3067.                callback_time 0.0
  3068.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  3069.                nb_node (start, end, changed) (0, 0, 0)
  3070.                init io_toposort 6.91413879395e-06
  3071.                loop time 1.19209289551e-06
  3072.                callback_time 0.0
  3073.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  3074.                nb_node (start, end, changed) (0, 0, 0)
  3075.                init io_toposort 6.19888305664e-06
  3076.                loop time 0.0
  3077.                callback_time 0.0
  3078.              TopoOptimizer  scanOp_merge_inouts
  3079.                nb_node (start, end, changed) (0, 0, 0)
  3080.                init io_toposort 5.96046447754e-06
  3081.                loop time 9.53674316406e-07
  3082.                callback_time 0.0
  3083.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  3084.                nb_node (start, end, changed) (0, 0, 0)
  3085.                init io_toposort 5.96046447754e-06
  3086.                loop time 0.0
  3087.                callback_time 0.0
  3088.    0.000234s - ('merge3', 'MergeOptimizer', 51, 0, 0) - 0.000s
  3089.      MergeOptimizer
  3090.        nb fail=    0 merged=    0 constant=    0
  3091.        time replace=0.00 validate=0.00 callback=0.00
  3092.    0.000208s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 0, 0) - 0.000s
  3093.      FusionOptimizer
  3094.       nb_iter 1
  3095.       nb_replacement 0
  3096.       nb_inconsistency_replace 0
  3097.       validate_time 0.0
  3098.       callback_time 0.0
  3099.       time_toposort 2.14576721191e-06
  3100.    0.000205s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 1, 1) - 0.000s
  3101.      EquilibriumOptimizer      scan_eqopt1
  3102.        time 0.000s for 1 passes
  3103.        nb nodes (start, end,  max) 1 1 1
  3104.        time io_toposort 0.000s
  3105.        time in local optimizers 0.000s
  3106.        time in global optimizers 0.000s
  3107.        time in final optimizers 0.000s
  3108.        time in cleanup optimizers 0.000s
  3109.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  3110.      Global, final and clean up optimizers
  3111.      Iter 0
  3112.              SeqOptimizer              all_pushout_opt  time 0.000s for 1/1 nodes before/after optimization
  3113.                0.000s for callback
  3114.                    0.000s for fgraph.validate()
  3115.                0.000061s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 1, 1) - 0.000s
  3116.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  3117.                    nb_node (start, end, changed) (1, 1, 0)
  3118.                    init io_toposort 2.09808349609e-05
  3119.                    loop time 4.05311584473e-06
  3120.                    callback_time 0.0
  3121.                0.000013s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 1, 1) - 0.000s
  3122.                0.000008s - ('scan_pushout_dot1', 'PushOutDot1', 3, 1, 1) - 0.000s
  3123.                0.000008s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 1, 1) - 0.000s
  3124.                0.000008s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 1, 1) - 0.000s
  3125.  
  3126.    0.000196s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 0, 0) - 0.000s
  3127.      FusionOptimizer
  3128.       nb_iter 1
  3129.       nb_replacement 0
  3130.       nb_inconsistency_replace 0
  3131.       validate_time 0.0
  3132.       callback_time 0.0
  3133.       time_toposort 9.53674316406e-07
  3134.    0.000131s - ('merge2', 'MergeOptimizer', 22, 0, 0) - 0.000s
  3135.      MergeOptimizer
  3136.        nb fail=    0 merged=    0 constant=    0
  3137.        time replace=0.00 validate=0.00 callback=0.00
  3138.    0.000124s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 0, 0) - 0.000s
  3139.      EquilibriumOptimizer      uncanonicalize
  3140.        time 0.000s for 1 passes
  3141.        nb nodes (start, end,  max) 0 0 0
  3142.        time io_toposort 0.000s
  3143.        time in local optimizers 0.000s
  3144.        time in global optimizers 0.000s
  3145.        time in final optimizers 0.000s
  3146.        time in cleanup optimizers 0.000s
  3147.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3148.      Global, final and clean up optimizers
  3149.      Iter 0
  3150.              TopoOptimizer  topo_constant_folding
  3151.                nb_node (start, end, changed) (0, 0, 0)
  3152.                init io_toposort 6.91413879395e-06
  3153.                loop time 0.0
  3154.                callback_time 0.0
  3155.    0.000108s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 0, 0) - 0.000s
  3156.      TopoOptimizer  local_dnna_conv_inplace
  3157.        nb_node (start, end, changed) (0, 0, 0)
  3158.        init io_toposort 7.86781311035e-06
  3159.        loop time 0.0
  3160.        callback_time 0.0
  3161.  LocalOptGroup
  3162.  ---------------------
  3163.    time taken - times applied - times tried - name - node_created:
  3164.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  3165.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  3166.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  3167.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  3168.  
  3169.    0.000104s - ('specialize_device', 'EquilibriumOptimizer', 17, 0, 0) - 0.000s
  3170.      EquilibriumOptimizer      specialize_device
  3171.        time 0.000s for 1 passes
  3172.        nb nodes (start, end,  max) 0 0 0
  3173.        time io_toposort 0.000s
  3174.        time in local optimizers 0.000s
  3175.        time in global optimizers 0.000s
  3176.        time in final optimizers 0.000s
  3177.        time in cleanup optimizers 0.000s
  3178.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3179.    0.000096s - ('blas_opt_inplace', 'TopoOptimizer', 34, 0, 0) - 0.000s
  3180.      TopoOptimizer  InplaceBlasOpt
  3181.        nb_node (start, end, changed) (0, 0, 0)
  3182.        init io_toposort 7.86781311035e-06
  3183.        loop time 0.0
  3184.        callback_time 0.0
  3185.  LocalOptGroup
  3186.  ---------------------
  3187.  --- The Optimizer wasn't successful ---
  3188.  
  3189.   0.000094s - ('merge1', 'MergeOptimizer', 0, 1, 1) - 0.000s
  3190.     MergeOptimizer
  3191.       nb fail=    0 merged=    0 constant=    0
  3192.       time replace=0.00 validate=0.00 callback=0.00
  3193.   0.000085s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 0, 0) - 0.000s
  3194.     TopoOptimizer  InplaceGpuBlasOpt
  3195.       nb_node (start, end, changed) (0, 0, 0)
  3196.       init io_toposort 8.10623168945e-06
  3197.       loop time 0.0
  3198.       callback_time 0.0
  3199. LocalOptGroup
  3200. ---------------------
  3201. --- The Optimizer wasn't successful ---
  3202.  
  3203.    0.000085s - ('useless', 'TopoOptimizer', 3, 1, 1) - 0.000s
  3204.      TopoOptimizer  useless
  3205.        nb_node (start, end, changed) (1, 1, 0)
  3206.        init io_toposort 1.81198120117e-05
  3207.        loop time 3.69548797607e-05
  3208.        callback_time 0.0
  3209.  LocalOptGroup
  3210.  ---------------------
  3211.    time taken - times applied - times tried - name - node_created:
  3212.    -0.000s - 0 - 1 - local_useless_switch - 0
  3213.    -0.000s - 0 - 1 - local_useless_elemwise_comparison - 0
  3214.    -0.000s - 0 - 1 - local_useless_elemwise - 0
  3215.    0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  3216.  
  3217.    0.000084s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 0, 0) - 0.000s
  3218.      TopoOptimizer  InplaceGpuaBlasOpt
  3219.        nb_node (start, end, changed) (0, 0, 0)
  3220.        init io_toposort 8.10623168945e-06
  3221.        loop time 0.0
  3222.        callback_time 0.0
  3223.  LocalOptGroup
  3224.  ---------------------
  3225.    time taken - times applied - times tried - name - node_created:
  3226.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  3227.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  3228.  
  3229.    0.000083s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 0, 0) - 0.000s
  3230.      TopoOptimizer  local_dnn_conv_inplace
  3231.        nb_node (start, end, changed) (0, 0, 0)
  3232.        init io_toposort 6.91413879395e-06
  3233.        loop time 0.0
  3234.        callback_time 0.0
  3235.  LocalOptGroup
  3236.  ---------------------
  3237.  --- The Optimizer wasn't successful ---
  3238.  
  3239.   0.000072s - ('ShapeOpt', 'ShapeOptimizer', 2, 1, 1) - 0.000s
  3240.   0.000065s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 0, 0) - 0.000s
  3241.     TopoOptimizer  local_inplace_setsubtensor
  3242.       nb_node (start, end, changed) (0, 0, 0)
  3243.       init io_toposort 7.15255737305e-06
  3244.       loop time 0.0
  3245.       callback_time 0.0
  3246.   0.000058s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 0, 0) - 0.000s
  3247.   0.000058s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 0, 0) - 0.000s
  3248.     TopoOptimizer  local_inplace_sparseblockouter
  3249.       nb_node (start, end, changed) (0, 0, 0)
  3250.       init io_toposort 6.91413879395e-06
  3251.       loop time 0.0
  3252.       callback_time 0.0
  3253.   0.000058s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 0, 0) - 0.000s
  3254.     TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  3255.       nb_node (start, end, changed) (0, 0, 0)
  3256.       init io_toposort 8.82148742676e-06
  3257.       loop time 0.0
  3258.       callback_time 0.0
  3259.   0.000055s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 0, 0) - 0.000s
  3260.     TopoOptimizer  local_inplace_sparse_block_gemv
  3261.       nb_node (start, end, changed) (0, 0, 0)
  3262.       init io_toposort 7.86781311035e-06
  3263.       loop time 1.19209289551e-06
  3264.       callback_time 0.0
  3265.   0.000052s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 0, 0) - 0.000s
  3266.     TopoOptimizer  local_gemm16_inplace
  3267.       nb_node (start, end, changed) (0, 0, 0)
  3268.       init io_toposort 8.10623168945e-06
  3269.       loop time 9.53674316406e-07
  3270.       callback_time 0.0
  3271.   0.000051s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 0, 0) - 0.000s
  3272.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  3273.       nb_node (start, end, changed) (0, 0, 0)
  3274.       init io_toposort 8.10623168945e-06
  3275.       loop time 0.0
  3276.       callback_time 0.0
  3277.   0.000050s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 0, 0) - 0.000s
  3278.     TopoOptimizer  local_inplace_incsubtensor1
  3279.       nb_node (start, end, changed) (0, 0, 0)
  3280.       init io_toposort 8.10623168945e-06
  3281.       loop time 0.0
  3282.       callback_time 0.0
  3283.   0.000050s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 0, 0) - 0.000s
  3284.     TopoOptimizer  local_inplace_sparse_block_outer
  3285.       nb_node (start, end, changed) (0, 0, 0)
  3286.       init io_toposort 7.86781311035e-06
  3287.       loop time 0.0
  3288.       callback_time 0.0
  3289.   0.000049s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 0, 0) - 0.000s
  3290.     TopoOptimizer  local_inplace_sparseblockgemv
  3291.       nb_node (start, end, changed) (0, 0, 0)
  3292.       init io_toposort 7.86781311035e-06
  3293.       loop time 9.53674316406e-07
  3294.       callback_time 0.0
  3295.   0.000046s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 1, 1) - 0.000s
  3296.     TopoOptimizer  pre_local_IncSubtensor_serialize
  3297.       nb_node (start, end, changed) (1, 1, 0)
  3298.       init io_toposort 1.59740447998e-05
  3299.       loop time 7.15255737305e-06
  3300.       callback_time 0.0
  3301.   0.000045s - ('cond_make_inplace', 'TopoOptimizer', 47, 0, 0) - 0.000s
  3302.     TopoOptimizer  cond_make_inplace
  3303.       nb_node (start, end, changed) (0, 0, 0)
  3304.       init io_toposort 1.00135803223e-05
  3305.       loop time 0.0
  3306.       callback_time 0.0
  3307.   0.000045s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 0, 0) - 0.000s
  3308.     TopoOptimizer  dimshuffle_as_view
  3309.       nb_node (start, end, changed) (0, 0, 0)
  3310.       init io_toposort 1.31130218506e-05
  3311.       loop time 9.53674316406e-07
  3312.       callback_time 0.0
  3313.   0.000043s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 0, 0) - 0.000s
  3314.   0.000042s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 0, 0) - 0.000s
  3315.   0.000039s - ('local_destructive', 'TopoOptimizer', 48, 0, 0) - 0.000s
  3316.     TopoOptimizer  CURAND_destructive
  3317.       nb_node (start, end, changed) (0, 0, 0)
  3318.       init io_toposort 9.05990600586e-06
  3319.       loop time 1.19209289551e-06
  3320.       callback_time 0.0
  3321.   0.000037s - ('random_make_inplace', 'TopoOptimizer', 49, 0, 0) - 0.000s
  3322.     TopoOptimizer  random_make_inplace
  3323.       nb_node (start, end, changed) (0, 0, 0)
  3324.       init io_toposort 7.86781311035e-06
  3325.       loop time 0.0
  3326.       callback_time 0.0
  3327.   0.000036s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 0, 0) - 0.000s
  3328.     TopoOptimizer  random_make_inplace_mrg
  3329.       nb_node (start, end, changed) (0, 0, 0)
  3330.       init io_toposort 8.10623168945e-06
  3331.       loop time 0.0
  3332.       callback_time 0.0
  3333.   0.000036s - ('merge1.2', 'MergeOptimizer', 7, 0, 0) - 0.000s
  3334.     MergeOptimizer
  3335.       nb fail=    0 merged=    0 constant=    0
  3336.       time replace=0.00 validate=0.00 callback=0.00
  3337.   0.000034s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 0, 0) - 0.000s
  3338.     TopoOptimizer  local_advincsub1_gpua_inplace
  3339.       nb_node (start, end, changed) (0, 0, 0)
  3340.       init io_toposort 8.10623168945e-06
  3341.       loop time 0.0
  3342.       callback_time 0.0
  3343.   0.000033s - ('make_ger_destructive', 'TopoOptimizer', 41, 0, 0) - 0.000s
  3344.     TopoOptimizer  make_scipy_blas_destructive
  3345.       nb_node (start, end, changed) (0, 0, 0)
  3346.       init io_toposort 6.91413879395e-06
  3347.       loop time 0.0
  3348.       callback_time 0.0
  3349.   0.000032s - ('c_blas_destructive', 'TopoOptimizer', 37, 0, 0) - 0.000s
  3350.     TopoOptimizer  c_blas_destructive
  3351.       nb_node (start, end, changed) (0, 0, 0)
  3352.       init io_toposort 6.91413879395e-06
  3353.       loop time 1.19209289551e-06
  3354.       callback_time 0.0
  3355. LocalOptGroup
  3356. ---------------------
  3357. --- The Optimizer wasn't successful ---
  3358.  
  3359.    0.000030s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 0, 0) - 0.000s
  3360.    0.000030s - ('AbstractConvCheck', 'TopoOptimizer', 18, 0, 0) - 0.000s
  3361.      TopoOptimizer  AbstractConvCheck
  3362.        nb_node (start, end, changed) (0, 0, 0)
  3363.        init io_toposort 8.10623168945e-06
  3364.        loop time 0.0
  3365.        callback_time 0.0
  3366.    0.000029s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 0, 0) - 0.000s
  3367.      TopoOptimizer  local_fill_to_alloc
  3368.        nb_node (start, end, changed) (0, 0, 0)
  3369.        init io_toposort 7.86781311035e-06
  3370.        loop time 9.53674316406e-07
  3371.        callback_time 0.0
  3372.    0.000026s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 0, 0) - 0.000s
  3373.      TopoOptimizer  local_elemwise_alloc
  3374.        nb_node (start, end, changed) (0, 0, 0)
  3375.        init io_toposort 6.19888305664e-06
  3376.        loop time 0.0
  3377.        callback_time 0.0
  3378.    0.000020s - ('merge1.1', 'MergeOptimizer', 4, 1, 1) - 0.000s
  3379.      MergeOptimizer
  3380.        nb fail=    0 merged=    0 constant=    0
  3381.        time replace=0.00 validate=0.00 callback=0.00
  3382.    0.000005s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 0, 0) - 0.000s
  3383.  
  3384. Here are tips to potentially make your code run faster
  3385.                  (if you think of new ones, suggest them on the mailing list).
  3386.                  Test them first, as they are not guaranteed to always provide a speedup.
  3387.   Sorry, no tip for today.
  3388. Function profiling
  3389. ==================
  3390.   Message: /Users/Ramana/projects/SBRNN/sb/utils.py:27
  3391.   Time in 1 calls to Function.__call__: 1.621246e-05s
  3392.   Time in Function.fn.__call__: 5.960464e-06s (36.765%)
  3393.   Time in thunks: 4.053116e-06s (25.000%)
  3394.   Total compile time: 2.882910e-02s
  3395.     Number of Apply nodes: 1
  3396.     Theano Optimizer time: 9.731054e-03s
  3397.        Theano validate time: 8.273125e-05s
  3398.     Theano Linker time (includes C, CUDA code generation/compiling): 6.179810e-04s
  3399.        Import time 0.000000e+00s
  3400.        Node make_thunk time 4.899502e-04s
  3401.  
  3402. Time in all call to theano.grad() 2.656322e+00s
  3403. Time since theano import 477.814s
  3404. Class
  3405. ---
  3406. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  3407.   100.0%   100.0%       0.000s       4.05e-06s     C        1       1   theano.compile.ops.DeepCopyOp
  3408.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  3409.  
  3410. Ops
  3411. ---
  3412. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  3413.   100.0%   100.0%       0.000s       4.05e-06s     C        1        1   DeepCopyOp
  3414.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  3415.  
  3416. Apply
  3417. ------
  3418. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  3419.   100.0%   100.0%       0.000s       4.05e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  3420.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  3421.  
  3422. Optimizer Profile
  3423. -----------------
  3424.  SeqOptimizer  OPT_FAST_RUN  time 0.009s for 1/0 nodes before/after optimization
  3425.    0.001s for callback
  3426.        0.000s for fgraph.validate()
  3427.    time      - (name, class, index, nodes before, nodes after) - validate time
  3428.    0.002802s - ('canonicalize', 'EquilibriumOptimizer', 6, 1, 0) - 0.000s
  3429.      EquilibriumOptimizer      canonicalize
  3430.        time 0.002s for 2 passes
  3431.        nb nodes (start, end,  max) 1 0 1
  3432.        time io_toposort 0.000s
  3433.        time in local optimizers 0.001s
  3434.        time in global optimizers 0.000s
  3435.        time in final optimizers 0.001s
  3436.        time in cleanup optimizers 0.000s
  3437.         0 - 0.002s 4 (0.001s in global opts, 0.000s io_toposort) - 1 nodes - ('topo_constant_folding', 1) ('local_upcast_elemwise_constant_inputs', 1) ('local_dimshuffle_lift', 1) ('MergeOptimizer', 1)
  3438.         1 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3439.        times - times applied - nb node created - name:
  3440.        0.001s - 1 - 3 - local_upcast_elemwise_constant_inputs
  3441.        0.001s - 1 - 0 - topo_constant_folding
  3442.        0.000s - 1 - 1 - local_dimshuffle_lift
  3443.        0.000s - 1 - 1 - MergeOptimizer
  3444.        0.000s - in 85 optimization that were not used (display only those with a runtime > 0)
  3445.          0.000s - local_fill_sink
  3446.          0.000s - local_useless_elemwise
  3447.          0.000s - local_func_inv
  3448.          0.000s - local_useless_elemwise_comparison
  3449.          0.000s - local_merge_switch_same_cond
  3450.          0.000s - local_track_shape_i
  3451.          0.000s - local_cast_cast
  3452.          0.000s - local_fill_cut
  3453.          0.000s - local_expm1
  3454.          0.000s - local_lift_transpose_through_dot
  3455.          0.000s - local_useless_switch
  3456.  
  3457.      Global, final and clean up optimizers
  3458.      Iter 0
  3459.              TopoOptimizer  topo_constant_folding
  3460.                nb_node (start, end, changed) (2, 0, 2)
  3461.                init io_toposort 3.31401824951e-05
  3462.                loop time 0.000442981719971
  3463.                callback_time 0.000193357467651
  3464.              MergeOptimizer
  3465.                nb fail=    0 merged=    1 constant=    0
  3466.                time replace=0.00 validate=0.00 callback=0.00
  3467.      Iter 1
  3468.              TopoOptimizer  topo_constant_folding
  3469.                nb_node (start, end, changed) (0, 0, 0)
  3470.                init io_toposort 7.86781311035e-06
  3471.                loop time 0.0
  3472.                callback_time 0.0
  3473.              MergeOptimizer
  3474.                nb fail=    0 merged=    0 constant=    0
  3475.                time replace=0.00 validate=0.00 callback=0.00
  3476.    0.000772s - ('BlasOpt', 'SeqOptimizer', 12, 0, 0) - 0.000s
  3477.      SeqOptimizer      BlasOpt  time 0.001s for 0/0 nodes before/after optimization
  3478.        0.000s for callback
  3479.            0.000s for fgraph.validate()
  3480.        0.000245s - ('gemm_optimizer', 'GemmOptimizer', 1, 0, 0) - 0.000s
  3481.          GemmOptimizer
  3482.           nb_iter 1
  3483.           nb_replacement 0
  3484.           nb_replacement_didn_t_remove 0
  3485.           nb_inconsistency_make 0
  3486.           nb_inconsistency_replace 0
  3487.           time_canonicalize 0
  3488.           time_factor_can 0
  3489.           time_factor_list 0
  3490.           time_toposort 1.50203704834e-05
  3491.           validate_time 0.0
  3492.           callback_time 0.0
  3493.        0.000178s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  3494.          EquilibriumOptimizer          local_gemm_to_gemv
  3495.            time 0.000s for 1 passes
  3496.            nb nodes (start, end,  max) 0 0 0
  3497.            time io_toposort 0.000s
  3498.            time in local optimizers 0.000s
  3499.            time in global optimizers 0.000s
  3500.            time in final optimizers 0.000s
  3501.            time in cleanup optimizers 0.000s
  3502.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3503.        0.000059s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 0, 0) - 0.000s
  3504.          TopoOptimizer  local_dot22_to_dot22scalar
  3505.            nb_node (start, end, changed) (0, 0, 0)
  3506.            init io_toposort 1.71661376953e-05
  3507.            loop time 9.53674316406e-07
  3508.            callback_time 0.0
  3509.        0.000056s - ('use_c_blas', 'TopoOptimizer', 4, 0, 0) - 0.000s
  3510.          TopoOptimizer  use_c_blas
  3511.            nb_node (start, end, changed) (0, 0, 0)
  3512.            init io_toposort 1.50203704834e-05
  3513.            loop time 9.53674316406e-07
  3514.            callback_time 0.0
  3515.  LocalOptGroup
  3516.  ---------------------
  3517.  --- The Optimizer wasn't successful ---
  3518.  
  3519.       0.000054s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 0, 0) - 0.000s
  3520.         TopoOptimizer  local_dot_to_dot22
  3521.           nb_node (start, end, changed) (0, 0, 0)
  3522.           init io_toposort 1.50203704834e-05
  3523.           loop time 1.19209289551e-06
  3524.           callback_time 0.0
  3525.       0.000048s - ('use_scipy_ger', 'TopoOptimizer', 5, 0, 0) - 0.000s
  3526.         TopoOptimizer  scipy_blas
  3527.           nb_node (start, end, changed) (0, 0, 0)
  3528.           init io_toposort 1.28746032715e-05
  3529.           loop time 1.19209289551e-06
  3530.           callback_time 0.0
  3531.  
  3532.   0.000682s - ('gpuarray_opt', 'SeqOptimizer', 16, 0, 0) - 0.000s
  3533.     SeqOptimizer      gpuarray_opt  time 0.001s for 0/0 nodes before/after optimization
  3534.       0.000s for callback
  3535.           0.000s for fgraph.validate()
  3536.       0.000387s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 0, 0) - 0.000s
  3537.         EquilibriumOptimizer          gpuarray_local_optimizations
  3538.           time 0.000s for 1 passes
  3539.           nb nodes (start, end,  max) 0 0 0
  3540.           time io_toposort 0.000s
  3541.           time in local optimizers 0.000s
  3542.           time in global optimizers 0.000s
  3543.           time in final optimizers 0.000s
  3544.           time in cleanup optimizers 0.000s
  3545.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3546.       0.000085s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  3547.         EquilibriumOptimizer          gpuarray_cut_transfers
  3548.           time 0.000s for 1 passes
  3549.           nb nodes (start, end,  max) 0 0 0
  3550.           time io_toposort 0.000s
  3551.           time in local optimizers 0.000s
  3552.           time in global optimizers 0.000s
  3553.           time in final optimizers 0.000s
  3554.           time in cleanup optimizers 0.000s
  3555.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3556.       0.000066s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 0, 0) - 0.000s
  3557.         GraphToGPUOptimizer          gpuarray_graph_optimization
  3558.           time io_toposort 0.000s
  3559.         Total time taken by local optimizers 0.000s
  3560.       0.000009s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 0, 0) - 0.000s
  3561.  
  3562.   0.000558s - ('add_destroy_handler', 'AddDestroyHandler', 23, 0, 0) - 0.000s
  3563.   0.000556s - ('elemwise_fusion', 'SeqOptimizer', 19, 0, 0) - 0.000s
  3564.     SeqOptimizer      elemwise_fusion  time 0.000s for 0/0 nodes before/after optimization
  3565.       0.000s for callback
  3566.           0.000s for fgraph.validate()
  3567.       0.000215s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 0, 0) - 0.000s
  3568.         FusionOptimizer
  3569.          nb_iter 1
  3570.          nb_replacement 0
  3571.          nb_inconsistency_replace 0
  3572.          validate_time 0.0
  3573.          callback_time 0.0
  3574.          time_toposort 2.14576721191e-06
  3575.       0.000206s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 0, 0) - 0.000s
  3576.         FusionOptimizer
  3577.          nb_iter 1
  3578.          nb_replacement 0
  3579.          nb_inconsistency_replace 0
  3580.          validate_time 0.0
  3581.          callback_time 0.0
  3582.          time_toposort 9.53674316406e-07
  3583.  
  3584.   0.000501s - ('specialize', 'EquilibriumOptimizer', 13, 0, 0) - 0.000s
  3585.     EquilibriumOptimizer      specialize
  3586.       time 0.000s for 1 passes
  3587.       nb nodes (start, end,  max) 0 0 0
  3588.       time io_toposort 0.000s
  3589.       time in local optimizers 0.000s
  3590.       time in global optimizers 0.000s
  3591.       time in final optimizers 0.000s
  3592.       time in cleanup optimizers 0.000s
  3593.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3594.     Global, final and clean up optimizers
  3595.     Iter 0
  3596.             TopoOptimizer  topo_constant_folding
  3597.               nb_node (start, end, changed) (0, 0, 0)
  3598.               init io_toposort 9.05990600586e-06
  3599.               loop time 0.0
  3600.               callback_time 0.0
  3601.   0.000425s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 0, 0) - 0.000s
  3602.     EquilibriumOptimizer      scan_eqopt2
  3603.       time 0.000s for 1 passes
  3604.       nb nodes (start, end,  max) 0 0 0
  3605.       time io_toposort 0.000s
  3606.       time in local optimizers 0.000s
  3607.       time in global optimizers 0.000s
  3608.       time in final optimizers 0.000s
  3609.       time in cleanup optimizers 0.000s
  3610.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3611.     Global, final and clean up optimizers
  3612.     Iter 0
  3613.             TopoOptimizer  constant_folding_for_scan2
  3614.               nb_node (start, end, changed) (0, 0, 0)
  3615.               init io_toposort 1.50203704834e-05
  3616.               loop time 0.0
  3617.               callback_time 0.0
  3618.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  3619.               nb_node (start, end, changed) (0, 0, 0)
  3620.               init io_toposort 7.86781311035e-06
  3621.               loop time 0.0
  3622.               callback_time 0.0
  3623.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  3624.               nb_node (start, end, changed) (0, 0, 0)
  3625.               init io_toposort 6.91413879395e-06
  3626.               loop time 0.0
  3627.               callback_time 0.0
  3628.             TopoOptimizer  scanOp_merge_inouts
  3629.               nb_node (start, end, changed) (0, 0, 0)
  3630.               init io_toposort 5.96046447754e-06
  3631.               loop time 0.0
  3632.               callback_time 0.0
  3633.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  3634.               nb_node (start, end, changed) (0, 0, 0)
  3635.               init io_toposort 5.96046447754e-06
  3636.               loop time 0.0
  3637.               callback_time 0.0
  3638.   0.000274s - ('stabilize', 'EquilibriumOptimizer', 8, 0, 0) - 0.000s
  3639.     EquilibriumOptimizer      stabilize
  3640.       time 0.000s for 1 passes
  3641.       nb nodes (start, end,  max) 0 0 0
  3642.       time io_toposort 0.000s
  3643.       time in local optimizers 0.000s
  3644.       time in global optimizers 0.000s
  3645.       time in final optimizers 0.000s
  3646.       time in cleanup optimizers 0.000s
  3647.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3648.     Global, final and clean up optimizers
  3649.     Iter 0
  3650.             TopoOptimizer  topo_constant_folding
  3651.               nb_node (start, end, changed) (0, 0, 0)
  3652.               init io_toposort 7.15255737305e-06
  3653.               loop time 0.0
  3654.               callback_time 0.0
  3655.   0.000224s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 1, 1) - 0.000s
  3656.     EquilibriumOptimizer      scan_eqopt1
  3657.       time 0.000s for 1 passes
  3658.       nb nodes (start, end,  max) 1 1 1
  3659.       time io_toposort 0.000s
  3660.       time in local optimizers 0.000s
  3661.       time in global optimizers 0.000s
  3662.       time in final optimizers 0.000s
  3663.       time in cleanup optimizers 0.000s
  3664.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  3665.     Global, final and clean up optimizers
  3666.     Iter 0
  3667.             SeqOptimizer              all_pushout_opt  time 0.000s for 1/1 nodes before/after optimization
  3668.               0.000s for callback
  3669.                   0.000s for fgraph.validate()
  3670.               0.000057s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 1, 1) - 0.000s
  3671.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  3672.                   nb_node (start, end, changed) (1, 1, 0)
  3673.                   init io_toposort 2.09808349609e-05
  3674.                   loop time 4.05311584473e-06
  3675.                   callback_time 0.0
  3676.               0.000013s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 1, 1) - 0.000s
  3677.               0.000009s - ('scan_pushout_dot1', 'PushOutDot1', 3, 1, 1) - 0.000s
  3678.               0.000009s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 1, 1) - 0.000s
  3679.               0.000008s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 1, 1) - 0.000s
  3680.  
  3681.   0.000196s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 0, 0) - 0.000s
  3682.     FusionOptimizer
  3683.      nb_iter 1
  3684.      nb_replacement 0
  3685.      nb_inconsistency_replace 0
  3686.      validate_time 0.0
  3687.      callback_time 0.0
  3688.      time_toposort 9.53674316406e-07
  3689.   0.000193s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 0, 0) - 0.000s
  3690.     FusionOptimizer
  3691.      nb_iter 1
  3692.      nb_replacement 0
  3693.      nb_inconsistency_replace 0
  3694.      validate_time 0.0
  3695.      callback_time 0.0
  3696.      time_toposort 1.19209289551e-06
  3697.   0.000189s - ('merge3', 'MergeOptimizer', 51, 0, 0) - 0.000s
  3698.     MergeOptimizer
  3699.       nb fail=    0 merged=    0 constant=    0
  3700.       time replace=0.00 validate=0.00 callback=0.00
  3701.   0.000153s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 0, 0) - 0.000s
  3702.     EquilibriumOptimizer      uncanonicalize
  3703.       time 0.000s for 1 passes
  3704.       nb nodes (start, end,  max) 0 0 0
  3705.       time io_toposort 0.000s
  3706.       time in local optimizers 0.000s
  3707.       time in global optimizers 0.000s
  3708.       time in final optimizers 0.000s
  3709.       time in cleanup optimizers 0.000s
  3710.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3711.     Global, final and clean up optimizers
  3712.     Iter 0
  3713.             TopoOptimizer  topo_constant_folding
  3714.               nb_node (start, end, changed) (0, 0, 0)
  3715.               init io_toposort 7.15255737305e-06
  3716.               loop time 9.53674316406e-07
  3717.               callback_time 0.0
  3718.   0.000129s - ('merge2', 'MergeOptimizer', 22, 0, 0) - 0.000s
  3719.     MergeOptimizer
  3720.       nb fail=    0 merged=    0 constant=    0
  3721.       time replace=0.00 validate=0.00 callback=0.00
  3722.   0.000096s - ('specialize_device', 'EquilibriumOptimizer', 17, 0, 0) - 0.000s
  3723.     EquilibriumOptimizer      specialize_device
  3724.       time 0.000s for 1 passes
  3725.       nb nodes (start, end,  max) 0 0 0
  3726.       time io_toposort 0.000s
  3727.       time in local optimizers 0.000s
  3728.       time in global optimizers 0.000s
  3729.       time in final optimizers 0.000s
  3730.       time in cleanup optimizers 0.000s
  3731.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  3732.   0.000093s - ('merge1', 'MergeOptimizer', 0, 1, 1) - 0.000s
  3733.     MergeOptimizer
  3734.       nb fail=    0 merged=    0 constant=    0
  3735.       time replace=0.00 validate=0.00 callback=0.00
  3736.   0.000085s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 0, 0) - 0.000s
  3737.     TopoOptimizer  InplaceGpuBlasOpt
  3738.       nb_node (start, end, changed) (0, 0, 0)
  3739.       init io_toposort 8.10623168945e-06
  3740.       loop time 0.0
  3741.       callback_time 0.0
  3742. LocalOptGroup
  3743. ---------------------
  3744. --- The Optimizer wasn't successful ---
  3745.  
  3746.    0.000082s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 0, 0) - 0.000s
  3747.      TopoOptimizer  InplaceGpuaBlasOpt
  3748.        nb_node (start, end, changed) (0, 0, 0)
  3749.        init io_toposort 8.10623168945e-06
  3750.        loop time 0.0
  3751.        callback_time 0.0
  3752.  LocalOptGroup
  3753.  ---------------------
  3754.    time taken - times applied - times tried - name - node_created:
  3755.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  3756.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  3757.  
  3758.    0.000079s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 0, 0) - 0.000s
  3759.      TopoOptimizer  local_dnn_conv_inplace
  3760.        nb_node (start, end, changed) (0, 0, 0)
  3761.        init io_toposort 7.15255737305e-06
  3762.        loop time 9.53674316406e-07
  3763.        callback_time 0.0
  3764.  LocalOptGroup
  3765.  ---------------------
  3766.  --- The Optimizer wasn't successful ---
  3767.  
  3768.   0.000079s - ('blas_opt_inplace', 'TopoOptimizer', 34, 0, 0) - 0.000s
  3769.     TopoOptimizer  InplaceBlasOpt
  3770.       nb_node (start, end, changed) (0, 0, 0)
  3771.       init io_toposort 6.91413879395e-06
  3772.       loop time 0.0
  3773.       callback_time 0.0
  3774. LocalOptGroup
  3775. ---------------------
  3776. --- The Optimizer wasn't successful ---
  3777.  
  3778.    0.000077s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 0, 0) - 0.000s
  3779.      TopoOptimizer  local_dnna_conv_inplace
  3780.        nb_node (start, end, changed) (0, 0, 0)
  3781.        init io_toposort 6.91413879395e-06
  3782.        loop time 0.0
  3783.        callback_time 0.0
  3784.  LocalOptGroup
  3785.  ---------------------
  3786.    time taken - times applied - times tried - name - node_created:
  3787.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  3788.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  3789.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  3790.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  3791.  
  3792.    0.000077s - ('useless', 'TopoOptimizer', 3, 1, 1) - 0.000s
  3793.      TopoOptimizer  useless
  3794.        nb_node (start, end, changed) (1, 1, 0)
  3795.        init io_toposort 1.78813934326e-05
  3796.        loop time 2.88486480713e-05
  3797.        callback_time 0.0
  3798.  LocalOptGroup
  3799.  ---------------------
  3800.    time taken - times applied - times tried - name - node_created:
  3801.    -0.000s - 0 - 1 - local_useless_switch - 0
  3802.    -0.000s - 0 - 1 - local_useless_elemwise_comparison - 0
  3803.    -0.000s - 0 - 1 - local_useless_elemwise - 0
  3804.    0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  3805.  
  3806.    0.000072s - ('ShapeOpt', 'ShapeOptimizer', 2, 1, 1) - 0.000s
  3807.    0.000059s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 0, 0) - 0.000s
  3808.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  3809.        nb_node (start, end, changed) (0, 0, 0)
  3810.        init io_toposort 9.05990600586e-06
  3811.        loop time 0.0
  3812.        callback_time 0.0
  3813.    0.000055s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 0, 0) - 0.000s
  3814.      TopoOptimizer  local_gemm16_inplace
  3815.        nb_node (start, end, changed) (0, 0, 0)
  3816.        init io_toposort 6.91413879395e-06
  3817.        loop time 9.53674316406e-07
  3818.        callback_time 0.0
  3819.    0.000054s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 0, 0) - 0.000s
  3820.      TopoOptimizer  local_inplace_sparseblockouter
  3821.        nb_node (start, end, changed) (0, 0, 0)
  3822.        init io_toposort 7.86781311035e-06
  3823.        loop time 0.0
  3824.        callback_time 0.0
  3825.    0.000051s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 0, 0) - 0.000s
  3826.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  3827.        nb_node (start, end, changed) (0, 0, 0)
  3828.        init io_toposort 9.05990600586e-06
  3829.        loop time 0.0
  3830.        callback_time 0.0
  3831.    0.000048s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 0, 0) - 0.000s
  3832.      TopoOptimizer  local_inplace_incsubtensor1
  3833.        nb_node (start, end, changed) (0, 0, 0)
  3834.        init io_toposort 6.91413879395e-06
  3835.        loop time 0.0
  3836.        callback_time 0.0
  3837.    0.000047s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 0, 0) - 0.000s
  3838.      TopoOptimizer  local_inplace_sparse_block_outer
  3839.        nb_node (start, end, changed) (0, 0, 0)
  3840.        init io_toposort 6.91413879395e-06
  3841.        loop time 1.19209289551e-06
  3842.        callback_time 0.0
  3843.    0.000047s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 0, 0) - 0.000s
  3844.      TopoOptimizer  local_inplace_sparse_block_gemv
  3845.        nb_node (start, end, changed) (0, 0, 0)
  3846.        init io_toposort 6.91413879395e-06
  3847.        loop time 0.0
  3848.        callback_time 0.0
  3849.    0.000046s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 0, 0) - 0.000s
  3850.      TopoOptimizer  local_inplace_setsubtensor
  3851.        nb_node (start, end, changed) (0, 0, 0)
  3852.        init io_toposort 7.15255737305e-06
  3853.        loop time 0.0
  3854.        callback_time 0.0
  3855.    0.000046s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 1, 1) - 0.000s
  3856.      TopoOptimizer  pre_local_IncSubtensor_serialize
  3857.        nb_node (start, end, changed) (1, 1, 0)
  3858.        init io_toposort 1.81198120117e-05
  3859.        loop time 7.15255737305e-06
  3860.        callback_time 0.0
  3861.    0.000046s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 0, 0) - 0.000s
  3862.      TopoOptimizer  local_inplace_sparseblockgemv
  3863.        nb_node (start, end, changed) (0, 0, 0)
  3864.        init io_toposort 7.15255737305e-06
  3865.        loop time 9.53674316406e-07
  3866.        callback_time 0.0
  3867.    0.000041s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 0, 0) - 0.000s
  3868.      TopoOptimizer  dimshuffle_as_view
  3869.        nb_node (start, end, changed) (0, 0, 0)
  3870.        init io_toposort 1.19209289551e-05
  3871.        loop time 0.0
  3872.        callback_time 0.0
  3873.    0.000040s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 0, 0) - 0.000s
  3874.    0.000038s - ('merge1.2', 'MergeOptimizer', 7, 0, 0) - 0.000s
  3875.      MergeOptimizer
  3876.        nb fail=    0 merged=    0 constant=    0
  3877.        time replace=0.00 validate=0.00 callback=0.00
  3878.    0.000034s - ('AbstractConvCheck', 'TopoOptimizer', 18, 0, 0) - 0.000s
  3879.      TopoOptimizer  AbstractConvCheck
  3880.        nb_node (start, end, changed) (0, 0, 0)
  3881.        init io_toposort 8.10623168945e-06
  3882.        loop time 0.0
  3883.        callback_time 0.0
  3884.    0.000034s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 0, 0) - 0.000s
  3885.    0.000033s - ('cond_make_inplace', 'TopoOptimizer', 47, 0, 0) - 0.000s
  3886.      TopoOptimizer  cond_make_inplace
  3887.        nb_node (start, end, changed) (0, 0, 0)
  3888.        init io_toposort 8.10623168945e-06
  3889.        loop time 0.0
  3890.        callback_time 0.0
  3891.    0.000032s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 0, 0) - 0.000s
  3892.      TopoOptimizer  local_advincsub1_gpua_inplace
  3893.        nb_node (start, end, changed) (0, 0, 0)
  3894.        init io_toposort 8.10623168945e-06
  3895.        loop time 0.0
  3896.        callback_time 0.0
  3897.    0.000031s - ('c_blas_destructive', 'TopoOptimizer', 37, 0, 0) - 0.000s
  3898.      TopoOptimizer  c_blas_destructive
  3899.        nb_node (start, end, changed) (0, 0, 0)
  3900.        init io_toposort 6.91413879395e-06
  3901.        loop time 0.0
  3902.        callback_time 0.0
  3903.  LocalOptGroup
  3904.  ---------------------
  3905.  --- The Optimizer wasn't successful ---
  3906.  
  3907.   0.000031s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 0, 0) - 0.000s
  3908.     TopoOptimizer  local_fill_to_alloc
  3909.       nb_node (start, end, changed) (0, 0, 0)
  3910.       init io_toposort 8.10623168945e-06
  3911.       loop time 0.0
  3912.       callback_time 0.0
  3913.   0.000030s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 0, 0) - 0.000s
  3914.   0.000030s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 0, 0) - 0.000s
  3915.   0.000029s - ('random_make_inplace', 'TopoOptimizer', 49, 0, 0) - 0.000s
  3916.     TopoOptimizer  random_make_inplace
  3917.       nb_node (start, end, changed) (0, 0, 0)
  3918.       init io_toposort 5.96046447754e-06
  3919.       loop time 0.0
  3920.       callback_time 0.0
  3921.   0.000029s - ('local_destructive', 'TopoOptimizer', 48, 0, 0) - 0.000s
  3922.     TopoOptimizer  CURAND_destructive
  3923.       nb_node (start, end, changed) (0, 0, 0)
  3924.       init io_toposort 6.91413879395e-06
  3925.       loop time 9.53674316406e-07
  3926.       callback_time 0.0
  3927.   0.000029s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 0, 0) - 0.000s
  3928.     TopoOptimizer  local_elemwise_alloc
  3929.       nb_node (start, end, changed) (0, 0, 0)
  3930.       init io_toposort 7.15255737305e-06
  3931.       loop time 9.53674316406e-07
  3932.       callback_time 0.0
  3933.   0.000028s - ('make_ger_destructive', 'TopoOptimizer', 41, 0, 0) - 0.000s
  3934.     TopoOptimizer  make_scipy_blas_destructive
  3935.       nb_node (start, end, changed) (0, 0, 0)
  3936.       init io_toposort 7.15255737305e-06
  3937.       loop time 9.53674316406e-07
  3938.       callback_time 0.0
  3939.   0.000028s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 0, 0) - 0.000s
  3940.     TopoOptimizer  random_make_inplace_mrg
  3941.       nb_node (start, end, changed) (0, 0, 0)
  3942.       init io_toposort 6.19888305664e-06
  3943.       loop time 1.19209289551e-06
  3944.       callback_time 0.0
  3945.   0.000025s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 0, 0) - 0.000s
  3946.   0.000021s - ('merge1.1', 'MergeOptimizer', 4, 1, 1) - 0.000s
  3947.     MergeOptimizer
  3948.       nb fail=    0 merged=    0 constant=    0
  3949.       time replace=0.00 validate=0.00 callback=0.00
  3950.   0.000004s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 0, 0) - 0.000s
  3951.  
  3952. Here are tips to potentially make your code run faster
  3953.                 (if you think of new ones, suggest them on the mailing list).
  3954.                 Test them first, as they are not guaranteed to always provide a speedup.
  3955.  Sorry, no tip for today.
  3956. Function profiling
  3957. ==================
  3958.  Message: /Users/Ramana/projects/SBRNN/sb/utils.py:56
  3959.  Time in 1 calls to Function.__call__: 1.502037e-05s
  3960.  Time in Function.fn.__call__: 5.960464e-06s (39.683%)
  3961.  Time in thunks: 3.099442e-06s (20.635%)
  3962.  Total compile time: 2.807999e-02s
  3963.    Number of Apply nodes: 1
  3964.    Theano Optimizer time: 9.526968e-03s
  3965.       Theano validate time: 8.201599e-05s
  3966.    Theano Linker time (includes C, CUDA code generation/compiling): 5.979538e-04s
  3967.       Import time 0.000000e+00s
  3968.       Node make_thunk time 4.749298e-04s
  3969.  
  3970. Time in all call to theano.grad() 2.656322e+00s
  3971. Time since theano import 477.820s
  3972. Class
  3973. ---
  3974. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  3975.  100.0%   100.0%       0.000s       3.10e-06s     C        1       1   theano.compile.ops.DeepCopyOp
  3976.   ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  3977.  
  3978. Ops
  3979. ---
  3980. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  3981.  100.0%   100.0%       0.000s       3.10e-06s     C        1        1   DeepCopyOp
  3982.   ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  3983.  
  3984. Apply
  3985. ------
  3986. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  3987.  100.0%   100.0%       0.000s       3.10e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  3988.   ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  3989.  
  3990. Optimizer Profile
  3991. -----------------
  3992. SeqOptimizer  OPT_FAST_RUN  time 0.009s for 1/0 nodes before/after optimization
  3993.   0.001s for callback
  3994.       0.000s for fgraph.validate()
  3995.   time      - (name, class, index, nodes before, nodes after) - validate time
  3996.   0.002808s - ('canonicalize', 'EquilibriumOptimizer', 6, 1, 0) - 0.000s
  3997.     EquilibriumOptimizer      canonicalize
  3998.       time 0.002s for 2 passes
  3999.       nb nodes (start, end,  max) 1 0 1
  4000.       time io_toposort 0.000s
  4001.       time in local optimizers 0.001s
  4002.       time in global optimizers 0.000s
  4003.       time in final optimizers 0.001s
  4004.       time in cleanup optimizers 0.000s
  4005.        0 - 0.002s 4 (0.001s in global opts, 0.000s io_toposort) - 1 nodes - ('topo_constant_folding', 1) ('local_upcast_elemwise_constant_inputs', 1) ('local_dimshuffle_lift', 1) ('MergeOptimizer', 1)
  4006.        1 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4007.       times - times applied - nb node created - name:
  4008.       0.001s - 1 - 3 - local_upcast_elemwise_constant_inputs
  4009.       0.001s - 1 - 0 - topo_constant_folding
  4010.       0.000s - 1 - 1 - local_dimshuffle_lift
  4011.       0.000s - 1 - 1 - MergeOptimizer
  4012.       0.000s - in 85 optimization that were not used (display only those with a runtime > 0)
  4013.         0.000s - local_fill_sink
  4014.         0.000s - local_useless_elemwise
  4015.         0.000s - local_func_inv
  4016.         0.000s - local_track_shape_i
  4017.         0.000s - local_merge_switch_same_cond
  4018.         0.000s - local_fill_cut
  4019.         0.000s - local_cast_cast
  4020.         0.000s - local_useless_switch
  4021.         0.000s - local_expm1
  4022.         0.000s - local_useless_elemwise_comparison
  4023.         0.000s - local_lift_transpose_through_dot
  4024.  
  4025.     Global, final and clean up optimizers
  4026.     Iter 0
  4027.             TopoOptimizer  topo_constant_folding
  4028.               nb_node (start, end, changed) (2, 0, 2)
  4029.               init io_toposort 3.09944152832e-05
  4030.               loop time 0.000473022460938
  4031.               callback_time 0.000208616256714
  4032.             MergeOptimizer
  4033.               nb fail=    0 merged=    1 constant=    0
  4034.               time replace=0.00 validate=0.00 callback=0.00
  4035.     Iter 1
  4036.             TopoOptimizer  topo_constant_folding
  4037.               nb_node (start, end, changed) (0, 0, 0)
  4038.               init io_toposort 1.00135803223e-05
  4039.               loop time 0.0
  4040.               callback_time 0.0
  4041.             MergeOptimizer
  4042.               nb fail=    0 merged=    0 constant=    0
  4043.               time replace=0.00 validate=0.00 callback=0.00
  4044.   0.000602s - ('add_destroy_handler', 'AddDestroyHandler', 23, 0, 0) - 0.000s
  4045.   0.000592s - ('gpuarray_opt', 'SeqOptimizer', 16, 0, 0) - 0.000s
  4046.     SeqOptimizer      gpuarray_opt  time 0.000s for 0/0 nodes before/after optimization
  4047.       0.000s for callback
  4048.           0.000s for fgraph.validate()
  4049.       0.000336s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 0, 0) - 0.000s
  4050.         EquilibriumOptimizer          gpuarray_local_optimizations
  4051.           time 0.000s for 1 passes
  4052.           nb nodes (start, end,  max) 0 0 0
  4053.           time io_toposort 0.000s
  4054.           time in local optimizers 0.000s
  4055.           time in global optimizers 0.000s
  4056.           time in final optimizers 0.000s
  4057.           time in cleanup optimizers 0.000s
  4058.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4059.       0.000070s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  4060.         EquilibriumOptimizer          gpuarray_cut_transfers
  4061.           time 0.000s for 1 passes
  4062.           nb nodes (start, end,  max) 0 0 0
  4063.           time io_toposort 0.000s
  4064.           time in local optimizers 0.000s
  4065.           time in global optimizers 0.000s
  4066.           time in final optimizers 0.000s
  4067.           time in cleanup optimizers 0.000s
  4068.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4069.       0.000058s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 0, 0) - 0.000s
  4070.         GraphToGPUOptimizer          gpuarray_graph_optimization
  4071.           time io_toposort 0.000s
  4072.         Total time taken by local optimizers 0.000s
  4073.       0.000007s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 0, 0) - 0.000s
  4074.  
  4075.   0.000566s - ('elemwise_fusion', 'SeqOptimizer', 19, 0, 0) - 0.000s
  4076.     SeqOptimizer      elemwise_fusion  time 0.000s for 0/0 nodes before/after optimization
  4077.       0.000s for callback
  4078.           0.000s for fgraph.validate()
  4079.       0.000195s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 0, 0) - 0.000s
  4080.         FusionOptimizer
  4081.          nb_iter 1
  4082.          nb_replacement 0
  4083.          nb_inconsistency_replace 0
  4084.          validate_time 0.0
  4085.          callback_time 0.0
  4086.          time_toposort 9.53674316406e-07
  4087.       0.000186s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 0, 0) - 0.000s
  4088.         FusionOptimizer
  4089.          nb_iter 1
  4090.          nb_replacement 0
  4091.          nb_inconsistency_replace 0
  4092.          validate_time 0.0
  4093.          callback_time 0.0
  4094.          time_toposort 1.19209289551e-06
  4095.  
  4096.   0.000460s - ('BlasOpt', 'SeqOptimizer', 12, 0, 0) - 0.000s
  4097.     SeqOptimizer      BlasOpt  time 0.000s for 0/0 nodes before/after optimization
  4098.       0.000s for callback
  4099.           0.000s for fgraph.validate()
  4100.       0.000144s - ('gemm_optimizer', 'GemmOptimizer', 1, 0, 0) - 0.000s
  4101.         GemmOptimizer
  4102.          nb_iter 1
  4103.          nb_replacement 0
  4104.          nb_replacement_didn_t_remove 0
  4105.          nb_inconsistency_make 0
  4106.          nb_inconsistency_replace 0
  4107.          time_canonicalize 0
  4108.          time_factor_can 0
  4109.          time_factor_list 0
  4110.          time_toposort 6.91413879395e-06
  4111.          validate_time 0.0
  4112.          callback_time 0.0
  4113.       0.000097s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  4114.         EquilibriumOptimizer          local_gemm_to_gemv
  4115.           time 0.000s for 1 passes
  4116.           nb nodes (start, end,  max) 0 0 0
  4117.           time io_toposort 0.000s
  4118.           time in local optimizers 0.000s
  4119.           time in global optimizers 0.000s
  4120.           time in final optimizers 0.000s
  4121.           time in cleanup optimizers 0.000s
  4122.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4123.       0.000031s - ('use_c_blas', 'TopoOptimizer', 4, 0, 0) - 0.000s
  4124.         TopoOptimizer  use_c_blas
  4125.           nb_node (start, end, changed) (0, 0, 0)
  4126.           init io_toposort 7.86781311035e-06
  4127.           loop time 9.53674316406e-07
  4128.           callback_time 0.0
  4129. LocalOptGroup
  4130. ---------------------
  4131. --- The Optimizer wasn't successful ---
  4132.  
  4133.        0.000030s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 0, 0) - 0.000s
  4134.          TopoOptimizer  local_dot_to_dot22
  4135.            nb_node (start, end, changed) (0, 0, 0)
  4136.            init io_toposort 8.10623168945e-06
  4137.            loop time 0.0
  4138.            callback_time 0.0
  4139.        0.000029s - ('use_scipy_ger', 'TopoOptimizer', 5, 0, 0) - 0.000s
  4140.          TopoOptimizer  scipy_blas
  4141.            nb_node (start, end, changed) (0, 0, 0)
  4142.            init io_toposort 6.91413879395e-06
  4143.            loop time 0.0
  4144.            callback_time 0.0
  4145.        0.000029s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 0, 0) - 0.000s
  4146.          TopoOptimizer  local_dot22_to_dot22scalar
  4147.            nb_node (start, end, changed) (0, 0, 0)
  4148.            init io_toposort 6.91413879395e-06
  4149.            loop time 0.0
  4150.            callback_time 0.0
  4151.  
  4152.    0.000395s - ('specialize', 'EquilibriumOptimizer', 13, 0, 0) - 0.000s
  4153.      EquilibriumOptimizer      specialize
  4154.        time 0.000s for 1 passes
  4155.        nb nodes (start, end,  max) 0 0 0
  4156.        time io_toposort 0.000s
  4157.        time in local optimizers 0.000s
  4158.        time in global optimizers 0.000s
  4159.        time in final optimizers 0.000s
  4160.        time in cleanup optimizers 0.000s
  4161.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4162.      Global, final and clean up optimizers
  4163.      Iter 0
  4164.              TopoOptimizer  topo_constant_folding
  4165.                nb_node (start, end, changed) (0, 0, 0)
  4166.                init io_toposort 6.91413879395e-06
  4167.                loop time 0.0
  4168.                callback_time 0.0
  4169.    0.000281s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 0, 0) - 0.000s
  4170.      TopoOptimizer  local_dnna_conv_inplace
  4171.        nb_node (start, end, changed) (0, 0, 0)
  4172.        init io_toposort 1.09672546387e-05
  4173.        loop time 0.0
  4174.        callback_time 0.0
  4175.  LocalOptGroup
  4176.  ---------------------
  4177.    time taken - times applied - times tried - name - node_created:
  4178.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  4179.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  4180.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  4181.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  4182.  
  4183.    0.000276s - ('stabilize', 'EquilibriumOptimizer', 8, 0, 0) - 0.000s
  4184.      EquilibriumOptimizer      stabilize
  4185.        time 0.000s for 1 passes
  4186.        nb nodes (start, end,  max) 0 0 0
  4187.        time io_toposort 0.000s
  4188.        time in local optimizers 0.000s
  4189.        time in global optimizers 0.000s
  4190.        time in final optimizers 0.000s
  4191.        time in cleanup optimizers 0.000s
  4192.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4193.      Global, final and clean up optimizers
  4194.      Iter 0
  4195.              TopoOptimizer  topo_constant_folding
  4196.                nb_node (start, end, changed) (0, 0, 0)
  4197.                init io_toposort 6.91413879395e-06
  4198.                loop time 0.0
  4199.                callback_time 0.0
  4200.    0.000272s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 1, 1) - 0.000s
  4201.      EquilibriumOptimizer      scan_eqopt1
  4202.        time 0.000s for 1 passes
  4203.        nb nodes (start, end,  max) 1 1 1
  4204.        time io_toposort 0.000s
  4205.        time in local optimizers 0.000s
  4206.        time in global optimizers 0.000s
  4207.        time in final optimizers 0.000s
  4208.        time in cleanup optimizers 0.000s
  4209.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  4210.      Global, final and clean up optimizers
  4211.      Iter 0
  4212.              SeqOptimizer              all_pushout_opt  time 0.000s for 1/1 nodes before/after optimization
  4213.                0.000s for callback
  4214.                    0.000s for fgraph.validate()
  4215.                0.000073s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 1, 1) - 0.000s
  4216.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  4217.                    nb_node (start, end, changed) (1, 1, 0)
  4218.                    init io_toposort 2.88486480713e-05
  4219.                    loop time 5.00679016113e-06
  4220.                    callback_time 0.0
  4221.                0.000018s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 1, 1) - 0.000s
  4222.                0.000011s - ('scan_pushout_dot1', 'PushOutDot1', 3, 1, 1) - 0.000s
  4223.                0.000011s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 1, 1) - 0.000s
  4224.                0.000010s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 1, 1) - 0.000s
  4225.  
  4226.    0.000261s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 0, 0) - 0.000s
  4227.      EquilibriumOptimizer      scan_eqopt2
  4228.        time 0.000s for 1 passes
  4229.        nb nodes (start, end,  max) 0 0 0
  4230.        time io_toposort 0.000s
  4231.        time in local optimizers 0.000s
  4232.        time in global optimizers 0.000s
  4233.        time in final optimizers 0.000s
  4234.        time in cleanup optimizers 0.000s
  4235.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4236.      Global, final and clean up optimizers
  4237.      Iter 0
  4238.              TopoOptimizer  constant_folding_for_scan2
  4239.                nb_node (start, end, changed) (0, 0, 0)
  4240.                init io_toposort 7.86781311035e-06
  4241.                loop time 9.53674316406e-07
  4242.                callback_time 0.0
  4243.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  4244.                nb_node (start, end, changed) (0, 0, 0)
  4245.                init io_toposort 7.15255737305e-06
  4246.                loop time 0.0
  4247.                callback_time 0.0
  4248.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  4249.                nb_node (start, end, changed) (0, 0, 0)
  4250.                init io_toposort 6.91413879395e-06
  4251.                loop time 0.0
  4252.                callback_time 0.0
  4253.              TopoOptimizer  scanOp_merge_inouts
  4254.                nb_node (start, end, changed) (0, 0, 0)
  4255.                init io_toposort 9.05990600586e-06
  4256.                loop time 0.0
  4257.                callback_time 0.0
  4258.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  4259.                nb_node (start, end, changed) (0, 0, 0)
  4260.                init io_toposort 6.19888305664e-06
  4261.                loop time 0.0
  4262.                callback_time 0.0
  4263.    0.000234s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 0, 0) - 0.000s
  4264.      FusionOptimizer
  4265.       nb_iter 1
  4266.       nb_replacement 0
  4267.       nb_inconsistency_replace 0
  4268.       validate_time 0.0
  4269.       callback_time 0.0
  4270.       time_toposort 1.90734863281e-06
  4271.    0.000221s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 0, 0) - 0.000s
  4272.      FusionOptimizer
  4273.       nb_iter 1
  4274.       nb_replacement 0
  4275.       nb_inconsistency_replace 0
  4276.       validate_time 0.0
  4277.       callback_time 0.0
  4278.       time_toposort 1.90734863281e-06
  4279.    0.000188s - ('merge2', 'MergeOptimizer', 22, 0, 0) - 0.000s
  4280.      MergeOptimizer
  4281.        nb fail=    0 merged=    0 constant=    0
  4282.        time replace=0.00 validate=0.00 callback=0.00
  4283.    0.000181s - ('merge3', 'MergeOptimizer', 51, 0, 0) - 0.000s
  4284.      MergeOptimizer
  4285.        nb fail=    0 merged=    0 constant=    0
  4286.        time replace=0.00 validate=0.00 callback=0.00
  4287.    0.000137s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 0, 0) - 0.000s
  4288.      EquilibriumOptimizer      uncanonicalize
  4289.        time 0.000s for 1 passes
  4290.        nb nodes (start, end,  max) 0 0 0
  4291.        time io_toposort 0.000s
  4292.        time in local optimizers 0.000s
  4293.        time in global optimizers 0.000s
  4294.        time in final optimizers 0.000s
  4295.        time in cleanup optimizers 0.000s
  4296.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4297.      Global, final and clean up optimizers
  4298.      Iter 0
  4299.              TopoOptimizer  topo_constant_folding
  4300.                nb_node (start, end, changed) (0, 0, 0)
  4301.                init io_toposort 8.10623168945e-06
  4302.                loop time 0.0
  4303.                callback_time 0.0
  4304.    0.000117s - ('merge1', 'MergeOptimizer', 0, 1, 1) - 0.000s
  4305.      MergeOptimizer
  4306.        nb fail=    0 merged=    0 constant=    0
  4307.        time replace=0.00 validate=0.00 callback=0.00
  4308.    0.000091s - ('blas_opt_inplace', 'TopoOptimizer', 34, 0, 0) - 0.000s
  4309.      TopoOptimizer  InplaceBlasOpt
  4310.        nb_node (start, end, changed) (0, 0, 0)
  4311.        init io_toposort 7.15255737305e-06
  4312.        loop time 0.0
  4313.        callback_time 0.0
  4314.  LocalOptGroup
  4315.  ---------------------
  4316.  --- The Optimizer wasn't successful ---
  4317.  
  4318.   0.000090s - ('useless', 'TopoOptimizer', 3, 1, 1) - 0.000s
  4319.     TopoOptimizer  useless
  4320.       nb_node (start, end, changed) (1, 1, 0)
  4321.       init io_toposort 2.21729278564e-05
  4322.       loop time 3.31401824951e-05
  4323.       callback_time 0.0
  4324. LocalOptGroup
  4325. ---------------------
  4326.   time taken - times applied - times tried - name - node_created:
  4327.   -0.000s - 0 - 1 - local_useless_elemwise_comparison - 0
  4328.   -0.000s - 0 - 1 - local_useless_switch - 0
  4329.   -0.000s - 0 - 1 - local_useless_elemwise - 0
  4330.   0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  4331.  
  4332.   0.000090s - ('ShapeOpt', 'ShapeOptimizer', 2, 1, 1) - 0.000s
  4333.   0.000086s - ('specialize_device', 'EquilibriumOptimizer', 17, 0, 0) - 0.000s
  4334.     EquilibriumOptimizer      specialize_device
  4335.       time 0.000s for 1 passes
  4336.       nb nodes (start, end,  max) 0 0 0
  4337.       time io_toposort 0.000s
  4338.       time in local optimizers 0.000s
  4339.       time in global optimizers 0.000s
  4340.       time in final optimizers 0.000s
  4341.       time in cleanup optimizers 0.000s
  4342.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4343.   0.000076s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 0, 0) - 0.000s
  4344.     TopoOptimizer  InplaceGpuaBlasOpt
  4345.       nb_node (start, end, changed) (0, 0, 0)
  4346.       init io_toposort 8.10623168945e-06
  4347.       loop time 0.0
  4348.       callback_time 0.0
  4349. LocalOptGroup
  4350. ---------------------
  4351.   time taken - times applied - times tried - name - node_created:
  4352.   -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  4353.   0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  4354.  
  4355.   0.000075s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 0, 0) - 0.000s
  4356.     TopoOptimizer  InplaceGpuBlasOpt
  4357.       nb_node (start, end, changed) (0, 0, 0)
  4358.       init io_toposort 6.91413879395e-06
  4359.       loop time 0.0
  4360.       callback_time 0.0
  4361. LocalOptGroup
  4362. ---------------------
  4363. --- The Optimizer wasn't successful ---
  4364.  
  4365.    0.000075s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 0, 0) - 0.000s
  4366.      TopoOptimizer  local_dnn_conv_inplace
  4367.        nb_node (start, end, changed) (0, 0, 0)
  4368.        init io_toposort 5.96046447754e-06
  4369.        loop time 0.0
  4370.        callback_time 0.0
  4371.  LocalOptGroup
  4372.  ---------------------
  4373.  --- The Optimizer wasn't successful ---
  4374.  
  4375.   0.000059s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 0, 0) - 0.000s
  4376.     TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  4377.       nb_node (start, end, changed) (0, 0, 0)
  4378.       init io_toposort 9.05990600586e-06
  4379.       loop time 0.0
  4380.       callback_time 0.0
  4381.   0.000053s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 1, 1) - 0.000s
  4382.     TopoOptimizer  pre_local_IncSubtensor_serialize
  4383.       nb_node (start, end, changed) (1, 1, 0)
  4384.       init io_toposort 2.00271606445e-05
  4385.       loop time 8.10623168945e-06
  4386.       callback_time 0.0
  4387.   0.000053s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 0, 0) - 0.000s
  4388.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  4389.       nb_node (start, end, changed) (0, 0, 0)
  4390.       init io_toposort 9.05990600586e-06
  4391.       loop time 0.0
  4392.       callback_time 0.0
  4393.   0.000051s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 0, 0) - 0.000s
  4394.     TopoOptimizer  local_inplace_incsubtensor1
  4395.       nb_node (start, end, changed) (0, 0, 0)
  4396.       init io_toposort 8.10623168945e-06
  4397.       loop time 0.0
  4398.       callback_time 0.0
  4399.   0.000051s - ('merge1.2', 'MergeOptimizer', 7, 0, 0) - 0.000s
  4400.     MergeOptimizer
  4401.       nb fail=    0 merged=    0 constant=    0
  4402.       time replace=0.00 validate=0.00 callback=0.00
  4403.   0.000050s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 0, 0) - 0.000s
  4404.     TopoOptimizer  local_inplace_sparse_block_gemv
  4405.       nb_node (start, end, changed) (0, 0, 0)
  4406.       init io_toposort 8.10623168945e-06
  4407.       loop time 0.0
  4408.       callback_time 0.0
  4409.   0.000050s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 0, 0) - 0.000s
  4410.     TopoOptimizer  local_inplace_setsubtensor
  4411.       nb_node (start, end, changed) (0, 0, 0)
  4412.       init io_toposort 7.15255737305e-06
  4413.       loop time 0.0
  4414.       callback_time 0.0
  4415.   0.000049s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 0, 0) - 0.000s
  4416.     TopoOptimizer  local_gemm16_inplace
  4417.       nb_node (start, end, changed) (0, 0, 0)
  4418.       init io_toposort 8.10623168945e-06
  4419.       loop time 9.53674316406e-07
  4420.       callback_time 0.0
  4421.   0.000049s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 0, 0) - 0.000s
  4422.     TopoOptimizer  local_inplace_sparseblockouter
  4423.       nb_node (start, end, changed) (0, 0, 0)
  4424.       init io_toposort 7.15255737305e-06
  4425.       loop time 9.53674316406e-07
  4426.       callback_time 0.0
  4427.   0.000049s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 0, 0) - 0.000s
  4428.     TopoOptimizer  local_inplace_sparse_block_outer
  4429.       nb_node (start, end, changed) (0, 0, 0)
  4430.       init io_toposort 6.91413879395e-06
  4431.       loop time 0.0
  4432.       callback_time 0.0
  4433.   0.000049s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 0, 0) - 0.000s
  4434.     TopoOptimizer  local_inplace_sparseblockgemv
  4435.       nb_node (start, end, changed) (0, 0, 0)
  4436.       init io_toposort 7.86781311035e-06
  4437.       loop time 9.53674316406e-07
  4438.       callback_time 0.0
  4439.   0.000048s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 0, 0) - 0.000s
  4440.     TopoOptimizer  dimshuffle_as_view
  4441.       nb_node (start, end, changed) (0, 0, 0)
  4442.       init io_toposort 1.38282775879e-05
  4443.       loop time 9.53674316406e-07
  4444.       callback_time 0.0
  4445.   0.000041s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 0, 0) - 0.000s
  4446.   0.000038s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 0, 0) - 0.000s
  4447.     TopoOptimizer  local_advincsub1_gpua_inplace
  4448.       nb_node (start, end, changed) (0, 0, 0)
  4449.       init io_toposort 8.82148742676e-06
  4450.       loop time 0.0
  4451.       callback_time 0.0
  4452.   0.000035s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 0, 0) - 0.000s
  4453.     TopoOptimizer  local_fill_to_alloc
  4454.       nb_node (start, end, changed) (0, 0, 0)
  4455.       init io_toposort 7.86781311035e-06
  4456.       loop time 0.0
  4457.       callback_time 0.0
  4458.   0.000031s - ('cond_make_inplace', 'TopoOptimizer', 47, 0, 0) - 0.000s
  4459.     TopoOptimizer  cond_make_inplace
  4460.       nb_node (start, end, changed) (0, 0, 0)
  4461.       init io_toposort 6.91413879395e-06
  4462.       loop time 0.0
  4463.       callback_time 0.0
  4464.   0.000031s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 0, 0) - 0.000s
  4465.   0.000030s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 0, 0) - 0.000s
  4466.   0.000029s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 0, 0) - 0.000s
  4467.   0.000029s - ('make_ger_destructive', 'TopoOptimizer', 41, 0, 0) - 0.000s
  4468.     TopoOptimizer  make_scipy_blas_destructive
  4469.       nb_node (start, end, changed) (0, 0, 0)
  4470.       init io_toposort 6.91413879395e-06
  4471.       loop time 0.0
  4472.       callback_time 0.0
  4473.   0.000029s - ('c_blas_destructive', 'TopoOptimizer', 37, 0, 0) - 0.000s
  4474.     TopoOptimizer  c_blas_destructive
  4475.       nb_node (start, end, changed) (0, 0, 0)
  4476.       init io_toposort 5.96046447754e-06
  4477.       loop time 0.0
  4478.       callback_time 0.0
  4479. LocalOptGroup
  4480. ---------------------
  4481. --- The Optimizer wasn't successful ---
  4482.  
  4483.    0.000029s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 0, 0) - 0.000s
  4484.      TopoOptimizer  local_elemwise_alloc
  4485.        nb_node (start, end, changed) (0, 0, 0)
  4486.        init io_toposort 6.91413879395e-06
  4487.        loop time 0.0
  4488.        callback_time 0.0
  4489.    0.000029s - ('random_make_inplace', 'TopoOptimizer', 49, 0, 0) - 0.000s
  4490.      TopoOptimizer  random_make_inplace
  4491.        nb_node (start, end, changed) (0, 0, 0)
  4492.        init io_toposort 5.96046447754e-06
  4493.        loop time 0.0
  4494.        callback_time 0.0
  4495.    0.000029s - ('AbstractConvCheck', 'TopoOptimizer', 18, 0, 0) - 0.000s
  4496.      TopoOptimizer  AbstractConvCheck
  4497.        nb_node (start, end, changed) (0, 0, 0)
  4498.        init io_toposort 7.15255737305e-06
  4499.        loop time 0.0
  4500.        callback_time 0.0
  4501.    0.000028s - ('local_destructive', 'TopoOptimizer', 48, 0, 0) - 0.000s
  4502.      TopoOptimizer  CURAND_destructive
  4503.        nb_node (start, end, changed) (0, 0, 0)
  4504.        init io_toposort 5.96046447754e-06
  4505.        loop time 0.0
  4506.        callback_time 0.0
  4507.    0.000027s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 0, 0) - 0.000s
  4508.      TopoOptimizer  random_make_inplace_mrg
  4509.        nb_node (start, end, changed) (0, 0, 0)
  4510.        init io_toposort 5.96046447754e-06
  4511.        loop time 0.0
  4512.        callback_time 0.0
  4513.    0.000024s - ('merge1.1', 'MergeOptimizer', 4, 1, 1) - 0.000s
  4514.      MergeOptimizer
  4515.        nb fail=    0 merged=    0 constant=    0
  4516.        time replace=0.00 validate=0.00 callback=0.00
  4517.    0.000023s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 0, 0) - 0.000s
  4518.    0.000005s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 0, 0) - 0.000s
  4519.  
  4520. Here are tips to potentially make your code run faster
  4521.                  (if you think of new ones, suggest them on the mailing list).
  4522.                  Test them first, as they are not guaranteed to always provide a speedup.
  4523.   Sorry, no tip for today.
  4524. Function profiling
  4525. ==================
  4526.   Message: /Users/Ramana/projects/SBRNN/sb/utils.py:27
  4527.   Time in 1 calls to Function.__call__: 1.287460e-05s
  4528.   Time in Function.fn.__call__: 5.006790e-06s (38.889%)
  4529.   Time in thunks: 1.907349e-06s (14.815%)
  4530.   Total compile time: 2.661610e-02s
  4531.     Number of Apply nodes: 1
  4532.     Theano Optimizer time: 9.035826e-03s
  4533.        Theano validate time: 8.153915e-05s
  4534.     Theano Linker time (includes C, CUDA code generation/compiling): 5.738735e-04s
  4535.        Import time 0.000000e+00s
  4536.        Node make_thunk time 4.589558e-04s
  4537.  
  4538. Time in all call to theano.grad() 2.656322e+00s
  4539. Time since theano import 477.860s
  4540. Class
  4541. ---
  4542. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  4543.   100.0%   100.0%       0.000s       1.91e-06s     C        1       1   theano.compile.ops.DeepCopyOp
  4544.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  4545.  
  4546. Ops
  4547. ---
  4548. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  4549.   100.0%   100.0%       0.000s       1.91e-06s     C        1        1   DeepCopyOp
  4550.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  4551.  
  4552. Apply
  4553. ------
  4554. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  4555.   100.0%   100.0%       0.000s       1.91e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  4556.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  4557.  
  4558. Optimizer Profile
  4559. -----------------
  4560.  SeqOptimizer  OPT_FAST_RUN  time 0.009s for 1/0 nodes before/after optimization
  4561.    0.001s for callback
  4562.        0.000s for fgraph.validate()
  4563.    time      - (name, class, index, nodes before, nodes after) - validate time
  4564.    0.002627s - ('canonicalize', 'EquilibriumOptimizer', 6, 1, 0) - 0.000s
  4565.      EquilibriumOptimizer      canonicalize
  4566.        time 0.002s for 2 passes
  4567.        nb nodes (start, end,  max) 1 0 1
  4568.        time io_toposort 0.000s
  4569.        time in local optimizers 0.001s
  4570.        time in global optimizers 0.000s
  4571.        time in final optimizers 0.001s
  4572.        time in cleanup optimizers 0.000s
  4573.         0 - 0.002s 4 (0.001s in global opts, 0.000s io_toposort) - 1 nodes - ('topo_constant_folding', 1) ('local_upcast_elemwise_constant_inputs', 1) ('local_dimshuffle_lift', 1) ('MergeOptimizer', 1)
  4574.         1 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4575.        times - times applied - nb node created - name:
  4576.        0.001s - 1 - 0 - topo_constant_folding
  4577.        0.001s - 1 - 3 - local_upcast_elemwise_constant_inputs
  4578.        0.000s - 1 - 1 - MergeOptimizer
  4579.        0.000s - 1 - 1 - local_dimshuffle_lift
  4580.        0.000s - in 85 optimization that were not used (display only those with a runtime > 0)
  4581.          0.000s - local_useless_elemwise
  4582.          0.000s - local_fill_sink
  4583.          0.000s - local_func_inv
  4584.          0.000s - local_cast_cast
  4585.          0.000s - local_track_shape_i
  4586.          0.000s - local_merge_switch_same_cond
  4587.          0.000s - local_fill_cut
  4588.          0.000s - local_expm1
  4589.          0.000s - local_useless_switch
  4590.          0.000s - local_useless_elemwise_comparison
  4591.          0.000s - local_lift_transpose_through_dot
  4592.  
  4593.      Global, final and clean up optimizers
  4594.      Iter 0
  4595.              TopoOptimizer  topo_constant_folding
  4596.                nb_node (start, end, changed) (2, 0, 2)
  4597.                init io_toposort 3.48091125488e-05
  4598.                loop time 0.000529050827026
  4599.                callback_time 0.000237703323364
  4600.              MergeOptimizer
  4601.                nb fail=    0 merged=    1 constant=    0
  4602.                time replace=0.00 validate=0.00 callback=0.00
  4603.      Iter 1
  4604.              TopoOptimizer  topo_constant_folding
  4605.                nb_node (start, end, changed) (0, 0, 0)
  4606.                init io_toposort 1.00135803223e-05
  4607.                loop time 0.0
  4608.                callback_time 0.0
  4609.              MergeOptimizer
  4610.                nb fail=    0 merged=    0 constant=    0
  4611.                time replace=0.00 validate=0.00 callback=0.00
  4612.    0.000537s - ('gpuarray_opt', 'SeqOptimizer', 16, 0, 0) - 0.000s
  4613.      SeqOptimizer      gpuarray_opt  time 0.000s for 0/0 nodes before/after optimization
  4614.        0.000s for callback
  4615.            0.000s for fgraph.validate()
  4616.        0.000301s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 0, 0) - 0.000s
  4617.          EquilibriumOptimizer          gpuarray_local_optimizations
  4618.            time 0.000s for 1 passes
  4619.            nb nodes (start, end,  max) 0 0 0
  4620.            time io_toposort 0.000s
  4621.            time in local optimizers 0.000s
  4622.            time in global optimizers 0.000s
  4623.            time in final optimizers 0.000s
  4624.            time in cleanup optimizers 0.000s
  4625.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4626.        0.000067s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  4627.          EquilibriumOptimizer          gpuarray_cut_transfers
  4628.            time 0.000s for 1 passes
  4629.            nb nodes (start, end,  max) 0 0 0
  4630.            time io_toposort 0.000s
  4631.            time in local optimizers 0.000s
  4632.            time in global optimizers 0.000s
  4633.            time in final optimizers 0.000s
  4634.            time in cleanup optimizers 0.000s
  4635.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4636.        0.000044s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 0, 0) - 0.000s
  4637.          GraphToGPUOptimizer          gpuarray_graph_optimization
  4638.            time io_toposort 0.000s
  4639.          Total time taken by local optimizers 0.000s
  4640.        0.000007s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 0, 0) - 0.000s
  4641.  
  4642.    0.000476s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 0, 0) - 0.000s
  4643.      EquilibriumOptimizer      scan_eqopt2
  4644.        time 0.000s for 1 passes
  4645.        nb nodes (start, end,  max) 0 0 0
  4646.        time io_toposort 0.000s
  4647.        time in local optimizers 0.000s
  4648.        time in global optimizers 0.000s
  4649.        time in final optimizers 0.000s
  4650.        time in cleanup optimizers 0.000s
  4651.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4652.      Global, final and clean up optimizers
  4653.      Iter 0
  4654.              TopoOptimizer  constant_folding_for_scan2
  4655.                nb_node (start, end, changed) (0, 0, 0)
  4656.                init io_toposort 1.31130218506e-05
  4657.                loop time 1.19209289551e-06
  4658.                callback_time 0.0
  4659.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  4660.                nb_node (start, end, changed) (0, 0, 0)
  4661.                init io_toposort 1.00135803223e-05
  4662.                loop time 0.0
  4663.                callback_time 0.0
  4664.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  4665.                nb_node (start, end, changed) (0, 0, 0)
  4666.                init io_toposort 9.05990600586e-06
  4667.                loop time 0.0
  4668.                callback_time 0.0
  4669.              TopoOptimizer  scanOp_merge_inouts
  4670.                nb_node (start, end, changed) (0, 0, 0)
  4671.                init io_toposort 8.10623168945e-06
  4672.                loop time 0.0
  4673.                callback_time 0.0
  4674.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  4675.                nb_node (start, end, changed) (0, 0, 0)
  4676.                init io_toposort 8.10623168945e-06
  4677.                loop time 0.0
  4678.                callback_time 0.0
  4679.    0.000474s - ('elemwise_fusion', 'SeqOptimizer', 19, 0, 0) - 0.000s
  4680.      SeqOptimizer      elemwise_fusion  time 0.000s for 0/0 nodes before/after optimization
  4681.        0.000s for callback
  4682.            0.000s for fgraph.validate()
  4683.        0.000180s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 0, 0) - 0.000s
  4684.          FusionOptimizer
  4685.           nb_iter 1
  4686.           nb_replacement 0
  4687.           nb_inconsistency_replace 0
  4688.           validate_time 0.0
  4689.           callback_time 0.0
  4690.           time_toposort 9.53674316406e-07
  4691.        0.000174s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 0, 0) - 0.000s
  4692.          FusionOptimizer
  4693.           nb_iter 1
  4694.           nb_replacement 0
  4695.           nb_inconsistency_replace 0
  4696.           validate_time 0.0
  4697.           callback_time 0.0
  4698.           time_toposort 9.53674316406e-07
  4699.  
  4700.    0.000467s - ('BlasOpt', 'SeqOptimizer', 12, 0, 0) - 0.000s
  4701.      SeqOptimizer      BlasOpt  time 0.000s for 0/0 nodes before/after optimization
  4702.        0.000s for callback
  4703.            0.000s for fgraph.validate()
  4704.        0.000150s - ('gemm_optimizer', 'GemmOptimizer', 1, 0, 0) - 0.000s
  4705.          GemmOptimizer
  4706.           nb_iter 1
  4707.           nb_replacement 0
  4708.           nb_replacement_didn_t_remove 0
  4709.           nb_inconsistency_make 0
  4710.           nb_inconsistency_replace 0
  4711.           time_canonicalize 0
  4712.           time_factor_can 0
  4713.           time_factor_list 0
  4714.           time_toposort 8.10623168945e-06
  4715.           validate_time 0.0
  4716.           callback_time 0.0
  4717.        0.000090s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  4718.          EquilibriumOptimizer          local_gemm_to_gemv
  4719.            time 0.000s for 1 passes
  4720.            nb nodes (start, end,  max) 0 0 0
  4721.            time io_toposort 0.000s
  4722.            time in local optimizers 0.000s
  4723.            time in global optimizers 0.000s
  4724.            time in final optimizers 0.000s
  4725.            time in cleanup optimizers 0.000s
  4726.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4727.        0.000042s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 0, 0) - 0.000s
  4728.          TopoOptimizer  local_dot_to_dot22
  4729.            nb_node (start, end, changed) (0, 0, 0)
  4730.            init io_toposort 1.78813934326e-05
  4731.            loop time 0.0
  4732.            callback_time 0.0
  4733.        0.000037s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 0, 0) - 0.000s
  4734.          TopoOptimizer  local_dot22_to_dot22scalar
  4735.            nb_node (start, end, changed) (0, 0, 0)
  4736.            init io_toposort 1.62124633789e-05
  4737.            loop time 9.53674316406e-07
  4738.            callback_time 0.0
  4739.        0.000029s - ('use_c_blas', 'TopoOptimizer', 4, 0, 0) - 0.000s
  4740.          TopoOptimizer  use_c_blas
  4741.            nb_node (start, end, changed) (0, 0, 0)
  4742.            init io_toposort 6.91413879395e-06
  4743.            loop time 9.53674316406e-07
  4744.            callback_time 0.0
  4745.  LocalOptGroup
  4746.  ---------------------
  4747.  --- The Optimizer wasn't successful ---
  4748.  
  4749.       0.000026s - ('use_scipy_ger', 'TopoOptimizer', 5, 0, 0) - 0.000s
  4750.         TopoOptimizer  scipy_blas
  4751.           nb_node (start, end, changed) (0, 0, 0)
  4752.           init io_toposort 5.96046447754e-06
  4753.           loop time 0.0
  4754.           callback_time 0.0
  4755.  
  4756.   0.000455s - ('add_destroy_handler', 'AddDestroyHandler', 23, 0, 0) - 0.000s
  4757.   0.000364s - ('specialize', 'EquilibriumOptimizer', 13, 0, 0) - 0.000s
  4758.     EquilibriumOptimizer      specialize
  4759.       time 0.000s for 1 passes
  4760.       nb nodes (start, end,  max) 0 0 0
  4761.       time io_toposort 0.000s
  4762.       time in local optimizers 0.000s
  4763.       time in global optimizers 0.000s
  4764.       time in final optimizers 0.000s
  4765.       time in cleanup optimizers 0.000s
  4766.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4767.     Global, final and clean up optimizers
  4768.     Iter 0
  4769.             TopoOptimizer  topo_constant_folding
  4770.               nb_node (start, end, changed) (0, 0, 0)
  4771.               init io_toposort 6.91413879395e-06
  4772.               loop time 0.0
  4773.               callback_time 0.0
  4774.   0.000341s - ('stabilize', 'EquilibriumOptimizer', 8, 0, 0) - 0.000s
  4775.     EquilibriumOptimizer      stabilize
  4776.       time 0.000s for 1 passes
  4777.       nb nodes (start, end,  max) 0 0 0
  4778.       time io_toposort 0.000s
  4779.       time in local optimizers 0.000s
  4780.       time in global optimizers 0.000s
  4781.       time in final optimizers 0.000s
  4782.       time in cleanup optimizers 0.000s
  4783.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4784.     Global, final and clean up optimizers
  4785.     Iter 0
  4786.             TopoOptimizer  topo_constant_folding
  4787.               nb_node (start, end, changed) (0, 0, 0)
  4788.               init io_toposort 1.00135803223e-05
  4789.               loop time 0.0
  4790.               callback_time 0.0
  4791.   0.000232s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 0, 0) - 0.000s
  4792.     FusionOptimizer
  4793.      nb_iter 1
  4794.      nb_replacement 0
  4795.      nb_inconsistency_replace 0
  4796.      validate_time 0.0
  4797.      callback_time 0.0
  4798.      time_toposort 0.0
  4799.   0.000198s - ('merge3', 'MergeOptimizer', 51, 0, 0) - 0.000s
  4800.     MergeOptimizer
  4801.       nb fail=    0 merged=    0 constant=    0
  4802.       time replace=0.00 validate=0.00 callback=0.00
  4803.   0.000197s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 1, 1) - 0.000s
  4804.     EquilibriumOptimizer      scan_eqopt1
  4805.       time 0.000s for 1 passes
  4806.       nb nodes (start, end,  max) 1 1 1
  4807.       time io_toposort 0.000s
  4808.       time in local optimizers 0.000s
  4809.       time in global optimizers 0.000s
  4810.       time in final optimizers 0.000s
  4811.       time in cleanup optimizers 0.000s
  4812.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  4813.     Global, final and clean up optimizers
  4814.     Iter 0
  4815.             SeqOptimizer              all_pushout_opt  time 0.000s for 1/1 nodes before/after optimization
  4816.               0.000s for callback
  4817.                   0.000s for fgraph.validate()
  4818.               0.000055s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 1, 1) - 0.000s
  4819.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  4820.                   nb_node (start, end, changed) (1, 1, 0)
  4821.                   init io_toposort 2.09808349609e-05
  4822.                   loop time 5.00679016113e-06
  4823.                   callback_time 0.0
  4824.               0.000013s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 1, 1) - 0.000s
  4825.               0.000008s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 1, 1) - 0.000s
  4826.               0.000008s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 1, 1) - 0.000s
  4827.               0.000008s - ('scan_pushout_dot1', 'PushOutDot1', 3, 1, 1) - 0.000s
  4828.  
  4829.   0.000179s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 0, 0) - 0.000s
  4830.     FusionOptimizer
  4831.      nb_iter 1
  4832.      nb_replacement 0
  4833.      nb_inconsistency_replace 0
  4834.      validate_time 0.0
  4835.      callback_time 0.0
  4836.      time_toposort 0.0
  4837.   0.000148s - ('blas_opt_inplace', 'TopoOptimizer', 34, 0, 0) - 0.000s
  4838.     TopoOptimizer  InplaceBlasOpt
  4839.       nb_node (start, end, changed) (0, 0, 0)
  4840.       init io_toposort 7.86781311035e-06
  4841.       loop time 0.0
  4842.       callback_time 0.0
  4843. LocalOptGroup
  4844. ---------------------
  4845. --- The Optimizer wasn't successful ---
  4846.  
  4847.    0.000130s - ('merge2', 'MergeOptimizer', 22, 0, 0) - 0.000s
  4848.      MergeOptimizer
  4849.        nb fail=    0 merged=    0 constant=    0
  4850.        time replace=0.00 validate=0.00 callback=0.00
  4851.    0.000120s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 0, 0) - 0.000s
  4852.      EquilibriumOptimizer      uncanonicalize
  4853.        time 0.000s for 1 passes
  4854.        nb nodes (start, end,  max) 0 0 0
  4855.        time io_toposort 0.000s
  4856.        time in local optimizers 0.000s
  4857.        time in global optimizers 0.000s
  4858.        time in final optimizers 0.000s
  4859.        time in cleanup optimizers 0.000s
  4860.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4861.      Global, final and clean up optimizers
  4862.      Iter 0
  4863.              TopoOptimizer  topo_constant_folding
  4864.                nb_node (start, end, changed) (0, 0, 0)
  4865.                init io_toposort 5.96046447754e-06
  4866.                loop time 0.0
  4867.                callback_time 0.0
  4868.    0.000116s - ('merge1', 'MergeOptimizer', 0, 1, 1) - 0.000s
  4869.      MergeOptimizer
  4870.        nb fail=    0 merged=    0 constant=    0
  4871.        time replace=0.00 validate=0.00 callback=0.00
  4872.    0.000113s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 0, 0) - 0.000s
  4873.      TopoOptimizer  InplaceGpuBlasOpt
  4874.        nb_node (start, end, changed) (0, 0, 0)
  4875.        init io_toposort 1.4066696167e-05
  4876.        loop time 0.0
  4877.        callback_time 0.0
  4878.  LocalOptGroup
  4879.  ---------------------
  4880.  --- The Optimizer wasn't successful ---
  4881.  
  4882.   0.000087s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 0, 0) - 0.000s
  4883.     TopoOptimizer  local_dnn_conv_inplace
  4884.       nb_node (start, end, changed) (0, 0, 0)
  4885.       init io_toposort 9.05990600586e-06
  4886.       loop time 0.0
  4887.       callback_time 0.0
  4888. LocalOptGroup
  4889. ---------------------
  4890. --- The Optimizer wasn't successful ---
  4891.  
  4892.    0.000085s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 0, 0) - 0.000s
  4893.      TopoOptimizer  InplaceGpuaBlasOpt
  4894.        nb_node (start, end, changed) (0, 0, 0)
  4895.        init io_toposort 1.00135803223e-05
  4896.        loop time 0.0
  4897.        callback_time 0.0
  4898.  LocalOptGroup
  4899.  ---------------------
  4900.    time taken - times applied - times tried - name - node_created:
  4901.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  4902.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  4903.  
  4904.    0.000084s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 0, 0) - 0.000s
  4905.      TopoOptimizer  local_dnna_conv_inplace
  4906.        nb_node (start, end, changed) (0, 0, 0)
  4907.        init io_toposort 8.10623168945e-06
  4908.        loop time 0.0
  4909.        callback_time 0.0
  4910.  LocalOptGroup
  4911.  ---------------------
  4912.    time taken - times applied - times tried - name - node_created:
  4913.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  4914.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  4915.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  4916.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  4917.  
  4918.    0.000077s - ('specialize_device', 'EquilibriumOptimizer', 17, 0, 0) - 0.000s
  4919.      EquilibriumOptimizer      specialize_device
  4920.        time 0.000s for 1 passes
  4921.        nb nodes (start, end,  max) 0 0 0
  4922.        time io_toposort 0.000s
  4923.        time in local optimizers 0.000s
  4924.        time in global optimizers 0.000s
  4925.        time in final optimizers 0.000s
  4926.        time in cleanup optimizers 0.000s
  4927.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  4928.    0.000076s - ('useless', 'TopoOptimizer', 3, 1, 1) - 0.000s
  4929.      TopoOptimizer  useless
  4930.        nb_node (start, end, changed) (1, 1, 0)
  4931.        init io_toposort 1.78813934326e-05
  4932.        loop time 2.90870666504e-05
  4933.        callback_time 0.0
  4934.  LocalOptGroup
  4935.  ---------------------
  4936.    time taken - times applied - times tried - name - node_created:
  4937.    -0.000s - 0 - 1 - local_useless_elemwise_comparison - 0
  4938.    -0.000s - 0 - 1 - local_useless_switch - 0
  4939.    -0.000s - 0 - 1 - local_useless_elemwise - 0
  4940.    0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  4941.  
  4942.    0.000076s - ('ShapeOpt', 'ShapeOptimizer', 2, 1, 1) - 0.000s
  4943.    0.000053s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 0, 0) - 0.000s
  4944.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  4945.        nb_node (start, end, changed) (0, 0, 0)
  4946.        init io_toposort 8.10623168945e-06
  4947.        loop time 0.0
  4948.        callback_time 0.0
  4949.    0.000049s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 0, 0) - 0.000s
  4950.      TopoOptimizer  local_gemm16_inplace
  4951.        nb_node (start, end, changed) (0, 0, 0)
  4952.        init io_toposort 7.15255737305e-06
  4953.        loop time 0.0
  4954.        callback_time 0.0
  4955.    0.000049s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 0, 0) - 0.000s
  4956.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  4957.        nb_node (start, end, changed) (0, 0, 0)
  4958.        init io_toposort 6.91413879395e-06
  4959.        loop time 0.0
  4960.        callback_time 0.0
  4961.    0.000048s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 0, 0) - 0.000s
  4962.      TopoOptimizer  local_inplace_incsubtensor1
  4963.        nb_node (start, end, changed) (0, 0, 0)
  4964.        init io_toposort 6.91413879395e-06
  4965.        loop time 9.53674316406e-07
  4966.        callback_time 0.0
  4967.    0.000047s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 0, 0) - 0.000s
  4968.      TopoOptimizer  local_inplace_sparseblockgemv
  4969.        nb_node (start, end, changed) (0, 0, 0)
  4970.        init io_toposort 7.15255737305e-06
  4971.        loop time 9.53674316406e-07
  4972.        callback_time 0.0
  4973.    0.000047s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 0, 0) - 0.000s
  4974.      TopoOptimizer  local_inplace_sparse_block_gemv
  4975.        nb_node (start, end, changed) (0, 0, 0)
  4976.        init io_toposort 7.86781311035e-06
  4977.        loop time 0.0
  4978.        callback_time 0.0
  4979.    0.000046s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 0, 0) - 0.000s
  4980.      TopoOptimizer  local_inplace_sparseblockouter
  4981.        nb_node (start, end, changed) (0, 0, 0)
  4982.        init io_toposort 7.15255737305e-06
  4983.        loop time 0.0
  4984.        callback_time 0.0
  4985.    0.000046s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 0, 0) - 0.000s
  4986.      TopoOptimizer  local_inplace_sparse_block_outer
  4987.        nb_node (start, end, changed) (0, 0, 0)
  4988.        init io_toposort 6.91413879395e-06
  4989.        loop time 0.0
  4990.        callback_time 0.0
  4991.    0.000046s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 0, 0) - 0.000s
  4992.      TopoOptimizer  local_inplace_setsubtensor
  4993.        nb_node (start, end, changed) (0, 0, 0)
  4994.        init io_toposort 6.91413879395e-06
  4995.        loop time 0.0
  4996.        callback_time 0.0
  4997.    0.000046s - ('c_blas_destructive', 'TopoOptimizer', 37, 0, 0) - 0.000s
  4998.      TopoOptimizer  c_blas_destructive
  4999.        nb_node (start, end, changed) (0, 0, 0)
  5000.        init io_toposort 8.10623168945e-06
  5001.        loop time 0.0
  5002.        callback_time 0.0
  5003.  LocalOptGroup
  5004.  ---------------------
  5005.  --- The Optimizer wasn't successful ---
  5006.  
  5007.   0.000044s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 1, 1) - 0.000s
  5008.     TopoOptimizer  pre_local_IncSubtensor_serialize
  5009.       nb_node (start, end, changed) (1, 1, 0)
  5010.       init io_toposort 1.59740447998e-05
  5011.       loop time 5.96046447754e-06
  5012.       callback_time 0.0
  5013.   0.000044s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 0, 0) - 0.000s
  5014.   0.000044s - ('merge1.2', 'MergeOptimizer', 7, 0, 0) - 0.000s
  5015.     MergeOptimizer
  5016.       nb fail=    0 merged=    0 constant=    0
  5017.       time replace=0.00 validate=0.00 callback=0.00
  5018.   0.000040s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 0, 0) - 0.000s
  5019.     TopoOptimizer  dimshuffle_as_view
  5020.       nb_node (start, end, changed) (0, 0, 0)
  5021.       init io_toposort 1.21593475342e-05
  5022.       loop time 9.53674316406e-07
  5023.       callback_time 0.0
  5024.   0.000037s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 0, 0) - 0.000s
  5025.     TopoOptimizer  local_fill_to_alloc
  5026.       nb_node (start, end, changed) (0, 0, 0)
  5027.       init io_toposort 1.00135803223e-05
  5028.       loop time 0.0
  5029.       callback_time 0.0
  5030.   0.000034s - ('cond_make_inplace', 'TopoOptimizer', 47, 0, 0) - 0.000s
  5031.     TopoOptimizer  cond_make_inplace
  5032.       nb_node (start, end, changed) (0, 0, 0)
  5033.       init io_toposort 8.10623168945e-06
  5034.       loop time 0.0
  5035.       callback_time 0.0
  5036.   0.000034s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 0, 0) - 0.000s
  5037.   0.000034s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 0, 0) - 0.000s
  5038.     TopoOptimizer  local_elemwise_alloc
  5039.       nb_node (start, end, changed) (0, 0, 0)
  5040.       init io_toposort 8.10623168945e-06
  5041.       loop time 0.0
  5042.       callback_time 0.0
  5043.   0.000033s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 0, 0) - 0.000s
  5044.   0.000033s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 0, 0) - 0.000s
  5045.     TopoOptimizer  local_advincsub1_gpua_inplace
  5046.       nb_node (start, end, changed) (0, 0, 0)
  5047.       init io_toposort 8.10623168945e-06
  5048.       loop time 0.0
  5049.       callback_time 0.0
  5050.   0.000032s - ('local_destructive', 'TopoOptimizer', 48, 0, 0) - 0.000s
  5051.     TopoOptimizer  CURAND_destructive
  5052.       nb_node (start, end, changed) (0, 0, 0)
  5053.       init io_toposort 6.91413879395e-06
  5054.       loop time 0.0
  5055.       callback_time 0.0
  5056.   0.000032s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 0, 0) - 0.000s
  5057.   0.000031s - ('random_make_inplace', 'TopoOptimizer', 49, 0, 0) - 0.000s
  5058.     TopoOptimizer  random_make_inplace
  5059.       nb_node (start, end, changed) (0, 0, 0)
  5060.       init io_toposort 5.96046447754e-06
  5061.       loop time 0.0
  5062.       callback_time 0.0
  5063.   0.000031s - ('make_ger_destructive', 'TopoOptimizer', 41, 0, 0) - 0.000s
  5064.     TopoOptimizer  make_scipy_blas_destructive
  5065.       nb_node (start, end, changed) (0, 0, 0)
  5066.       init io_toposort 6.91413879395e-06
  5067.       loop time 0.0
  5068.       callback_time 0.0
  5069.   0.000030s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 0, 0) - 0.000s
  5070.     TopoOptimizer  random_make_inplace_mrg
  5071.       nb_node (start, end, changed) (0, 0, 0)
  5072.       init io_toposort 7.15255737305e-06
  5073.       loop time 0.0
  5074.       callback_time 0.0
  5075.   0.000029s - ('AbstractConvCheck', 'TopoOptimizer', 18, 0, 0) - 0.000s
  5076.     TopoOptimizer  AbstractConvCheck
  5077.       nb_node (start, end, changed) (0, 0, 0)
  5078.       init io_toposort 8.10623168945e-06
  5079.       loop time 1.19209289551e-06
  5080.       callback_time 0.0
  5081.   0.000025s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 0, 0) - 0.000s
  5082.   0.000020s - ('merge1.1', 'MergeOptimizer', 4, 1, 1) - 0.000s
  5083.     MergeOptimizer
  5084.       nb fail=    0 merged=    0 constant=    0
  5085.       time replace=0.00 validate=0.00 callback=0.00
  5086.   0.000004s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 0, 0) - 0.000s
  5087.  
  5088. Here are tips to potentially make your code run faster
  5089.                 (if you think of new ones, suggest them on the mailing list).
  5090.                 Test them first, as they are not guaranteed to always provide a speedup.
  5091.  Sorry, no tip for today.
  5092. Function profiling
  5093. ==================
  5094.  Message: /Users/Ramana/projects/SBRNN/sb/utils.py:56
  5095.  Time in 1 calls to Function.__call__: 1.502037e-05s
  5096.  Time in Function.fn.__call__: 5.960464e-06s (39.683%)
  5097.  Time in thunks: 4.053116e-06s (26.984%)
  5098.  Total compile time: 1.754310e-01s
  5099.    Number of Apply nodes: 1
  5100.    Theano Optimizer time: 1.580000e-01s
  5101.       Theano validate time: 7.796288e-05s
  5102.    Theano Linker time (includes C, CUDA code generation/compiling): 6.060600e-04s
  5103.       Import time 0.000000e+00s
  5104.       Node make_thunk time 4.839897e-04s
  5105.  
  5106. Time in all call to theano.grad() 2.656322e+00s
  5107. Time since theano import 477.866s
  5108. Class
  5109. ---
  5110. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  5111.  100.0%   100.0%       0.000s       4.05e-06s     C        1       1   theano.compile.ops.DeepCopyOp
  5112.   ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  5113.  
  5114. Ops
  5115. ---
  5116. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  5117.  100.0%   100.0%       0.000s       4.05e-06s     C        1        1   DeepCopyOp
  5118.   ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  5119.  
  5120. Apply
  5121. ------
  5122. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  5123.  100.0%   100.0%       0.000s       4.05e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  5124.   ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  5125.  
  5126. Optimizer Profile
  5127. -----------------
  5128. SeqOptimizer  OPT_FAST_RUN  time 0.158s for 1/0 nodes before/after optimization
  5129.   0.001s for callback
  5130.       0.000s for fgraph.validate()
  5131.   time      - (name, class, index, nodes before, nodes after) - validate time
  5132.   0.149964s - ('BlasOpt', 'SeqOptimizer', 12, 0, 0) - 0.000s
  5133.     SeqOptimizer      BlasOpt  time 0.150s for 0/0 nodes before/after optimization
  5134.       0.000s for callback
  5135.           0.000s for fgraph.validate()
  5136.       0.149595s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  5137.         EquilibriumOptimizer          local_gemm_to_gemv
  5138.           time 0.000s for 1 passes
  5139.           nb nodes (start, end,  max) 0 0 0
  5140.           time io_toposort 0.000s
  5141.           time in local optimizers 0.000s
  5142.           time in global optimizers 0.000s
  5143.           time in final optimizers 0.000s
  5144.           time in cleanup optimizers 0.000s
  5145.            0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5146.       0.000132s - ('gemm_optimizer', 'GemmOptimizer', 1, 0, 0) - 0.000s
  5147.         GemmOptimizer
  5148.          nb_iter 1
  5149.          nb_replacement 0
  5150.          nb_replacement_didn_t_remove 0
  5151.          nb_inconsistency_make 0
  5152.          nb_inconsistency_replace 0
  5153.          time_canonicalize 0
  5154.          time_factor_can 0
  5155.          time_factor_list 0
  5156.          time_toposort 8.10623168945e-06
  5157.          validate_time 0.0
  5158.          callback_time 0.0
  5159.       0.000042s - ('use_c_blas', 'TopoOptimizer', 4, 0, 0) - 0.000s
  5160.         TopoOptimizer  use_c_blas
  5161.           nb_node (start, end, changed) (0, 0, 0)
  5162.           init io_toposort 1.00135803223e-05
  5163.           loop time 0.0
  5164.           callback_time 0.0
  5165. LocalOptGroup
  5166. ---------------------
  5167. --- The Optimizer wasn't successful ---
  5168.  
  5169.        0.000030s - ('use_scipy_ger', 'TopoOptimizer', 5, 0, 0) - 0.000s
  5170.          TopoOptimizer  scipy_blas
  5171.            nb_node (start, end, changed) (0, 0, 0)
  5172.            init io_toposort 8.10623168945e-06
  5173.            loop time 0.0
  5174.            callback_time 0.0
  5175.        0.000027s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 0, 0) - 0.000s
  5176.          TopoOptimizer  local_dot22_to_dot22scalar
  5177.            nb_node (start, end, changed) (0, 0, 0)
  5178.            init io_toposort 5.96046447754e-06
  5179.            loop time 9.53674316406e-07
  5180.            callback_time 0.0
  5181.        0.000027s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 0, 0) - 0.000s
  5182.          TopoOptimizer  local_dot_to_dot22
  5183.            nb_node (start, end, changed) (0, 0, 0)
  5184.            init io_toposort 7.15255737305e-06
  5185.            loop time 1.19209289551e-06
  5186.            callback_time 0.0
  5187.  
  5188.    0.002318s - ('canonicalize', 'EquilibriumOptimizer', 6, 1, 0) - 0.000s
  5189.      EquilibriumOptimizer      canonicalize
  5190.        time 0.002s for 2 passes
  5191.        nb nodes (start, end,  max) 1 0 1
  5192.        time io_toposort 0.000s
  5193.        time in local optimizers 0.001s
  5194.        time in global optimizers 0.000s
  5195.        time in final optimizers 0.000s
  5196.        time in cleanup optimizers 0.000s
  5197.         0 - 0.002s 4 (0.000s in global opts, 0.000s io_toposort) - 1 nodes - ('topo_constant_folding', 1) ('local_upcast_elemwise_constant_inputs', 1) ('local_dimshuffle_lift', 1) ('MergeOptimizer', 1)
  5198.         1 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5199.        times - times applied - nb node created - name:
  5200.        0.001s - 1 - 3 - local_upcast_elemwise_constant_inputs
  5201.        0.000s - 1 - 0 - topo_constant_folding
  5202.        0.000s - 1 - 1 - local_dimshuffle_lift
  5203.        0.000s - 1 - 1 - MergeOptimizer
  5204.        0.000s - in 85 optimization that were not used (display only those with a runtime > 0)
  5205.          0.000s - local_fill_sink
  5206.          0.000s - local_useless_elemwise
  5207.          0.000s - local_func_inv
  5208.          0.000s - local_track_shape_i
  5209.          0.000s - local_merge_switch_same_cond
  5210.          0.000s - local_fill_cut
  5211.          0.000s - local_cast_cast
  5212.          0.000s - local_useless_elemwise_comparison
  5213.          0.000s - local_lift_transpose_through_dot
  5214.          0.000s - local_expm1
  5215.          0.000s - local_useless_switch
  5216.  
  5217.      Global, final and clean up optimizers
  5218.      Iter 0
  5219.              TopoOptimizer  topo_constant_folding
  5220.                nb_node (start, end, changed) (2, 0, 2)
  5221.                init io_toposort 2.59876251221e-05
  5222.                loop time 0.00040602684021
  5223.                callback_time 0.000176191329956
  5224.              MergeOptimizer
  5225.                nb fail=    0 merged=    1 constant=    0
  5226.                time replace=0.00 validate=0.00 callback=0.00
  5227.      Iter 1
  5228.              TopoOptimizer  topo_constant_folding
  5229.                nb_node (start, end, changed) (0, 0, 0)
  5230.                init io_toposort 7.15255737305e-06
  5231.                loop time 0.0
  5232.                callback_time 0.0
  5233.              MergeOptimizer
  5234.                nb fail=    0 merged=    0 constant=    0
  5235.                time replace=0.00 validate=0.00 callback=0.00
  5236.    0.000641s - ('gpuarray_opt', 'SeqOptimizer', 16, 0, 0) - 0.000s
  5237.      SeqOptimizer      gpuarray_opt  time 0.001s for 0/0 nodes before/after optimization
  5238.        0.000s for callback
  5239.            0.000s for fgraph.validate()
  5240.        0.000343s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 0, 0) - 0.000s
  5241.          EquilibriumOptimizer          gpuarray_local_optimizations
  5242.            time 0.000s for 1 passes
  5243.            nb nodes (start, end,  max) 0 0 0
  5244.            time io_toposort 0.000s
  5245.            time in local optimizers 0.000s
  5246.            time in global optimizers 0.000s
  5247.            time in final optimizers 0.000s
  5248.            time in cleanup optimizers 0.000s
  5249.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5250.        0.000088s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  5251.          EquilibriumOptimizer          gpuarray_cut_transfers
  5252.            time 0.000s for 1 passes
  5253.            nb nodes (start, end,  max) 0 0 0
  5254.            time io_toposort 0.000s
  5255.            time in local optimizers 0.000s
  5256.            time in global optimizers 0.000s
  5257.            time in final optimizers 0.000s
  5258.            time in cleanup optimizers 0.000s
  5259.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5260.        0.000067s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 0, 0) - 0.000s
  5261.          GraphToGPUOptimizer          gpuarray_graph_optimization
  5262.            time io_toposort 0.000s
  5263.          Total time taken by local optimizers 0.000s
  5264.        0.000009s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 0, 0) - 0.000s
  5265.  
  5266.    0.000533s - ('elemwise_fusion', 'SeqOptimizer', 19, 0, 0) - 0.000s
  5267.      SeqOptimizer      elemwise_fusion  time 0.000s for 0/0 nodes before/after optimization
  5268.        0.000s for callback
  5269.            0.000s for fgraph.validate()
  5270.        0.000201s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 0, 0) - 0.000s
  5271.          FusionOptimizer
  5272.           nb_iter 1
  5273.           nb_replacement 0
  5274.           nb_inconsistency_replace 0
  5275.           validate_time 0.0
  5276.           callback_time 0.0
  5277.           time_toposort 9.53674316406e-07
  5278.        0.000199s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 0, 0) - 0.000s
  5279.          FusionOptimizer
  5280.           nb_iter 1
  5281.           nb_replacement 0
  5282.           nb_inconsistency_replace 0
  5283.           validate_time 0.0
  5284.           callback_time 0.0
  5285.           time_toposort 9.53674316406e-07
  5286.  
  5287.    0.000514s - ('add_destroy_handler', 'AddDestroyHandler', 23, 0, 0) - 0.000s
  5288.    0.000436s - ('specialize', 'EquilibriumOptimizer', 13, 0, 0) - 0.000s
  5289.      EquilibriumOptimizer      specialize
  5290.        time 0.000s for 1 passes
  5291.        nb nodes (start, end,  max) 0 0 0
  5292.        time io_toposort 0.000s
  5293.        time in local optimizers 0.000s
  5294.        time in global optimizers 0.000s
  5295.        time in final optimizers 0.000s
  5296.        time in cleanup optimizers 0.000s
  5297.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5298.      Global, final and clean up optimizers
  5299.      Iter 0
  5300.              TopoOptimizer  topo_constant_folding
  5301.                nb_node (start, end, changed) (0, 0, 0)
  5302.                init io_toposort 7.86781311035e-06
  5303.                loop time 0.0
  5304.                callback_time 0.0
  5305.    0.000254s - ('stabilize', 'EquilibriumOptimizer', 8, 0, 0) - 0.000s
  5306.      EquilibriumOptimizer      stabilize
  5307.        time 0.000s for 1 passes
  5308.        nb nodes (start, end,  max) 0 0 0
  5309.        time io_toposort 0.000s
  5310.        time in local optimizers 0.000s
  5311.        time in global optimizers 0.000s
  5312.        time in final optimizers 0.000s
  5313.        time in cleanup optimizers 0.000s
  5314.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5315.      Global, final and clean up optimizers
  5316.      Iter 0
  5317.              TopoOptimizer  topo_constant_folding
  5318.                nb_node (start, end, changed) (0, 0, 0)
  5319.                init io_toposort 6.91413879395e-06
  5320.                loop time 0.0
  5321.                callback_time 0.0
  5322.    0.000245s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 0, 0) - 0.000s
  5323.      EquilibriumOptimizer      scan_eqopt2
  5324.        time 0.000s for 1 passes
  5325.        nb nodes (start, end,  max) 0 0 0
  5326.        time io_toposort 0.000s
  5327.        time in local optimizers 0.000s
  5328.        time in global optimizers 0.000s
  5329.        time in final optimizers 0.000s
  5330.        time in cleanup optimizers 0.000s
  5331.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5332.      Global, final and clean up optimizers
  5333.      Iter 0
  5334.              TopoOptimizer  constant_folding_for_scan2
  5335.                nb_node (start, end, changed) (0, 0, 0)
  5336.                init io_toposort 7.86781311035e-06
  5337.                loop time 0.0
  5338.                callback_time 0.0
  5339.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  5340.                nb_node (start, end, changed) (0, 0, 0)
  5341.                init io_toposort 7.15255737305e-06
  5342.                loop time 0.0
  5343.                callback_time 0.0
  5344.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  5345.                nb_node (start, end, changed) (0, 0, 0)
  5346.                init io_toposort 5.96046447754e-06
  5347.                loop time 9.53674316406e-07
  5348.                callback_time 0.0
  5349.              TopoOptimizer  scanOp_merge_inouts
  5350.                nb_node (start, end, changed) (0, 0, 0)
  5351.                init io_toposort 6.19888305664e-06
  5352.                loop time 0.0
  5353.                callback_time 0.0
  5354.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  5355.                nb_node (start, end, changed) (0, 0, 0)
  5356.                init io_toposort 5.96046447754e-06
  5357.                loop time 0.0
  5358.                callback_time 0.0
  5359.    0.000208s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 1, 1) - 0.000s
  5360.      EquilibriumOptimizer      scan_eqopt1
  5361.        time 0.000s for 1 passes
  5362.        nb nodes (start, end,  max) 1 1 1
  5363.        time io_toposort 0.000s
  5364.        time in local optimizers 0.000s
  5365.        time in global optimizers 0.000s
  5366.        time in final optimizers 0.000s
  5367.        time in cleanup optimizers 0.000s
  5368.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  5369.      Global, final and clean up optimizers
  5370.      Iter 0
  5371.              SeqOptimizer              all_pushout_opt  time 0.000s for 1/1 nodes before/after optimization
  5372.                0.000s for callback
  5373.                    0.000s for fgraph.validate()
  5374.                0.000057s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 1, 1) - 0.000s
  5375.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  5376.                    nb_node (start, end, changed) (1, 1, 0)
  5377.                    init io_toposort 2.09808349609e-05
  5378.                    loop time 5.00679016113e-06
  5379.                    callback_time 0.0
  5380.                0.000013s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 1, 1) - 0.000s
  5381.                0.000009s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 1, 1) - 0.000s
  5382.                0.000008s - ('scan_pushout_dot1', 'PushOutDot1', 3, 1, 1) - 0.000s
  5383.                0.000008s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 1, 1) - 0.000s
  5384.  
  5385.    0.000194s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 0, 0) - 0.000s
  5386.      FusionOptimizer
  5387.       nb_iter 1
  5388.       nb_replacement 0
  5389.       nb_inconsistency_replace 0
  5390.       validate_time 0.0
  5391.       callback_time 0.0
  5392.       time_toposort 9.53674316406e-07
  5393.    0.000193s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 0, 0) - 0.000s
  5394.      FusionOptimizer
  5395.       nb_iter 1
  5396.       nb_replacement 0
  5397.       nb_inconsistency_replace 0
  5398.       validate_time 0.0
  5399.       callback_time 0.0
  5400.       time_toposort 0.0
  5401.    0.000188s - ('merge3', 'MergeOptimizer', 51, 0, 0) - 0.000s
  5402.      MergeOptimizer
  5403.        nb fail=    0 merged=    0 constant=    0
  5404.        time replace=0.00 validate=0.00 callback=0.00
  5405.    0.000136s - ('merge2', 'MergeOptimizer', 22, 0, 0) - 0.000s
  5406.      MergeOptimizer
  5407.        nb fail=    0 merged=    0 constant=    0
  5408.        time replace=0.00 validate=0.00 callback=0.00
  5409.    0.000136s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 0, 0) - 0.000s
  5410.      EquilibriumOptimizer      uncanonicalize
  5411.        time 0.000s for 1 passes
  5412.        nb nodes (start, end,  max) 0 0 0
  5413.        time io_toposort 0.000s
  5414.        time in local optimizers 0.000s
  5415.        time in global optimizers 0.000s
  5416.        time in final optimizers 0.000s
  5417.        time in cleanup optimizers 0.000s
  5418.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5419.      Global, final and clean up optimizers
  5420.      Iter 0
  5421.              TopoOptimizer  topo_constant_folding
  5422.                nb_node (start, end, changed) (0, 0, 0)
  5423.                init io_toposort 7.86781311035e-06
  5424.                loop time 0.0
  5425.                callback_time 0.0
  5426.    0.000096s - ('merge1', 'MergeOptimizer', 0, 1, 1) - 0.000s
  5427.      MergeOptimizer
  5428.        nb fail=    0 merged=    0 constant=    0
  5429.        time replace=0.00 validate=0.00 callback=0.00
  5430.    0.000094s - ('specialize_device', 'EquilibriumOptimizer', 17, 0, 0) - 0.000s
  5431.      EquilibriumOptimizer      specialize_device
  5432.        time 0.000s for 1 passes
  5433.        nb nodes (start, end,  max) 0 0 0
  5434.        time io_toposort 0.000s
  5435.        time in local optimizers 0.000s
  5436.        time in global optimizers 0.000s
  5437.        time in final optimizers 0.000s
  5438.        time in cleanup optimizers 0.000s
  5439.         0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5440.    0.000080s - ('blas_opt_inplace', 'TopoOptimizer', 34, 0, 0) - 0.000s
  5441.      TopoOptimizer  InplaceBlasOpt
  5442.        nb_node (start, end, changed) (0, 0, 0)
  5443.        init io_toposort 7.86781311035e-06
  5444.        loop time 0.0
  5445.        callback_time 0.0
  5446.  LocalOptGroup
  5447.  ---------------------
  5448.  --- The Optimizer wasn't successful ---
  5449.  
  5450.   0.000079s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 0, 0) - 0.000s
  5451.     TopoOptimizer  InplaceGpuaBlasOpt
  5452.       nb_node (start, end, changed) (0, 0, 0)
  5453.       init io_toposort 6.91413879395e-06
  5454.       loop time 0.0
  5455.       callback_time 0.0
  5456. LocalOptGroup
  5457. ---------------------
  5458.   time taken - times applied - times tried - name - node_created:
  5459.   -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  5460.   0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  5461.  
  5462.   0.000079s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 0, 0) - 0.000s
  5463.     TopoOptimizer  InplaceGpuBlasOpt
  5464.       nb_node (start, end, changed) (0, 0, 0)
  5465.       init io_toposort 8.10623168945e-06
  5466.       loop time 0.0
  5467.       callback_time 0.0
  5468. LocalOptGroup
  5469. ---------------------
  5470. --- The Optimizer wasn't successful ---
  5471.  
  5472.    0.000079s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 0, 0) - 0.000s
  5473.      TopoOptimizer  local_dnn_conv_inplace
  5474.        nb_node (start, end, changed) (0, 0, 0)
  5475.        init io_toposort 7.86781311035e-06
  5476.        loop time 0.0
  5477.        callback_time 0.0
  5478.  LocalOptGroup
  5479.  ---------------------
  5480.  --- The Optimizer wasn't successful ---
  5481.  
  5482.   0.000078s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 0, 0) - 0.000s
  5483.     TopoOptimizer  local_dnna_conv_inplace
  5484.       nb_node (start, end, changed) (0, 0, 0)
  5485.       init io_toposort 7.15255737305e-06
  5486.       loop time 0.0
  5487.       callback_time 0.0
  5488. LocalOptGroup
  5489. ---------------------
  5490.   time taken - times applied - times tried - name - node_created:
  5491.   -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  5492.   -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  5493.   -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  5494.   0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  5495.  
  5496.   0.000076s - ('useless', 'TopoOptimizer', 3, 1, 1) - 0.000s
  5497.     TopoOptimizer  useless
  5498.       nb_node (start, end, changed) (1, 1, 0)
  5499.       init io_toposort 1.69277191162e-05
  5500.       loop time 3.00407409668e-05
  5501.       callback_time 0.0
  5502. LocalOptGroup
  5503. ---------------------
  5504.   time taken - times applied - times tried - name - node_created:
  5505.   -0.000s - 0 - 1 - local_useless_elemwise_comparison - 0
  5506.   -0.000s - 0 - 1 - local_useless_switch - 0
  5507.   -0.000s - 0 - 1 - local_useless_elemwise - 0
  5508.   0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  5509.  
  5510.   0.000072s - ('ShapeOpt', 'ShapeOptimizer', 2, 1, 1) - 0.000s
  5511.   0.000056s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 0, 0) - 0.000s
  5512.     TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  5513.       nb_node (start, end, changed) (0, 0, 0)
  5514.       init io_toposort 8.10623168945e-06
  5515.       loop time 0.0
  5516.       callback_time 0.0
  5517.   0.000049s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 0, 0) - 0.000s
  5518.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  5519.       nb_node (start, end, changed) (0, 0, 0)
  5520.       init io_toposort 7.86781311035e-06
  5521.       loop time 9.53674316406e-07
  5522.       callback_time 0.0
  5523.   0.000048s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 0, 0) - 0.000s
  5524.     TopoOptimizer  local_inplace_incsubtensor1
  5525.       nb_node (start, end, changed) (0, 0, 0)
  5526.       init io_toposort 7.86781311035e-06
  5527.       loop time 0.0
  5528.       callback_time 0.0
  5529.   0.000047s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 0, 0) - 0.000s
  5530.     TopoOptimizer  local_inplace_sparse_block_gemv
  5531.       nb_node (start, end, changed) (0, 0, 0)
  5532.       init io_toposort 7.15255737305e-06
  5533.       loop time 0.0
  5534.       callback_time 0.0
  5535.   0.000047s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 0, 0) - 0.000s
  5536.     TopoOptimizer  local_inplace_setsubtensor
  5537.       nb_node (start, end, changed) (0, 0, 0)
  5538.       init io_toposort 8.10623168945e-06
  5539.       loop time 0.0
  5540.       callback_time 0.0
  5541.   0.000047s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 0, 0) - 0.000s
  5542.     TopoOptimizer  dimshuffle_as_view
  5543.       nb_node (start, end, changed) (0, 0, 0)
  5544.       init io_toposort 1.09672546387e-05
  5545.       loop time 9.53674316406e-07
  5546.       callback_time 0.0
  5547.   0.000046s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 0, 0) - 0.000s
  5548.     TopoOptimizer  local_gemm16_inplace
  5549.       nb_node (start, end, changed) (0, 0, 0)
  5550.       init io_toposort 6.91413879395e-06
  5551.       loop time 0.0
  5552.       callback_time 0.0
  5553.   0.000046s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 0, 0) - 0.000s
  5554.     TopoOptimizer  local_inplace_sparseblockouter
  5555.       nb_node (start, end, changed) (0, 0, 0)
  5556.       init io_toposort 6.91413879395e-06
  5557.       loop time 0.0
  5558.       callback_time 0.0
  5559.   0.000046s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 0, 0) - 0.000s
  5560.     TopoOptimizer  local_inplace_sparseblockgemv
  5561.       nb_node (start, end, changed) (0, 0, 0)
  5562.       init io_toposort 6.91413879395e-06
  5563.       loop time 0.0
  5564.       callback_time 0.0
  5565.   0.000046s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 0, 0) - 0.000s
  5566.     TopoOptimizer  local_inplace_sparse_block_outer
  5567.       nb_node (start, end, changed) (0, 0, 0)
  5568.       init io_toposort 6.91413879395e-06
  5569.       loop time 0.0
  5570.       callback_time 0.0
  5571.   0.000045s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 1, 1) - 0.000s
  5572.     TopoOptimizer  pre_local_IncSubtensor_serialize
  5573.       nb_node (start, end, changed) (1, 1, 0)
  5574.       init io_toposort 1.62124633789e-05
  5575.       loop time 6.91413879395e-06
  5576.       callback_time 0.0
  5577.   0.000041s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 0, 0) - 0.000s
  5578.   0.000040s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 0, 0) - 0.000s
  5579.     TopoOptimizer  local_advincsub1_gpua_inplace
  5580.       nb_node (start, end, changed) (0, 0, 0)
  5581.       init io_toposort 1.28746032715e-05
  5582.       loop time 0.0
  5583.       callback_time 0.0
  5584.   0.000034s - ('merge1.2', 'MergeOptimizer', 7, 0, 0) - 0.000s
  5585.     MergeOptimizer
  5586.       nb fail=    0 merged=    0 constant=    0
  5587.       time replace=0.00 validate=0.00 callback=0.00
  5588.   0.000033s - ('AbstractConvCheck', 'TopoOptimizer', 18, 0, 0) - 0.000s
  5589.     TopoOptimizer  AbstractConvCheck
  5590.       nb_node (start, end, changed) (0, 0, 0)
  5591.       init io_toposort 8.10623168945e-06
  5592.       loop time 0.0
  5593.       callback_time 0.0
  5594.   0.000032s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 0, 0) - 0.000s
  5595.   0.000031s - ('cond_make_inplace', 'TopoOptimizer', 47, 0, 0) - 0.000s
  5596.     TopoOptimizer  cond_make_inplace
  5597.       nb_node (start, end, changed) (0, 0, 0)
  5598.       init io_toposort 6.91413879395e-06
  5599.       loop time 0.0
  5600.       callback_time 0.0
  5601.   0.000031s - ('c_blas_destructive', 'TopoOptimizer', 37, 0, 0) - 0.000s
  5602.     TopoOptimizer  c_blas_destructive
  5603.       nb_node (start, end, changed) (0, 0, 0)
  5604.       init io_toposort 6.91413879395e-06
  5605.       loop time 0.0
  5606.       callback_time 0.0
  5607. LocalOptGroup
  5608. ---------------------
  5609. --- The Optimizer wasn't successful ---
  5610.  
  5611.    0.000030s - ('local_destructive', 'TopoOptimizer', 48, 0, 0) - 0.000s
  5612.      TopoOptimizer  CURAND_destructive
  5613.        nb_node (start, end, changed) (0, 0, 0)
  5614.        init io_toposort 6.91413879395e-06
  5615.        loop time 1.19209289551e-06
  5616.        callback_time 0.0
  5617.    0.000030s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 0, 0) - 0.000s
  5618.    0.000030s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 0, 0) - 0.000s
  5619.    0.000030s - ('make_ger_destructive', 'TopoOptimizer', 41, 0, 0) - 0.000s
  5620.      TopoOptimizer  make_scipy_blas_destructive
  5621.        nb_node (start, end, changed) (0, 0, 0)
  5622.        init io_toposort 6.91413879395e-06
  5623.        loop time 0.0
  5624.        callback_time 0.0
  5625.    0.000029s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 0, 0) - 0.000s
  5626.      TopoOptimizer  random_make_inplace_mrg
  5627.        nb_node (start, end, changed) (0, 0, 0)
  5628.        init io_toposort 5.96046447754e-06
  5629.        loop time 0.0
  5630.        callback_time 0.0
  5631.    0.000029s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 0, 0) - 0.000s
  5632.      TopoOptimizer  local_fill_to_alloc
  5633.        nb_node (start, end, changed) (0, 0, 0)
  5634.        init io_toposort 6.91413879395e-06
  5635.        loop time 0.0
  5636.        callback_time 0.0
  5637.    0.000029s - ('random_make_inplace', 'TopoOptimizer', 49, 0, 0) - 0.000s
  5638.      TopoOptimizer  random_make_inplace
  5639.        nb_node (start, end, changed) (0, 0, 0)
  5640.        init io_toposort 6.91413879395e-06
  5641.        loop time 0.0
  5642.        callback_time 0.0
  5643.    0.000026s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 0, 0) - 0.000s
  5644.      TopoOptimizer  local_elemwise_alloc
  5645.        nb_node (start, end, changed) (0, 0, 0)
  5646.        init io_toposort 5.96046447754e-06
  5647.        loop time 0.0
  5648.        callback_time 0.0
  5649.    0.000024s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 0, 0) - 0.000s
  5650.    0.000020s - ('merge1.1', 'MergeOptimizer', 4, 1, 1) - 0.000s
  5651.      MergeOptimizer
  5652.        nb fail=    0 merged=    0 constant=    0
  5653.        time replace=0.00 validate=0.00 callback=0.00
  5654.    0.000005s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 0, 0) - 0.000s
  5655.  
  5656. Here are tips to potentially make your code run faster
  5657.                  (if you think of new ones, suggest them on the mailing list).
  5658.                  Test them first, as they are not guaranteed to always provide a speedup.
  5659.   Sorry, no tip for today.
  5660. Function profiling
  5661. ==================
  5662.   Message: /Users/Ramana/projects/SBRNN/sb/utils.py:27
  5663.   Time in 1 calls to Function.__call__: 1.311302e-05s
  5664.   Time in Function.fn.__call__: 4.053116e-06s (30.909%)
  5665.   Time in thunks: 2.145767e-06s (16.364%)
  5666.   Total compile time: 2.555609e-02s
  5667.     Number of Apply nodes: 1
  5668.     Theano Optimizer time: 8.413076e-03s
  5669.        Theano validate time: 7.486343e-05s
  5670.     Theano Linker time (includes C, CUDA code generation/compiling): 5.640984e-04s
  5671.        Import time 0.000000e+00s
  5672.        Node make_thunk time 4.489422e-04s
  5673.  
  5674. Time in all call to theano.grad() 2.656322e+00s
  5675. Time since theano import 477.907s
  5676. Class
  5677. ---
  5678. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  5679.   100.0%   100.0%       0.000s       2.15e-06s     C        1       1   theano.compile.ops.DeepCopyOp
  5680.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  5681.  
  5682. Ops
  5683. ---
  5684. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  5685.   100.0%   100.0%       0.000s       2.15e-06s     C        1        1   DeepCopyOp
  5686.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  5687.  
  5688. Apply
  5689. ------
  5690. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  5691.   100.0%   100.0%       0.000s       2.15e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  5692.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  5693.  
  5694. Optimizer Profile
  5695. -----------------
  5696.  SeqOptimizer  OPT_FAST_RUN  time 0.008s for 1/0 nodes before/after optimization
  5697.    0.001s for callback
  5698.        0.000s for fgraph.validate()
  5699.    time      - (name, class, index, nodes before, nodes after) - validate time
  5700.    0.002406s - ('canonicalize', 'EquilibriumOptimizer', 6, 1, 0) - 0.000s
  5701.      EquilibriumOptimizer      canonicalize
  5702.        time 0.002s for 2 passes
  5703.        nb nodes (start, end,  max) 1 0 1
  5704.        time io_toposort 0.000s
  5705.        time in local optimizers 0.001s
  5706.        time in global optimizers 0.000s
  5707.        time in final optimizers 0.001s
  5708.        time in cleanup optimizers 0.000s
  5709.         0 - 0.002s 4 (0.000s in global opts, 0.000s io_toposort) - 1 nodes - ('topo_constant_folding', 1) ('local_upcast_elemwise_constant_inputs', 1) ('local_dimshuffle_lift', 1) ('MergeOptimizer', 1)
  5710.         1 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5711.        times - times applied - nb node created - name:
  5712.        0.001s - 1 - 3 - local_upcast_elemwise_constant_inputs
  5713.        0.001s - 1 - 0 - topo_constant_folding
  5714.        0.000s - 1 - 1 - local_dimshuffle_lift
  5715.        0.000s - 1 - 1 - MergeOptimizer
  5716.        0.000s - in 85 optimization that were not used (display only those with a runtime > 0)
  5717.          0.000s - local_fill_sink
  5718.          0.000s - local_useless_elemwise
  5719.          0.000s - local_func_inv
  5720.          0.000s - local_track_shape_i
  5721.          0.000s - local_fill_cut
  5722.          0.000s - local_merge_switch_same_cond
  5723.          0.000s - local_cast_cast
  5724.          0.000s - local_useless_elemwise_comparison
  5725.          0.000s - local_expm1
  5726.          0.000s - local_useless_switch
  5727.          0.000s - local_lift_transpose_through_dot
  5728.  
  5729.      Global, final and clean up optimizers
  5730.      Iter 0
  5731.              TopoOptimizer  topo_constant_folding
  5732.                nb_node (start, end, changed) (2, 0, 2)
  5733.                init io_toposort 2.69412994385e-05
  5734.                loop time 0.000415802001953
  5735.                callback_time 0.00018310546875
  5736.              MergeOptimizer
  5737.                nb fail=    0 merged=    1 constant=    0
  5738.                time replace=0.00 validate=0.00 callback=0.00
  5739.      Iter 1
  5740.              TopoOptimizer  topo_constant_folding
  5741.                nb_node (start, end, changed) (0, 0, 0)
  5742.                init io_toposort 9.05990600586e-06
  5743.                loop time 0.0
  5744.                callback_time 0.0
  5745.              MergeOptimizer
  5746.                nb fail=    0 merged=    0 constant=    0
  5747.                time replace=0.00 validate=0.00 callback=0.00
  5748.    0.000558s - ('gpuarray_opt', 'SeqOptimizer', 16, 0, 0) - 0.000s
  5749.      SeqOptimizer      gpuarray_opt  time 0.000s for 0/0 nodes before/after optimization
  5750.        0.000s for callback
  5751.            0.000s for fgraph.validate()
  5752.        0.000309s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 0, 0) - 0.000s
  5753.          EquilibriumOptimizer          gpuarray_local_optimizations
  5754.            time 0.000s for 1 passes
  5755.            nb nodes (start, end,  max) 0 0 0
  5756.            time io_toposort 0.000s
  5757.            time in local optimizers 0.000s
  5758.            time in global optimizers 0.000s
  5759.            time in final optimizers 0.000s
  5760.            time in cleanup optimizers 0.000s
  5761.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5762.        0.000071s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  5763.          EquilibriumOptimizer          gpuarray_cut_transfers
  5764.            time 0.000s for 1 passes
  5765.            nb nodes (start, end,  max) 0 0 0
  5766.            time io_toposort 0.000s
  5767.            time in local optimizers 0.000s
  5768.            time in global optimizers 0.000s
  5769.            time in final optimizers 0.000s
  5770.            time in cleanup optimizers 0.000s
  5771.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5772.        0.000049s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 0, 0) - 0.000s
  5773.          GraphToGPUOptimizer          gpuarray_graph_optimization
  5774.            time io_toposort 0.000s
  5775.          Total time taken by local optimizers 0.000s
  5776.        0.000008s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 0, 0) - 0.000s
  5777.  
  5778.    0.000496s - ('elemwise_fusion', 'SeqOptimizer', 19, 0, 0) - 0.000s
  5779.      SeqOptimizer      elemwise_fusion  time 0.000s for 0/0 nodes before/after optimization
  5780.        0.000s for callback
  5781.            0.000s for fgraph.validate()
  5782.        0.000190s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 0, 0) - 0.000s
  5783.          FusionOptimizer
  5784.           nb_iter 1
  5785.           nb_replacement 0
  5786.           nb_inconsistency_replace 0
  5787.           validate_time 0.0
  5788.           callback_time 0.0
  5789.           time_toposort 1.90734863281e-06
  5790.        0.000184s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 0, 0) - 0.000s
  5791.          FusionOptimizer
  5792.           nb_iter 1
  5793.           nb_replacement 0
  5794.           nb_inconsistency_replace 0
  5795.           validate_time 0.0
  5796.           callback_time 0.0
  5797.           time_toposort 9.53674316406e-07
  5798.  
  5799.    0.000471s - ('add_destroy_handler', 'AddDestroyHandler', 23, 0, 0) - 0.000s
  5800.    0.000442s - ('BlasOpt', 'SeqOptimizer', 12, 0, 0) - 0.000s
  5801.      SeqOptimizer      BlasOpt  time 0.000s for 0/0 nodes before/after optimization
  5802.        0.000s for callback
  5803.            0.000s for fgraph.validate()
  5804.        0.000139s - ('gemm_optimizer', 'GemmOptimizer', 1, 0, 0) - 0.000s
  5805.          GemmOptimizer
  5806.           nb_iter 1
  5807.           nb_replacement 0
  5808.           nb_replacement_didn_t_remove 0
  5809.           nb_inconsistency_make 0
  5810.           nb_inconsistency_replace 0
  5811.           time_canonicalize 0
  5812.           time_factor_can 0
  5813.           time_factor_list 0
  5814.           time_toposort 6.91413879395e-06
  5815.           validate_time 0.0
  5816.           callback_time 0.0
  5817.        0.000094s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 0, 0) - 0.000s
  5818.          EquilibriumOptimizer          local_gemm_to_gemv
  5819.            time 0.000s for 1 passes
  5820.            nb nodes (start, end,  max) 0 0 0
  5821.            time io_toposort 0.000s
  5822.            time in local optimizers 0.000s
  5823.            time in global optimizers 0.000s
  5824.            time in final optimizers 0.000s
  5825.            time in cleanup optimizers 0.000s
  5826.             0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5827.        0.000030s - ('use_c_blas', 'TopoOptimizer', 4, 0, 0) - 0.000s
  5828.          TopoOptimizer  use_c_blas
  5829.            nb_node (start, end, changed) (0, 0, 0)
  5830.            init io_toposort 7.15255737305e-06
  5831.            loop time 1.19209289551e-06
  5832.            callback_time 0.0
  5833.  LocalOptGroup
  5834.  ---------------------
  5835.  --- The Optimizer wasn't successful ---
  5836.  
  5837.       0.000029s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 0, 0) - 0.000s
  5838.         TopoOptimizer  local_dot22_to_dot22scalar
  5839.           nb_node (start, end, changed) (0, 0, 0)
  5840.           init io_toposort 6.91413879395e-06
  5841.           loop time 0.0
  5842.           callback_time 0.0
  5843.       0.000029s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 0, 0) - 0.000s
  5844.         TopoOptimizer  local_dot_to_dot22
  5845.           nb_node (start, end, changed) (0, 0, 0)
  5846.           init io_toposort 7.86781311035e-06
  5847.           loop time 0.0
  5848.           callback_time 0.0
  5849.       0.000027s - ('use_scipy_ger', 'TopoOptimizer', 5, 0, 0) - 0.000s
  5850.         TopoOptimizer  scipy_blas
  5851.           nb_node (start, end, changed) (0, 0, 0)
  5852.           init io_toposort 6.91413879395e-06
  5853.           loop time 0.0
  5854.           callback_time 0.0
  5855.  
  5856.   0.000393s - ('stabilize', 'EquilibriumOptimizer', 8, 0, 0) - 0.000s
  5857.     EquilibriumOptimizer      stabilize
  5858.       time 0.000s for 1 passes
  5859.       nb nodes (start, end,  max) 0 0 0
  5860.       time io_toposort 0.000s
  5861.       time in local optimizers 0.000s
  5862.       time in global optimizers 0.000s
  5863.       time in final optimizers 0.000s
  5864.       time in cleanup optimizers 0.000s
  5865.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5866.     Global, final and clean up optimizers
  5867.     Iter 0
  5868.             TopoOptimizer  topo_constant_folding
  5869.               nb_node (start, end, changed) (0, 0, 0)
  5870.               init io_toposort 9.05990600586e-06
  5871.               loop time 0.0
  5872.               callback_time 0.0
  5873.   0.000374s - ('specialize', 'EquilibriumOptimizer', 13, 0, 0) - 0.000s
  5874.     EquilibriumOptimizer      specialize
  5875.       time 0.000s for 1 passes
  5876.       nb nodes (start, end,  max) 0 0 0
  5877.       time io_toposort 0.000s
  5878.       time in local optimizers 0.000s
  5879.       time in global optimizers 0.000s
  5880.       time in final optimizers 0.000s
  5881.       time in cleanup optimizers 0.000s
  5882.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5883.     Global, final and clean up optimizers
  5884.     Iter 0
  5885.             TopoOptimizer  topo_constant_folding
  5886.               nb_node (start, end, changed) (0, 0, 0)
  5887.               init io_toposort 6.91413879395e-06
  5888.               loop time 0.0
  5889.               callback_time 0.0
  5890.   0.000254s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 0, 0) - 0.000s
  5891.     EquilibriumOptimizer      scan_eqopt2
  5892.       time 0.000s for 1 passes
  5893.       nb nodes (start, end,  max) 0 0 0
  5894.       time io_toposort 0.000s
  5895.       time in local optimizers 0.000s
  5896.       time in global optimizers 0.000s
  5897.       time in final optimizers 0.000s
  5898.       time in cleanup optimizers 0.000s
  5899.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5900.     Global, final and clean up optimizers
  5901.     Iter 0
  5902.             TopoOptimizer  constant_folding_for_scan2
  5903.               nb_node (start, end, changed) (0, 0, 0)
  5904.               init io_toposort 8.10623168945e-06
  5905.               loop time 9.53674316406e-07
  5906.               callback_time 0.0
  5907.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  5908.               nb_node (start, end, changed) (0, 0, 0)
  5909.               init io_toposort 6.91413879395e-06
  5910.               loop time 0.0
  5911.               callback_time 0.0
  5912.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  5913.               nb_node (start, end, changed) (0, 0, 0)
  5914.               init io_toposort 5.96046447754e-06
  5915.               loop time 0.0
  5916.               callback_time 0.0
  5917.             TopoOptimizer  scanOp_merge_inouts
  5918.               nb_node (start, end, changed) (0, 0, 0)
  5919.               init io_toposort 5.96046447754e-06
  5920.               loop time 0.0
  5921.               callback_time 0.0
  5922.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  5923.               nb_node (start, end, changed) (0, 0, 0)
  5924.               init io_toposort 7.15255737305e-06
  5925.               loop time 9.53674316406e-07
  5926.               callback_time 0.0
  5927.   0.000218s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 1, 1) - 0.000s
  5928.     EquilibriumOptimizer      scan_eqopt1
  5929.       time 0.000s for 1 passes
  5930.       nb nodes (start, end,  max) 1 1 1
  5931.       time io_toposort 0.000s
  5932.       time in local optimizers 0.000s
  5933.       time in global optimizers 0.000s
  5934.       time in final optimizers 0.000s
  5935.       time in cleanup optimizers 0.000s
  5936.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 1 nodes -
  5937.     Global, final and clean up optimizers
  5938.     Iter 0
  5939.             SeqOptimizer              all_pushout_opt  time 0.000s for 1/1 nodes before/after optimization
  5940.               0.000s for callback
  5941.                   0.000s for fgraph.validate()
  5942.               0.000061s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 1, 1) - 0.000s
  5943.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  5944.                   nb_node (start, end, changed) (1, 1, 0)
  5945.                   init io_toposort 2.40802764893e-05
  5946.                   loop time 4.05311584473e-06
  5947.                   callback_time 0.0
  5948.               0.000015s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 1, 1) - 0.000s
  5949.               0.000009s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 1, 1) - 0.000s
  5950.               0.000008s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 1, 1) - 0.000s
  5951.               0.000008s - ('scan_pushout_dot1', 'PushOutDot1', 3, 1, 1) - 0.000s
  5952.  
  5953.   0.000185s - ('merge3', 'MergeOptimizer', 51, 0, 0) - 0.000s
  5954.     MergeOptimizer
  5955.       nb fail=    0 merged=    0 constant=    0
  5956.       time replace=0.00 validate=0.00 callback=0.00
  5957.   0.000181s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 0, 0) - 0.000s
  5958.     FusionOptimizer
  5959.      nb_iter 1
  5960.      nb_replacement 0
  5961.      nb_inconsistency_replace 0
  5962.      validate_time 0.0
  5963.      callback_time 0.0
  5964.      time_toposort 9.53674316406e-07
  5965.   0.000180s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 0, 0) - 0.000s
  5966.     FusionOptimizer
  5967.      nb_iter 1
  5968.      nb_replacement 0
  5969.      nb_inconsistency_replace 0
  5970.      validate_time 0.0
  5971.      callback_time 0.0
  5972.      time_toposort 9.53674316406e-07
  5973.   0.000127s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 0, 0) - 0.000s
  5974.     EquilibriumOptimizer      uncanonicalize
  5975.       time 0.000s for 1 passes
  5976.       nb nodes (start, end,  max) 0 0 0
  5977.       time io_toposort 0.000s
  5978.       time in local optimizers 0.000s
  5979.       time in global optimizers 0.000s
  5980.       time in final optimizers 0.000s
  5981.       time in cleanup optimizers 0.000s
  5982.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  5983.     Global, final and clean up optimizers
  5984.     Iter 0
  5985.             TopoOptimizer  topo_constant_folding
  5986.               nb_node (start, end, changed) (0, 0, 0)
  5987.               init io_toposort 7.15255737305e-06
  5988.               loop time 0.0
  5989.               callback_time 0.0
  5990.   0.000120s - ('merge2', 'MergeOptimizer', 22, 0, 0) - 0.000s
  5991.     MergeOptimizer
  5992.       nb fail=    0 merged=    0 constant=    0
  5993.       time replace=0.00 validate=0.00 callback=0.00
  5994.   0.000106s - ('merge1', 'MergeOptimizer', 0, 1, 1) - 0.000s
  5995.     MergeOptimizer
  5996.       nb fail=    0 merged=    0 constant=    0
  5997.       time replace=0.00 validate=0.00 callback=0.00
  5998.   0.000083s - ('specialize_device', 'EquilibriumOptimizer', 17, 0, 0) - 0.000s
  5999.     EquilibriumOptimizer      specialize_device
  6000.       time 0.000s for 1 passes
  6001.       nb nodes (start, end,  max) 0 0 0
  6002.       time io_toposort 0.000s
  6003.       time in local optimizers 0.000s
  6004.       time in global optimizers 0.000s
  6005.       time in final optimizers 0.000s
  6006.       time in cleanup optimizers 0.000s
  6007.        0 - 0.000s 0 (0.000s in global opts, 0.000s io_toposort) - 0 nodes -
  6008.   0.000081s - ('ShapeOpt', 'ShapeOptimizer', 2, 1, 1) - 0.000s
  6009.   0.000080s - ('blas_opt_inplace', 'TopoOptimizer', 34, 0, 0) - 0.000s
  6010.     TopoOptimizer  InplaceBlasOpt
  6011.       nb_node (start, end, changed) (0, 0, 0)
  6012.       init io_toposort 1.09672546387e-05
  6013.       loop time 9.53674316406e-07
  6014.       callback_time 0.0
  6015. LocalOptGroup
  6016. ---------------------
  6017. --- The Optimizer wasn't successful ---
  6018.  
  6019.    0.000080s - ('useless', 'TopoOptimizer', 3, 1, 1) - 0.000s
  6020.      TopoOptimizer  useless
  6021.        nb_node (start, end, changed) (1, 1, 0)
  6022.        init io_toposort 1.90734863281e-05
  6023.        loop time 3.09944152832e-05
  6024.        callback_time 0.0
  6025.  LocalOptGroup
  6026.  ---------------------
  6027.    time taken - times applied - times tried - name - node_created:
  6028.    -0.000s - 0 - 1 - local_useless_switch - 0
  6029.    -0.000s - 0 - 1 - local_useless_elemwise_comparison - 0
  6030.    -0.000s - 0 - 1 - local_useless_elemwise - 0
  6031.    0.000s - in 16 optimization that were not used (display those with runtime greater than 0)
  6032.  
  6033.    0.000077s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 0, 0) - 0.000s
  6034.      TopoOptimizer  local_dnn_conv_inplace
  6035.        nb_node (start, end, changed) (0, 0, 0)
  6036.        init io_toposort 6.91413879395e-06
  6037.        loop time 1.19209289551e-06
  6038.        callback_time 0.0
  6039.  LocalOptGroup
  6040.  ---------------------
  6041.  --- The Optimizer wasn't successful ---
  6042.  
  6043.   0.000077s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 0, 0) - 0.000s
  6044.     TopoOptimizer  InplaceGpuaBlasOpt
  6045.       nb_node (start, end, changed) (0, 0, 0)
  6046.       init io_toposort 6.91413879395e-06
  6047.       loop time 0.0
  6048.       callback_time 0.0
  6049. LocalOptGroup
  6050. ---------------------
  6051.   time taken - times applied - times tried - name - node_created:
  6052.   -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  6053.   0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  6054.  
  6055.   0.000077s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 0, 0) - 0.000s
  6056.     TopoOptimizer  InplaceGpuBlasOpt
  6057.       nb_node (start, end, changed) (0, 0, 0)
  6058.       init io_toposort 7.15255737305e-06
  6059.       loop time 0.0
  6060.       callback_time 0.0
  6061. LocalOptGroup
  6062. ---------------------
  6063. --- The Optimizer wasn't successful ---
  6064.  
  6065.    0.000076s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 0, 0) - 0.000s
  6066.      TopoOptimizer  local_dnna_conv_inplace
  6067.        nb_node (start, end, changed) (0, 0, 0)
  6068.        init io_toposort 7.86781311035e-06
  6069.        loop time 1.19209289551e-06
  6070.        callback_time 0.0
  6071.  LocalOptGroup
  6072.  ---------------------
  6073.    time taken - times applied - times tried - name - node_created:
  6074.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  6075.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  6076.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  6077.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  6078.  
  6079.    0.000052s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 0, 0) - 0.000s
  6080.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  6081.        nb_node (start, end, changed) (0, 0, 0)
  6082.        init io_toposort 8.10623168945e-06
  6083.        loop time 0.0
  6084.        callback_time 0.0
  6085.    0.000048s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 0, 0) - 0.000s
  6086.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  6087.        nb_node (start, end, changed) (0, 0, 0)
  6088.        init io_toposort 7.86781311035e-06
  6089.        loop time 0.0
  6090.        callback_time 0.0
  6091.    0.000047s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 1, 1) - 0.000s
  6092.      TopoOptimizer  pre_local_IncSubtensor_serialize
  6093.        nb_node (start, end, changed) (1, 1, 0)
  6094.        init io_toposort 1.71661376953e-05
  6095.        loop time 7.86781311035e-06
  6096.        callback_time 0.0
  6097.    0.000046s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 0, 0) - 0.000s
  6098.      TopoOptimizer  local_inplace_sparseblockouter
  6099.        nb_node (start, end, changed) (0, 0, 0)
  6100.        init io_toposort 7.15255737305e-06
  6101.        loop time 0.0
  6102.        callback_time 0.0
  6103.    0.000046s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 0, 0) - 0.000s
  6104.      TopoOptimizer  local_inplace_sparseblockgemv
  6105.        nb_node (start, end, changed) (0, 0, 0)
  6106.        init io_toposort 7.15255737305e-06
  6107.        loop time 0.0
  6108.        callback_time 0.0
  6109.    0.000046s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 0, 0) - 0.000s
  6110.      TopoOptimizer  local_inplace_setsubtensor
  6111.        nb_node (start, end, changed) (0, 0, 0)
  6112.        init io_toposort 7.15255737305e-06
  6113.        loop time 0.0
  6114.        callback_time 0.0
  6115.    0.000046s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 0, 0) - 0.000s
  6116.      TopoOptimizer  local_inplace_incsubtensor1
  6117.        nb_node (start, end, changed) (0, 0, 0)
  6118.        init io_toposort 7.15255737305e-06
  6119.        loop time 0.0
  6120.        callback_time 0.0
  6121.    0.000045s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 0, 0) - 0.000s
  6122.      TopoOptimizer  local_gemm16_inplace
  6123.        nb_node (start, end, changed) (0, 0, 0)
  6124.        init io_toposort 6.91413879395e-06
  6125.        loop time 0.0
  6126.        callback_time 0.0
  6127.    0.000045s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 0, 0) - 0.000s
  6128.      TopoOptimizer  local_inplace_sparse_block_gemv
  6129.        nb_node (start, end, changed) (0, 0, 0)
  6130.        init io_toposort 6.91413879395e-06
  6131.        loop time 0.0
  6132.        callback_time 0.0
  6133.    0.000045s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 0, 0) - 0.000s
  6134.      TopoOptimizer  local_inplace_sparse_block_outer
  6135.        nb_node (start, end, changed) (0, 0, 0)
  6136.        init io_toposort 7.15255737305e-06
  6137.        loop time 0.0
  6138.        callback_time 0.0
  6139.    0.000040s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 0, 0) - 0.000s
  6140.      TopoOptimizer  dimshuffle_as_view
  6141.        nb_node (start, end, changed) (0, 0, 0)
  6142.        init io_toposort 1.19209289551e-05
  6143.        loop time 9.53674316406e-07
  6144.        callback_time 0.0
  6145.    0.000038s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 0, 0) - 0.000s
  6146.    0.000036s - ('merge1.2', 'MergeOptimizer', 7, 0, 0) - 0.000s
  6147.      MergeOptimizer
  6148.        nb fail=    0 merged=    0 constant=    0
  6149.        time replace=0.00 validate=0.00 callback=0.00
  6150.    0.000035s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 0, 0) - 0.000s
  6151.    0.000033s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 0, 0) - 0.000s
  6152.    0.000032s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 0, 0) - 0.000s
  6153.      TopoOptimizer  local_advincsub1_gpua_inplace
  6154.        nb_node (start, end, changed) (0, 0, 0)
  6155.        init io_toposort 7.86781311035e-06
  6156.        loop time 1.19209289551e-06
  6157.        callback_time 0.0
  6158.    0.000031s - ('local_destructive', 'TopoOptimizer', 48, 0, 0) - 0.000s
  6159.      TopoOptimizer  CURAND_destructive
  6160.        nb_node (start, end, changed) (0, 0, 0)
  6161.        init io_toposort 7.15255737305e-06
  6162.        loop time 0.0
  6163.        callback_time 0.0
  6164.    0.000031s - ('cond_make_inplace', 'TopoOptimizer', 47, 0, 0) - 0.000s
  6165.      TopoOptimizer  cond_make_inplace
  6166.        nb_node (start, end, changed) (0, 0, 0)
  6167.        init io_toposort 7.15255737305e-06
  6168.        loop time 9.53674316406e-07
  6169.        callback_time 0.0
  6170.    0.000031s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 0, 0) - 0.000s
  6171.      TopoOptimizer  local_fill_to_alloc
  6172.        nb_node (start, end, changed) (0, 0, 0)
  6173.        init io_toposort 9.05990600586e-06
  6174.        loop time 0.0
  6175.        callback_time 0.0
  6176.    0.000030s - ('c_blas_destructive', 'TopoOptimizer', 37, 0, 0) - 0.000s
  6177.      TopoOptimizer  c_blas_destructive
  6178.        nb_node (start, end, changed) (0, 0, 0)
  6179.        init io_toposort 5.96046447754e-06
  6180.        loop time 0.0
  6181.        callback_time 0.0
  6182.  LocalOptGroup
  6183.  ---------------------
  6184.  --- The Optimizer wasn't successful ---
  6185.  
  6186.   0.000029s - ('make_ger_destructive', 'TopoOptimizer', 41, 0, 0) - 0.000s
  6187.     TopoOptimizer  make_scipy_blas_destructive
  6188.       nb_node (start, end, changed) (0, 0, 0)
  6189.       init io_toposort 5.96046447754e-06
  6190.       loop time 0.0
  6191.       callback_time 0.0
  6192.   0.000029s - ('AbstractConvCheck', 'TopoOptimizer', 18, 0, 0) - 0.000s
  6193.     TopoOptimizer  AbstractConvCheck
  6194.       nb_node (start, end, changed) (0, 0, 0)
  6195.       init io_toposort 7.15255737305e-06
  6196.       loop time 9.53674316406e-07
  6197.       callback_time 0.0
  6198.   0.000028s - ('random_make_inplace', 'TopoOptimizer', 49, 0, 0) - 0.000s
  6199.     TopoOptimizer  random_make_inplace
  6200.       nb_node (start, end, changed) (0, 0, 0)
  6201.       init io_toposort 6.91413879395e-06
  6202.       loop time 0.0
  6203.       callback_time 0.0
  6204.   0.000028s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 0, 0) - 0.000s
  6205.   0.000028s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 0, 0) - 0.000s
  6206.     TopoOptimizer  random_make_inplace_mrg
  6207.       nb_node (start, end, changed) (0, 0, 0)
  6208.       init io_toposort 6.19888305664e-06
  6209.       loop time 9.53674316406e-07
  6210.       callback_time 0.0
  6211.   0.000027s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 0, 0) - 0.000s
  6212.     TopoOptimizer  local_elemwise_alloc
  6213.       nb_node (start, end, changed) (0, 0, 0)
  6214.       init io_toposort 6.91413879395e-06
  6215.       loop time 0.0
  6216.       callback_time 0.0
  6217.   0.000023s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 0, 0) - 0.000s
  6218.   0.000021s - ('merge1.1', 'MergeOptimizer', 4, 1, 1) - 0.000s
  6219.     MergeOptimizer
  6220.       nb fail=    0 merged=    0 constant=    0
  6221.       time replace=0.00 validate=0.00 callback=0.00
  6222.   0.000006s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 0, 0) - 0.000s
  6223.  
  6224. Here are tips to potentially make your code run faster
  6225.                 (if you think of new ones, suggest them on the mailing list).
  6226.                 Test them first, as they are not guaranteed to always provide a speedup.
  6227.  Sorry, no tip for today.
  6228. Function profiling
  6229. ==================
  6230.  Message: sb/convnet/sb_resnet.py:327
  6231.  Time in 0 calls to Function.__call__: 0.000000e+00s
  6232.  Total compile time: 4.351297e+02s
  6233.    Number of Apply nodes: 2957
  6234.    Theano Optimizer time: 6.785364e+01s
  6235.       Theano validate time: 9.809670e+00s
  6236.    Theano Linker time (includes C, CUDA code generation/compiling): 3.640262e+02s
  6237.       Import time 1.506580e+00s
  6238.       Node make_thunk time 3.635044e+02s
  6239.  
  6240. Time in all call to theano.grad() 2.656322e+00s
  6241. Time since theano import 477.913s
  6242. Optimizer Profile
  6243. -----------------
  6244. SeqOptimizer  OPT_FAST_RUN  time 67.853s for 10599/2957 nodes before/after optimization
  6245.   22.259s for callback
  6246.       9.810s for fgraph.validate()
  6247.   callbacks_time
  6248.        <theano.gof.destroyhandler.DestroyHandler object at 0x1291a3f10> , 7.663418293
  6249.        <theano.tensor.opt.ShapeFeature object at 0x121562590> , 4.24110126495
  6250.        Updater{canonicalize} , 3.05698680878
  6251.        <theano.compile.function_module.Supervisor instance at 0x1227d3f38> , 2.56386876106
  6252.        <theano.gof.opt.MergeFeature object at 0x1227e7190> , 1.74654364586
  6253.        Updater{canonicalize} , 1.31303954124
  6254.        Updater{gpuarray_local_optimizations} , 0.261435270309
  6255.        Updater{gpuarray_cut_transfers} , 0.261384248734
  6256.        Updater{canonicalize} , 0.194210767746
  6257.        Updater{specialize} , 0.130860090256
  6258.        <theano.gof.toolbox.ReplaceValidate object at 0x12087ee50> , 0.11890411377
  6259.        <theano.gof.toolbox.PreserveVariableAttributes object at 0x122195850> , 0.0774388313293
  6260.        Updater{canonicalize} , 0.0659112930298
  6261.        <theano.gof.opt.ChangeTracker instance at 0x11f7d8f38> , 0.0626258850098
  6262.        Updater{canonicalize} , 0.0376682281494
  6263.        Updater{canonicalize} , 0.0349521636963
  6264.        Updater{gpuarray_local_optimizations} , 0.0294797420502
  6265.        Updater{specialize} , 0.026261806488
  6266.        <theano.gof.opt.ChangeTracker instance at 0x125a3e1b8> , 0.012256860733
  6267.        Updater{local_elemwise_alloc} , 0.00888752937317
  6268.        <theano.gof.opt.ChangeTracker instance at 0x124d8a098> , 0.00651884078979
  6269.        Updater{pre_local_IncSubtensor_serialize} , 0.00562787055969
  6270.        Updater{specialize} , 0.00509214401245
  6271.        Updater{dimshuffle_as_view} , 0.00395131111145
  6272.        Updater{topo_constant_folding} , 0.00292682647705
  6273.        Updater{useless} , 0.00235295295715
  6274.        <theano.gof.opt.ChangeTracker instance at 0x127399f38> , 0.00211262702942
  6275.        Updater{local_inplace_setsubtensor} , 0.0013701915741
  6276.        Updater{local_dnna_conv_inplace} , 0.00122761726379
  6277.        Updater{specialize} , 0.00121068954468
  6278.        Updater{constant_folding_for_scan2} , 0.000903367996216
  6279.        Updater{stabilize} , 0.000802278518677
  6280.        Updater{specialize} , 0.000703573226929
  6281.        Updater{topo_constant_folding} , 0.000659465789795
  6282.        <theano.gof.opt.ChangeTracker instance at 0x124bccfc8> , 0.000591039657593
  6283.        Updater{specialize} , 0.000572443008423
  6284.        Updater{GemmOptimizer} , 0.000359296798706
  6285.        Updater{local_dot_to_dot22} , 0.000351905822754
  6286.        Updater{topo_constant_folding} , 0.000331401824951
  6287.        <theano.gof.opt.ChangeTracker instance at 0x1226b5560> , 0.000254154205322
  6288.        Updater{random_make_inplace_mrg} , 0.000191688537598
  6289.        Updater{topo_constant_folding} , 0.000139713287354
  6290.        Updater{InplaceGpuaBlasOpt} , 9.91821289062e-05
  6291.        Updater{topo_constant_folding} , 6.60419464111e-05
  6292.        Updater{specialize} , 4.02927398682e-05
  6293.        Updater{specialize} , 2.93254852295e-05
  6294.        Updater{specialize} , 2.76565551758e-05
  6295.        Updater{specialize} , 2.55107879639e-05
  6296.        Updater{local_dot22_to_dot22scalar} , 2.121925354e-05
  6297.        Updater{topo_constant_folding} , 1.52587890625e-05
  6298.        Updater{topo_constant_folding} , 1.43051147461e-05
  6299.        Updater{topo_constant_folding} , 1.38282775879e-05
  6300.        Updater{topo_constant_folding} , 1.28746032715e-05
  6301.        Updater{topo_constant_folding} , 8.82148742676e-06
  6302.        Updater{topo_constant_folding} , 6.67572021484e-06
  6303.        Updater{gpuarray_local_optimizations} , 5.72204589844e-06
  6304.   time      - (name, class, index, nodes before, nodes after) - validate time
  6305.   19.936074s - ('canonicalize', 'EquilibriumOptimizer', 6, 7639, 4559) - 0.143s
  6306.     EquilibriumOptimizer      canonicalize
  6307.       time 19.936s for 7 passes
  6308.       nb nodes (start, end,  max) 7639 4559 7639
  6309.       time io_toposort 1.026s
  6310.       time in local optimizers 12.701s
  6311.       time in global optimizers 0.000s
  6312.       time in final optimizers 1.199s
  6313.       time in cleanup optimizers 4.571s
  6314.        0 - 10.519s 4693 (0.420s in global opts, 0.381s io_toposort) - 7635 nodes - ('MergeOptimizer', 1723) ('local_useless_fill', 643) ('local_mul_canonizer', 358) ('local_fill_sink', 315) ('local_neg_to_mul', 306) ...
  6315.        1 - 4.410s 1690 (0.136s in global opts, 0.426s io_toposort) - 6297 nodes - ('MergeOptimizer', 630) ('local_dimshuffle_lift', 216) ('local_mul_canonizer', 206) ('local_fill_sink', 203) ('local_upcast_elemwise_constant_inputs', 131) ...
  6316.        2 - 1.501s 497 (0.055s in global opts, 0.045s io_toposort) - 4794 nodes - ('MergeOptimizer', 142) ('local_fill_sink', 114) ('local_useless_fill', 57) ('local_zero_div', 57) ('local_sum_prod_div_dimshuffle', 56) ...
  6317.        3 - 0.844s 123 (0.049s in global opts, 0.043s io_toposort) - 4568 nodes - ('MergeOptimizer', 60) ('local_dimshuffle_lift', 56) ('local_useless_fill', 3) ('local_mul_zero', 3) ('topo_constant_folding', 1)
  6318.        4 - 0.762s 67 (0.048s in global opts, 0.041s io_toposort) - 4568 nodes - ('MergeOptimizer', 32) ('local_sum_prod_div_dimshuffle', 28) ('local_zero_div', 3) ('local_fill_sink', 3) ('topo_constant_folding', 1)
  6319.        5 - 1.199s 56 (0.441s in global opts, 0.047s io_toposort) - 4559 nodes - ('local_dimshuffle_lift', 28) ('MergeOptimizer', 28)
  6320.        6 - 0.702s 0 (0.051s in global opts, 0.044s io_toposort) - 4559 nodes -
  6321.       times - times applied - nb node created - name:
  6322.       4.571s - 2615 - 17 - MergeOptimizer
  6323.       2.217s - 565 - 1522 - local_mul_canonizer
  6324.       2.124s - 186 - 975 - local_greedy_distributor
  6325.       1.672s - 635 - 790 - local_fill_sink
  6326.       1.520s - 544 - 1738 - local_dimshuffle_lift
  6327.       1.199s - 5 - 0 - topo_constant_folding
  6328.       0.804s - 260 - 1317 - local_mul_zero
  6329.       0.773s - 251 - 401 - local_add_canonizer
  6330.       0.376s - 799 - 0 - local_useless_fill
  6331.       0.335s - 137 - 411 - local_upcast_elemwise_constant_inputs
  6332.       0.330s - 306 - 606 - local_neg_to_mul
  6333.       0.327s - 168 - 504 - local_sum_prod_div_dimshuffle
  6334.       0.169s - 134 - 0 - local_cut_gpu_transfers
  6335.       0.159s - 3 - 3 - local_useless_elemwise
  6336.       0.143s - 12 - 24 - local_reshape_to_dimshuffle
  6337.       0.141s - 14 - 28 - local_subtensor_merge
  6338.       0.138s - 146 - 449 - local_shape_to_shape_i
  6339.       0.135s - 90 - 180 - local_zero_div
  6340.       0.117s - 36 - 108 - local_mul_switch_sink
  6341.       0.088s - 33 - 99 - local_div_switch_sink
  6342.       0.073s - 16 - 9 - local_useless_switch
  6343.       0.065s - 9 - 0 - local_join_1
  6344.       0.049s - 9 - 9 - local_useless_dimshuffle_in_reshape
  6345.       0.047s - 81 - 32 - local_subtensor_make_vector
  6346.       0.035s - 31 - 62 - local_inv_canon
  6347.       0.020s - 19 - 0 - local_pow_canonicalize
  6348.       0.014s - 10 - 20 - local_subtensor_lift
  6349.       0.010s - 12 - 0 - local_intdiv_by_one
  6350.       0.820s - in 61 optimization that were not used (display only those with a runtime > 0)
  6351.         0.179s - local_func_inv
  6352.         0.113s - local_one_minus_erf2
  6353.         0.093s - local_merge_switch_same_cond
  6354.         0.089s - local_useless_elemwise_comparison
  6355.         0.062s - local_track_shape_i
  6356.         0.053s - local_fill_cut
  6357.         0.044s - local_expm1
  6358.         0.039s - local_cast_cast
  6359.         0.032s - local_IncSubtensor_serialize
  6360.         0.030s - local_one_minus_erf
  6361.         0.015s - local_useless_subtensor
  6362.         0.010s - local_sum_prod_all_to_none
  6363.         0.007s - local_lift_transpose_through_dot
  6364.         0.007s - local_op_of_op
  6365.         0.006s - local_useless_reduce
  6366.         0.006s - local_useless_slice
  6367.         0.005s - local_sumsqr2dot
  6368.         0.004s - local_reduce_join
  6369.         0.004s - local_dimshuffle_no_inplace_at_canonicalize
  6370.         0.004s - f
  6371.         0.004s - local_subtensor_remove_broadcastable_index
  6372.         0.002s - local_0_dot_x
  6373.         0.001s - local_abs_lift
  6374.         0.001s - local_incsubtensor_of_zeros
  6375.         0.001s - local_useless_reshape
  6376.         0.001s - local_subtensor_of_dot
  6377.         0.001s - local_subtensor_of_alloc
  6378.         0.001s - local_reshape_lift
  6379.         0.001s - local_useless_inc_subtensor
  6380.         0.001s - local_canonicalize_alloc
  6381.         0.000s - local_useless_inc_subtensor_alloc
  6382.         0.000s - local_useless_alloc
  6383.         0.000s - local_setsubtensor_of_constants
  6384.         0.000s - local_merge_alloc
  6385.         0.000s - local_scalar_tensor_scalar
  6386.  
  6387.     Global, final and clean up optimizers
  6388.     Iter 0
  6389.             TopoOptimizer  topo_constant_folding
  6390.               nb_node (start, end, changed) (6389, 6305, 84)
  6391.               init io_toposort 0.0622019767761
  6392.               loop time 0.357506990433
  6393.               callback_time 0.245550394058
  6394.             MergeOptimizer
  6395.               nb fail=    0 merged= 4140 constant= 1491
  6396.               time replace=3.41 validate=0.05 callback=3.05
  6397.               callbacks_time
  6398.                    <theano.gof.toolbox.PreserveVariableAttributes object at 0x122195850> , 0.00784587860107
  6399.                    <theano.gof.opt.ChangeTracker instance at 0x11f7d8f38> , 0.0107350349426
  6400.                    <theano.compile.function_module.Supervisor instance at 0x1227d3f38> , 0.0117847919464
  6401.                    <theano.gof.toolbox.ReplaceValidate object at 0x12087ee50> , 0.0135762691498
  6402.                    <theano.gof.opt.MergeFeature object at 0x1227e7190> , 0.25992846489
  6403.                    <theano.tensor.opt.ShapeFeature object at 0x121562590> , 0.968868017197
  6404.                    Updater{canonicalize} , 1.68857598305
  6405.     Iter 1
  6406.             TopoOptimizer  topo_constant_folding
  6407.               nb_node (start, end, changed) (4829, 4794, 35)
  6408.               init io_toposort 0.0512568950653
  6409.               loop time 0.083508014679
  6410.               callback_time 0.0425992012024
  6411.             MergeOptimizer
  6412.               nb fail=    0 merged= 1652 constant=  351
  6413.               time replace=0.88 validate=0.02 callback=0.73
  6414.     Iter 2
  6415.             TopoOptimizer  topo_constant_folding
  6416.               nb_node (start, end, changed) (4570, 4568, 2)
  6417.               init io_toposort 0.0442609786987
  6418.               loop time 0.0108721256256
  6419.               callback_time 0.004061460495
  6420.             MergeOptimizer
  6421.               nb fail=    0 merged=  350 constant=   86
  6422.               time replace=0.18 validate=0.00 callback=0.14
  6423.     Iter 3
  6424.             TopoOptimizer  topo_constant_folding
  6425.               nb_node (start, end, changed) (4569, 4568, 1)
  6426.               init io_toposort 0.0423350334167
  6427.               loop time 0.00659894943237
  6428.               callback_time 0.000518798828125
  6429.             MergeOptimizer
  6430.               nb fail=    0 merged=   62 constant=    4
  6431.               time replace=0.04 validate=0.00 callback=0.04
  6432.     Iter 4
  6433.             TopoOptimizer  topo_constant_folding
  6434.               nb_node (start, end, changed) (4563, 4562, 1)
  6435.               init io_toposort 0.0421988964081
  6436.               loop time 0.00592303276062
  6437.               callback_time 0.000275611877441
  6438.             MergeOptimizer
  6439.               nb fail=    0 merged=   37 constant=    4
  6440.               time replace=0.02 validate=0.00 callback=0.02
  6441.     Iter 5
  6442.             TopoOptimizer  topo_constant_folding
  6443.               nb_node (start, end, changed) (4559, 4559, 0)
  6444.               init io_toposort 0.435209989548
  6445.               loop time 0.00607204437256
  6446.               callback_time 0.0
  6447.             MergeOptimizer
  6448.               nb fail=    0 merged=   28 constant=    0
  6449.               time replace=0.02 validate=0.00 callback=0.02
  6450.     Iter 6
  6451.             TopoOptimizer  topo_constant_folding
  6452.               nb_node (start, end, changed) (4559, 4559, 0)
  6453.               init io_toposort 0.0452370643616
  6454.               loop time 0.0053391456604
  6455.               callback_time 0.0
  6456.             MergeOptimizer
  6457.               nb fail=    0 merged=    0 constant=    0
  6458.               time replace=0.00 validate=0.00 callback=0.00
  6459.   14.695019s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 2957, 2957) - 2.875s
  6460.   6.134235s - ('gpuarray_opt', 'SeqOptimizer', 16, 4633, 5796) - 0.029s
  6461.     SeqOptimizer      gpuarray_opt  time 6.134s for 4633/5796 nodes before/after optimization
  6462.       1.806s for callback
  6463.           0.029s for fgraph.validate()
  6464.       callbacks_time
  6465.            <theano.tensor.opt.ShapeFeature object at 0x121562590> , 0.800983190536
  6466.            <theano.gof.opt.MergeFeature object at 0x1227e7190> , 0.281701803207
  6467.            Updater{gpuarray_local_optimizations} , 0.261435270309
  6468.            Updater{gpuarray_cut_transfers} , 0.261384248734
  6469.            Updater{gpuarray_local_optimizations} , 0.0294797420502
  6470.            <theano.gof.toolbox.ReplaceValidate object at 0x12087ee50> , 0.0172119140625
  6471.            <theano.gof.opt.ChangeTracker instance at 0x125a3e1b8> , 0.012256860733
  6472.            <theano.gof.toolbox.PreserveVariableAttributes object at 0x122195850> , 0.0115115642548
  6473.            <theano.compile.function_module.Supervisor instance at 0x1227d3f38> , 0.00647950172424
  6474.            <theano.gof.opt.ChangeTracker instance at 0x127399f38> , 0.00211262702942
  6475.            Updater{gpuarray_local_optimizations} , 5.72204589844e-06
  6476.       3.286718s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 5910, 7186) - 0.018s
  6477.         EquilibriumOptimizer          gpuarray_local_optimizations
  6478.           time 3.286s for 4 passes
  6479.           nb nodes (start, end,  max) 5910 7186 7755
  6480.           time io_toposort 0.280s
  6481.           time in local optimizers 2.832s
  6482.           time in global optimizers 0.000s
  6483.           time in final optimizers 0.000s
  6484.           time in cleanup optimizers 0.000s
  6485.            0 - 2.157s 1135 (0.000s in global opts, 0.055s io_toposort) - 5910 nodes - ('constant_folding', 747) ('local_gpua_dimshuffle', 135) ('local_gpua_elemwise', 73) ('local_gpua_subtensor', 48) ('local_abstractconv_gw_cudnn', 43) ...
  6486.            1 - 0.729s 319 (0.000s in global opts, 0.079s io_toposort) - 7755 nodes - ('constant_folding', 271) ('local_dnn_convw_output_merge', 30) ('local_dnn_convi_output_merge', 9) ('local_gpualloc_memset_0', 9)
  6487.            2 - 0.201s 1 (0.000s in global opts, 0.074s io_toposort) - 7187 nodes - ('constant_folding', 1)
  6488.            3 - 0.199s 0 (0.000s in global opts, 0.072s io_toposort) - 7186 nodes -
  6489.           times - times applied - nb node created - name:
  6490.           1.373s - 1019 - 0 - constant_folding
  6491.           0.243s - 30 - 60 - local_dnn_convw_output_merge
  6492.           0.190s - 135 - 405 - local_gpua_dimshuffle
  6493.           0.173s - 36 - 684 - local_abstractconv_gi_cudnn
  6494.           0.149s - 43 - 774 - local_abstractconv_gw_cudnn
  6495.           0.137s - 9 - 18 - local_dnn_convi_output_merge
  6496.           0.128s - 29 - 734 - local_abstractconv_cudnn
  6497.           0.087s - 73 - 243 - local_gpua_elemwise
  6498.           0.048s - 48 - 144 - local_gpua_subtensor
  6499.           0.036s - 23 - 23 - local_gpu_elemwise_careduce
  6500.           0.029s - 1 - 3 - local_gpua_gemm_output_merge
  6501.           0.010s - 9 - 9 - local_gpualloc_memset_0
  6502.           0.230s - in 58 optimization that were not used (display only those with a runtime > 0)
  6503.             0.031s - local_track_shape_i
  6504.             0.024s - local_gpua_gemm_alpha_merge
  6505.             0.020s - local_dnn_conv_output_merge
  6506.             0.020s - local_gpua_gemmbatch_output_merge
  6507.             0.019s - local_dnn_conv_alpha_merge
  6508.             0.019s - local_gemm16_output_merge
  6509.             0.019s - local_gpua_gemmbatch_alpha_merge
  6510.             0.019s - local_gemm16_alpha_merge
  6511.             0.019s - local_dnn_convi_alpha_merge
  6512.             0.019s - local_dnn_convw_alpha_merge
  6513.             0.013s - local_log_softmax_dnn
  6514.             0.003s - local_gpua_assert
  6515.             0.002s - local_gpua_shape
  6516.             0.001s - local_gpu_contiguous_gpu_contiguous
  6517.             0.001s - local_gpua_abstractconv2d
  6518.  
  6519.       2.210998s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 4633, 5910) - 0.000s
  6520.         GraphToGPUOptimizer          gpuarray_graph_optimization
  6521.           time io_toposort 0.415s
  6522.         Total time taken by local optimizers 0.340s
  6523.           times - times applied - Node created - name:
  6524.           0.268s - 199 - 199 - local_gpua_careduce
  6525.           0.054s - 2631 - 3852 - local_gpua_elemwise
  6526.           0.006s - 162 - 162 - local_gpua_assert_graph
  6527.           0.005s - 108 - 108 - local_gpua_lift_abstractconv2d_graph
  6528.           0.003s - 24 - 24 - local_gpua_inc_subtensor
  6529.           0.003s - 226 - 226 - local_gpua_dimshuffle
  6530.           0.001s - 15 - 30 - local_gpua_mrg_graph
  6531.           0.000s - 1 - 8 - local_gpua_dot22scalar
  6532.           0.000s - 22 - 22 - local_gpua_subtensor_graph
  6533.           0.000s - 50 - 50 - local_gpua_reshape
  6534.           0.000s - 36 - 36 - local_gpua_dot22
  6535.           0.000s - 10 - 19 - local_gpua_alloc
  6536.           0.000s - 9 - 27 - local_gpua_gemm
  6537.           0.000s - 1 - 1 - local_gpua_crossentropysoftmaxargmax1hotwithbias
  6538.           0.000s - 1 - 2 - local_gpua_crossentropysoftmax1hotwithbiasdx
  6539.           0.000s - in 1 optimization that were not used (display only those with a runtime > 0)
  6540.  
  6541.       0.636133s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 7186, 5796) - 0.011s
  6542.         EquilibriumOptimizer          gpuarray_cut_transfers
  6543.           time 0.636s for 2 passes
  6544.           nb nodes (start, end,  max) 7186 5796 7186
  6545.           time io_toposort 0.141s
  6546.           time in local optimizers 0.455s
  6547.           time in global optimizers 0.000s
  6548.           time in final optimizers 0.000s
  6549.           time in cleanup optimizers 0.000s
  6550.            0 - 0.549s 927 (0.000s in global opts, 0.077s io_toposort) - 7186 nodes - ('local_cut_gpu_transfers', 927)
  6551.            1 - 0.087s 0 (0.000s in global opts, 0.064s io_toposort) - 5796 nodes -
  6552.           times - times applied - nb node created - name:
  6553.           0.437s - 927 - 0 - local_cut_gpu_transfers
  6554.           0.018s - in 1 optimization that were not used (display only those with a runtime > 0)
  6555.             0.018s - constant_folding
  6556.  
  6557.       0.000100s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 5910, 5910) - 0.000s
  6558.  
  6559.   5.952148s - ('specialize', 'EquilibriumOptimizer', 13, 5202, 4633) - 0.016s
  6560.     EquilibriumOptimizer      specialize
  6561.       time 5.952s for 11 passes
  6562.       nb nodes (start, end,  max) 5202 4633 5337
  6563.       time io_toposort 0.851s
  6564.       time in local optimizers 3.097s
  6565.       time in global optimizers 0.582s
  6566.       time in final optimizers 1.026s
  6567.       time in cleanup optimizers 0.000s
  6568.        0 - 0.868s 520 (0.168s in global opts, 0.048s io_toposort) - 5202 nodes - ('local_reduce_broadcastable', 140) ('local_pow_specialize', 93) ('local_shape_to_shape_i', 72) ('local_add_specialize', 63) ('local_mul_specialize', 52) ...
  6569.        1 - 1.194s 529 (0.193s in global opts, 0.418s io_toposort) - 5337 nodes - ('local_fill_to_alloc', 246) ('local_dimshuffle_lift', 168) ('local_subtensor_make_vector', 72) ('local_useless_elemwise', 30) ('local_mul_specialize', 7) ...
  6570.        2 - 0.428s 65 (0.093s in global opts, 0.043s io_toposort) - 4686 nodes - ('local_remove_useless_assert', 30) ('local_mul_specialize', 28) ('local_func_inv', 4) ('local_elemwise_alloc', 3)
  6571.        3 - 0.386s 3 (0.092s in global opts, 0.044s io_toposort) - 4645 nodes - ('local_neg_div_neg', 3)
  6572.        4 - 0.377s 4 (0.096s in global opts, 0.042s io_toposort) - 4642 nodes - ('local_zero_div', 3) ('topo_constant_folding', 1)
  6573.        5 - 0.387s 4 (0.104s in global opts, 0.043s io_toposort) - 4642 nodes - ('local_elemwise_alloc', 3) ('topo_constant_folding', 1)
  6574.        6 - 0.761s 4 (0.475s in global opts, 0.042s io_toposort) - 4639 nodes - ('local_zero_div', 3) ('topo_constant_folding', 1)
  6575.        7 - 0.413s 4 (0.105s in global opts, 0.043s io_toposort) - 4639 nodes - ('local_elemwise_alloc', 3) ('topo_constant_folding', 1)
  6576.        8 - 0.377s 3 (0.093s in global opts, 0.043s io_toposort) - 4636 nodes - ('local_add_specialize', 3)
  6577.        9 - 0.386s 6 (0.093s in global opts, 0.041s io_toposort) - 4639 nodes - ('local_fill_to_alloc', 6)
  6578.       10 - 0.375s 0 (0.095s in global opts, 0.042s io_toposort) - 4633 nodes -
  6579.       times - times applied - nb node created - name:
  6580.       1.026s - 6 - 0 - topo_constant_folding
  6581.       0.582s - 1 - 1 - crossentropy_to_crossentropy_with_softmax_with_bias
  6582.       0.412s - 66 - 255 - local_add_specialize
  6583.       0.231s - 6 - 0 - local_func_inv
  6584.       0.225s - 168 - 112 - local_dimshuffle_lift
  6585.       0.220s - 87 - 78 - local_mul_specialize
  6586.       0.207s - 9 - 15 - local_elemwise_alloc
  6587.       0.198s - 30 - 30 - local_useless_elemwise
  6588.       0.108s - 19 - 19 - local_div_to_inv
  6589.       0.098s - 39 - 30 - local_remove_useless_assert
  6590.       0.076s - 140 - 140 - local_reduce_broadcastable
  6591.       0.075s - 31 - 121 - local_sum_prod_mul_by_scalar
  6592.       0.070s - 7 - 0 - local_useless_switch
  6593.       0.068s - 93 - 93 - local_pow_specialize
  6594.       0.047s - 6 - 12 - local_zero_div
  6595.       0.043s - 252 - 3 - local_fill_to_alloc
  6596.       0.042s - 21 - 21 - local_mul_to_sqr
  6597.       0.039s - 72 - 72 - local_shape_to_shape_i
  6598.       0.039s - 83 - 0 - local_subtensor_make_vector
  6599.       0.007s - 3 - 3 - local_neg_div_neg
  6600.       0.001s - 1 - 1 - local_softmax_with_bias
  6601.       0.001s - 1 - 1 - local_softmax_grad_to_crossentropy_with_softmax_grad
  6602.       0.001s - 1 - 1 - local_useless_crossentropy_softmax_1hot_with_bias_dx_alloc
  6603.       0.888s - in 52 optimization that were not used (display only those with a runtime > 0)
  6604.         0.148s - local_one_minus_erf2
  6605.         0.134s - local_abs_merge
  6606.         0.111s - local_useless_elemwise_comparison
  6607.         0.064s - local_track_shape_i
  6608.         0.061s - local_mul_switch_sink
  6609.         0.049s - local_expm1
  6610.         0.048s - local_elemwise_sub_zeros
  6611.         0.047s - local_logsoftmax
  6612.         0.046s - local_cast_cast
  6613.         0.042s - local_one_minus_erf
  6614.         0.039s - local_alloc_unary
  6615.         0.036s - local_grad_log_erfc_neg
  6616.         0.015s - local_useless_subtensor
  6617.         0.009s - local_sum_prod_div_dimshuffle
  6618.         0.008s - local_log1p
  6619.         0.007s - local_useless_slice
  6620.         0.004s - local_sumsqr2dot
  6621.         0.004s - local_opt_alloc
  6622.         0.003s - local_subtensor_remove_broadcastable_index
  6623.         0.002s - local_neg_neg
  6624.         0.001s - local_useless_inc_subtensor
  6625.         0.001s - local_log_add
  6626.         0.001s - local_subtensor_merge
  6627.         0.001s - local_subtensor_of_alloc
  6628.         0.001s - local_subtensor_of_dot
  6629.         0.001s - local_log_erfc
  6630.         0.001s - local_useless_inc_subtensor_alloc
  6631.         0.001s - local_advanced_indexing_crossentropy_onehot
  6632.         0.001s - local_canonicalize_alloc
  6633.         0.000s - local_useless_alloc
  6634.         0.000s - local_merge_alloc
  6635.         0.000s - local_scalar_tensor_scalar
  6636.         0.000s - local_logsoftmax_grad
  6637.  
  6638.     Global, final and clean up optimizers
  6639.     Iter 0
  6640.             TopoOptimizer  topo_constant_folding
  6641.               nb_node (start, end, changed) (5411, 5338, 73)
  6642.               init io_toposort 0.0502741336823
  6643.               loop time 0.0653800964355
  6644.               callback_time 0.0122356414795
  6645.     Iter 1
  6646.             TopoOptimizer  topo_constant_folding
  6647.               nb_node (start, end, changed) (4714, 4686, 28)
  6648.               init io_toposort 0.0429389476776
  6649.               loop time 0.0369338989258
  6650.               callback_time 0.0113785266876
  6651.     Iter 2
  6652.             TopoOptimizer  topo_constant_folding
  6653.               nb_node (start, end, changed) (4645, 4645, 0)
  6654.               init io_toposort 0.0422580242157
  6655.               loop time 0.00536894798279
  6656.               callback_time 0.0
  6657.     Iter 3
  6658.             TopoOptimizer  topo_constant_folding
  6659.               nb_node (start, end, changed) (4642, 4642, 0)
  6660.               init io_toposort 0.0418980121613
  6661.               loop time 0.00537490844727
  6662.               callback_time 0.0
  6663.     Iter 4
  6664.             TopoOptimizer  topo_constant_folding
  6665.               nb_node (start, end, changed) (4645, 4642, 3)
  6666.               init io_toposort 0.044823884964
  6667.               loop time 0.0064218044281
  6668.               callback_time 0.000365972518921
  6669.     Iter 5
  6670.             TopoOptimizer  topo_constant_folding
  6671.               nb_node (start, end, changed) (4642, 4639, 3)
  6672.               init io_toposort 0.0490889549255
  6673.               loop time 0.00785398483276
  6674.               callback_time 0.000781059265137
  6675.     Iter 6
  6676.             TopoOptimizer  topo_constant_folding
  6677.               nb_node (start, end, changed) (4642, 4639, 3)
  6678.               init io_toposort 0.42392206192
  6679.               loop time 0.00623798370361
  6680.               callback_time 0.000385522842407
  6681.     Iter 7
  6682.             TopoOptimizer  topo_constant_folding
  6683.               nb_node (start, end, changed) (4639, 4636, 3)
  6684.               init io_toposort 0.0444281101227
  6685.               loop time 0.00804901123047
  6686.               callback_time 0.000782251358032
  6687.     Iter 8
  6688.             TopoOptimizer  topo_constant_folding
  6689.               nb_node (start, end, changed) (4639, 4639, 0)
  6690.               init io_toposort 0.0420498847961
  6691.               loop time 0.00549507141113
  6692.               callback_time 0.0
  6693.     Iter 9
  6694.             TopoOptimizer  topo_constant_folding
  6695.               nb_node (start, end, changed) (4633, 4633, 0)
  6696.               init io_toposort 0.0423140525818
  6697.               loop time 0.00554203987122
  6698.               callback_time 0.0
  6699.     Iter 10
  6700.             TopoOptimizer  topo_constant_folding
  6701.               nb_node (start, end, changed) (4633, 4633, 0)
  6702.               init io_toposort 0.0434467792511
  6703.               loop time 0.00532102584839
  6704.               callback_time 0.0
  6705.   3.769982s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 5272, 3758) - 0.006s
  6706.     FusionOptimizer
  6707.      nb_iter 3
  6708.      nb_replacement 571
  6709.      nb_inconsistency_replace 0
  6710.      validate_time 0.00642895698547
  6711.      callback_time 0.202159881592
  6712.      time_toposort 0.672288179398
  6713.   2.975995s - ('merge3', 'MergeOptimizer', 51, 2957, 2957) - 2.962s
  6714.     MergeOptimizer
  6715.       nb fail=    0 merged=   47 constant=   47
  6716.       time replace=2.98 validate=2.96 callback=2.97
  6717.       callbacks_time
  6718.            <theano.gof.toolbox.PreserveVariableAttributes object at 0x122195850> , 0.000165700912476
  6719.            <theano.gof.toolbox.ReplaceValidate object at 0x12087ee50> , 0.000275135040283
  6720.            <theano.gof.opt.MergeFeature object at 0x1227e7190> , 0.00250196456909
  6721.            <theano.tensor.opt.ShapeFeature object at 0x121562590> , 0.00316381454468
  6722.            <theano.compile.function_module.Supervisor instance at 0x1227d3f38> , 0.807505607605
  6723.            <theano.gof.destroyhandler.DestroyHandler object at 0x1291a3f10> , 2.15578103065
  6724.   2.688769s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 2956, 2957) - 2.480s
  6725.     TopoOptimizer  local_dnna_conv_inplace
  6726.       nb_node (start, end, changed) (2956, 2957, 108)
  6727.       init io_toposort 0.0287601947784
  6728.       loop time 2.65987706184
  6729.       callback_time 2.56827378273
  6730. LocalOptGroup
  6731. ---------------------
  6732.   time taken - times applied - times tried - name - node_created:
  6733.   -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  6734.   -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  6735.   -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  6736.   0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  6737.  
  6738.   1.944101s - ('elemwise_fusion', 'SeqOptimizer', 19, 5796, 5272) - 0.004s
  6739.     SeqOptimizer      elemwise_fusion  time 1.944s for 5796/5272 nodes before/after optimization
  6740.       0.120s for callback
  6741.           0.004s for fgraph.validate()
  6742.       1.477741s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 5632, 5272) - 0.001s
  6743.         FusionOptimizer
  6744.          nb_iter 3
  6745.          nb_replacement 101
  6746.          nb_inconsistency_replace 0
  6747.          validate_time 0.00115561485291
  6748.          callback_time 0.0301287174225
  6749.          time_toposort 0.670944929123
  6750.       0.466092s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 5796, 5632) - 0.003s
  6751.         FusionOptimizer
  6752.          nb_iter 3
  6753.          nb_replacement 259
  6754.          nb_inconsistency_replace 0
  6755.          validate_time 0.00277924537659
  6756.          callback_time 0.0899620056152
  6757.          time_toposort 0.185888051987
  6758.  
  6759.   1.419394s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 5403, 5205) - 0.003s
  6760.     EquilibriumOptimizer      scan_eqopt2
  6761.       time 1.419s for 2 passes
  6762.       nb nodes (start, end,  max) 5403 5205 5403
  6763.       time io_toposort 0.097s
  6764.       time in local optimizers 0.000s
  6765.       time in global optimizers 1.302s
  6766.       time in final optimizers 0.000s
  6767.       time in cleanup optimizers 0.000s
  6768.        0 - 0.982s 1 (0.922s in global opts, 0.049s io_toposort) - 5205 nodes - ('constant_folding', 1)
  6769.        1 - 0.437s 0 (0.381s in global opts, 0.048s io_toposort) - 5205 nodes -
  6770.       times - times applied - nb node created - name:
  6771.       0.263s - 1 - 0 - constant_folding
  6772.       1.039s - in 6 optimization that were not used (display only those with a runtime > 0)
  6773.         0.495s - remove_constants_and_unused_inputs_scan
  6774.         0.114s - scan_merge_inouts
  6775.         0.110s - remove_constants_and_unused_inputs_scan
  6776.         0.109s - <theano.scan_module.scan_opt.ScanMerge object at 0x10f69f750>
  6777.         0.108s - remove_constants_and_unused_inputs_scan
  6778.         0.102s - <theano.scan_module.scan_opt.ScanSaveMem object at 0x10f69fa10>
  6779.  
  6780.     Global, final and clean up optimizers
  6781.     Iter 0
  6782.             TopoOptimizer  constant_folding_for_scan2
  6783.               nb_node (start, end, changed) (5403, 5205, 198)
  6784.               init io_toposort 0.0526330471039
  6785.               loop time 0.149075984955
  6786.               callback_time 0.048490524292
  6787.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  6788.               nb_node (start, end, changed) (5205, 5205, 0)
  6789.               init io_toposort 0.0490438938141
  6790.               loop time 0.00537800788879
  6791.               callback_time 0.0
  6792.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  6793.               nb_node (start, end, changed) (5205, 5205, 0)
  6794.               init io_toposort 0.43702507019
  6795.               loop time 0.00532221794128
  6796.               callback_time 0.0
  6797.             TopoOptimizer  scanOp_merge_inouts
  6798.               nb_node (start, end, changed) (5205, 5205, 0)
  6799.               init io_toposort 0.0544281005859
  6800.               loop time 0.00595307350159
  6801.               callback_time 0.0
  6802.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  6803.               nb_node (start, end, changed) (5205, 5205, 0)
  6804.               init io_toposort 0.0488278865814
  6805.               loop time 0.00527000427246
  6806.               callback_time 0.0
  6807.     Iter 1
  6808.             TopoOptimizer  constant_folding_for_scan2
  6809.               nb_node (start, end, changed) (5205, 5205, 0)
  6810.               init io_toposort 0.0551071166992
  6811.               loop time 0.00627899169922
  6812.               callback_time 0.0
  6813.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  6814.               nb_node (start, end, changed) (5205, 5205, 0)
  6815.               init io_toposort 0.0485699176788
  6816.               loop time 0.00670099258423
  6817.               callback_time 0.0
  6818.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  6819.               nb_node (start, end, changed) (5205, 5205, 0)
  6820.               init io_toposort 0.0476229190826
  6821.               loop time 0.00529885292053
  6822.               callback_time 0.0
  6823.             TopoOptimizer  scanOp_merge_inouts
  6824.               nb_node (start, end, changed) (5205, 5205, 0)
  6825.               init io_toposort 0.0482048988342
  6826.               loop time 0.00564789772034
  6827.               callback_time 0.0
  6828.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  6829.               nb_node (start, end, changed) (5205, 5205, 0)
  6830.               init io_toposort 0.0479300022125
  6831.               loop time 0.00541400909424
  6832.               callback_time 0.0
  6833.   1.065656s - ('BlasOpt', 'SeqOptimizer', 12, 5205, 5202) - 0.001s
  6834.     SeqOptimizer      BlasOpt  time 1.065s for 5205/5202 nodes before/after optimization
  6835.       0.012s for callback
  6836.           0.001s for fgraph.validate()
  6837.       0.433675s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 5205, 5205) - 0.000s
  6838.         TopoOptimizer  local_dot_to_dot22
  6839.           nb_node (start, end, changed) (5205, 5205, 45)
  6840.           init io_toposort 0.406867027283
  6841.           loop time 0.0267560482025
  6842.           callback_time 0.00478005409241
  6843.       0.384683s - ('gemm_optimizer', 'GemmOptimizer', 1, 5205, 5201) - 0.000s
  6844.         GemmOptimizer
  6845.          nb_iter 2
  6846.          nb_replacement 9
  6847.          nb_replacement_didn_t_remove 2
  6848.          nb_inconsistency_make 0
  6849.          nb_inconsistency_replace 0
  6850.          time_canonicalize 0.112722396851
  6851.          time_factor_can 3.38554382324e-05
  6852.          time_factor_list 0.00463604927063
  6853.          time_toposort 0.0978970527649
  6854.          validate_time 0.000133037567139
  6855.          callback_time 0.00729250907898
  6856.       0.068988s - ('use_c_blas', 'TopoOptimizer', 4, 5202, 5202) - 0.000s
  6857.         TopoOptimizer  use_c_blas
  6858.           nb_node (start, end, changed) (5202, 5202, 0)
  6859.           init io_toposort 0.0484161376953
  6860.           loop time 0.0205109119415
  6861.           callback_time 0.0
  6862. LocalOptGroup
  6863. ---------------------
  6864. --- The Optimizer wasn't successful ---
  6865.  
  6866.        0.063015s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 5201, 5202) - 0.000s
  6867.          TopoOptimizer  local_dot22_to_dot22scalar
  6868.            nb_node (start, end, changed) (5201, 5202, 1)
  6869.            init io_toposort 0.0481488704681
  6870.            loop time 0.0148110389709
  6871.            callback_time 0.000154972076416
  6872.        0.059894s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 5202, 5202) - 0.000s
  6873.          EquilibriumOptimizer          local_gemm_to_gemv
  6874.            time 0.060s for 1 passes
  6875.            nb nodes (start, end,  max) 5202 5202 5202
  6876.            time io_toposort 0.047s
  6877.            time in local optimizers 0.003s
  6878.            time in global optimizers 0.000s
  6879.            time in final optimizers 0.000s
  6880.            time in cleanup optimizers 0.000s
  6881.             0 - 0.060s 0 (0.000s in global opts, 0.047s io_toposort) - 5202 nodes -
  6882.        0.055211s - ('use_scipy_ger', 'TopoOptimizer', 5, 5202, 5202) - 0.000s
  6883.          TopoOptimizer  scipy_blas
  6884.            nb_node (start, end, changed) (5202, 5202, 0)
  6885.            init io_toposort 0.0487020015717
  6886.            loop time 0.00645112991333
  6887.            callback_time 0.0
  6888.  
  6889.    0.870026s - ('stabilize', 'EquilibriumOptimizer', 8, 4559, 4622) - 0.000s
  6890.      EquilibriumOptimizer      stabilize
  6891.        time 0.870s for 2 passes
  6892.        nb nodes (start, end,  max) 4559 4622 4622
  6893.        time io_toposort 0.086s
  6894.        time in local optimizers 0.553s
  6895.        time in global optimizers 0.095s
  6896.        time in final optimizers 0.100s
  6897.        time in cleanup optimizers 0.000s
  6898.         0 - 0.444s 20 (0.097s in global opts, 0.043s io_toposort) - 4559 nodes - ('local_fill_to_alloc', 11) ('local_log1p', 9)
  6899.         1 - 0.426s 0 (0.098s in global opts, 0.043s io_toposort) - 4622 nodes -
  6900.        times - times applied - nb node created - name:
  6901.        0.012s - 11 - 66 - local_fill_to_alloc
  6902.        0.008s - 9 - 18 - local_log1p
  6903.        0.729s - in 40 optimization that were not used (display only those with a runtime > 0)
  6904.          0.381s - local_greedy_distributor
  6905.          0.100s - topo_constant_folding
  6906.          0.095s - crossentropy_to_crossentropy_with_softmax_with_bias
  6907.          0.052s - local_sigm_times_exp
  6908.          0.028s - local_one_minus_erf2
  6909.          0.024s - local_exp_over_1_plus_exp
  6910.          0.019s - local_useless_elemwise_comparison
  6911.          0.010s - local_expm1
  6912.          0.008s - local_grad_log_erfc_neg
  6913.          0.007s - local_one_minus_erf
  6914.          0.001s - Elemwise{log,no_inplace}(sigmoid(x)) -> Elemwise{neg,no_inplace}(softplus(Elemwise{neg,no_inplace}(x)))
  6915.          0.001s - Elemwise{log,no_inplace}(Elemwise{sub,no_inplace}(y subject to <function _is_1 at 0x10efc5758>, sigmoid(x))) -> Elemwise{neg,no_inplace}(softplus(x))
  6916.          0.001s - local_0_dot_x
  6917.          0.000s - local_incsubtensor_of_zeros
  6918.          0.000s - local_useless_reshape
  6919.          0.000s - local_canonicalize_alloc
  6920.          0.000s - local_subtensor_of_dot
  6921.          0.000s - local_log_add
  6922.          0.000s - Elemwise{log1p,no_inplace}(Elemwise{exp,no_inplace}(x)) -> softplus(x)
  6923.          0.000s - local_reshape_lift
  6924.          0.000s - local_log_erfc
  6925.          0.000s - Elemwise{log1p,no_inplace}(Elemwise{neg,no_inplace}(sigmoid(x))) -> Elemwise{neg,no_inplace}(softplus(x))
  6926.          0.000s - local_useless_alloc
  6927.          0.000s - local_useless_inc_subtensor_alloc
  6928.          0.000s - local_merge_alloc
  6929.          0.000s - local_setsubtensor_of_constants
  6930.  
  6931.      Global, final and clean up optimizers
  6932.      Iter 0
  6933.              TopoOptimizer  topo_constant_folding
  6934.                nb_node (start, end, changed) (4622, 4622, 0)
  6935.                init io_toposort 0.043016910553
  6936.                loop time 0.00549912452698
  6937.                callback_time 0.0
  6938.      Iter 1
  6939.              TopoOptimizer  topo_constant_folding
  6940.                nb_node (start, end, changed) (4622, 4622, 0)
  6941.                init io_toposort 0.0462830066681
  6942.                loop time 0.00541090965271
  6943.                callback_time 0.0
  6944.    0.837172s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 7753, 7753) - 0.000s
  6945.      EquilibriumOptimizer      scan_eqopt1
  6946.        time 0.837s for 1 passes
  6947.        nb nodes (start, end,  max) 7753 7753 7753
  6948.        time io_toposort 0.085s
  6949.        time in local optimizers 0.000s
  6950.        time in global optimizers 0.740s
  6951.        time in final optimizers 0.000s
  6952.        time in cleanup optimizers 0.000s
  6953.         0 - 0.837s 0 (0.740s in global opts, 0.085s io_toposort) - 7753 nodes -
  6954.      Global, final and clean up optimizers
  6955.      Iter 0
  6956.              SeqOptimizer              all_pushout_opt  time 0.740s for 7753/7753 nodes before/after optimization
  6957.                0.000s for callback
  6958.                    0.000s for fgraph.validate()
  6959.                0.396074s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 7753, 7753) - 0.000s
  6960.                0.091526s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 7753, 7753) - 0.000s
  6961.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  6962.                    nb_node (start, end, changed) (7753, 7753, 0)
  6963.                    init io_toposort 0.0833880901337
  6964.                    loop time 0.0080738067627
  6965.                    callback_time 0.0
  6966.                0.086859s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 7753, 7753) - 0.000s
  6967.                0.084704s - ('scan_pushout_dot1', 'PushOutDot1', 3, 7753, 7753) - 0.000s
  6968.                0.080556s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 7753, 7753) - 0.000s
  6969.  
  6970.    0.783922s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 2957, 2957) - 0.732s
  6971.      TopoOptimizer  random_make_inplace_mrg
  6972.        nb_node (start, end, changed) (2957, 2957, 15)
  6973.        init io_toposort 0.0296950340271
  6974.        loop time 0.754165887833
  6975.        callback_time 0.741482257843
  6976.    0.660968s - ('merge1', 'MergeOptimizer', 0, 10599, 7753) - 0.035s
  6977.      MergeOptimizer
  6978.        nb fail=    0 merged= 4133 constant= 1287
  6979.        time replace=0.45 validate=0.03 callback=0.21
  6980.    0.554039s - ('ShapeOpt', 'ShapeOptimizer', 2, 7753, 7753) - 0.000s
  6981.    0.524648s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 2956, 2956) - 0.091s
  6982.      TopoOptimizer  dimshuffle_as_view
  6983.        nb_node (start, end, changed) (2956, 2956, 299)
  6984.        init io_toposort 0.0304479598999
  6985.        loop time 0.494140148163
  6986.        callback_time 0.354798316956
  6987.    0.368721s - ('merge2', 'MergeOptimizer', 22, 3758, 2956) - 0.038s
  6988.      MergeOptimizer
  6989.        nb fail=    0 merged= 4273 constant= 1638
  6990.        time replace=0.37 validate=0.04 callback=0.23
  6991.    0.330622s - ('add_destroy_handler', 'AddDestroyHandler', 23, 2956, 2956) - 0.000s
  6992.    0.310139s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 2956, 2956) - 0.247s
  6993.      TopoOptimizer  local_inplace_setsubtensor
  6994.        nb_node (start, end, changed) (2956, 2956, 18)
  6995.        init io_toposort 0.0290389060974
  6996.        loop time 0.280987024307
  6997.        callback_time 0.256593227386
  6998.    0.272342s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 4622, 5403) - 0.002s
  6999.      TopoOptimizer  local_elemwise_alloc
  7000.        nb_node (start, end, changed) (4622, 5403, 132)
  7001.        init io_toposort 0.0435831546783
  7002.        loop time 0.22869181633
  7003.        callback_time 0.080276966095
  7004.    0.265057s - ('useless', 'TopoOptimizer', 3, 7753, 7645) - 0.002s
  7005.      TopoOptimizer  useless
  7006.        nb_node (start, end, changed) (7753, 7645, 166)
  7007.        init io_toposort 0.0862708091736
  7008.        loop time 0.17871594429
  7009.        callback_time 0.0392537117004
  7010.  LocalOptGroup
  7011.  ---------------------
  7012.    time taken - times applied - times tried - name - node_created:
  7013.    -0.000s - 9 - 9 - local_join_1 - 0
  7014.    -0.000s - 0 - 8 - local_merge_alloc - 0
  7015.    -0.000s - 0 - 8 - local_useless_alloc - 0
  7016.    -0.000s - 0 - 9 - local_join_make_vector - 0
  7017.    -0.000s - 0 - 24 - local_useless_inc_subtensor_alloc - 0
  7018.    -0.000s - 0 - 9 - local_join_empty - 0
  7019.    -0.000s - 0 - 24 - local_useless_inc_subtensor - 0
  7020.    -0.000s - 0 - 50 - local_useless_reshape - 0
  7021.    -0.000s - 0 - 323 - local_subtensor_of_alloc - 0
  7022.    -0.001s - 6 - 323 - local_subtensor_make_vector - 0
  7023.    -0.001s - 1 - 102 - local_useless_fill - 0
  7024.    -0.001s - 0 - 358 - local_useless_reduce - 0
  7025.    -0.002s - 0 - 323 - local_useless_slice - 0
  7026.    -0.004s - 6 - 6 - local_useless_split - 60
  7027.    -0.004s - 0 - 4958 - local_useless_switch - 0
  7028.    -0.010s - 0 - 4958 - local_useless_elemwise_comparison - 0
  7029.    -0.024s - 159 - 4958 - local_useless_elemwise - 0
  7030.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  7031.  
  7032.    0.199589s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 2956, 2956) - 0.143s
  7033.      TopoOptimizer  InplaceGpuaBlasOpt
  7034.        nb_node (start, end, changed) (2956, 2956, 10)
  7035.        init io_toposort 0.0286979675293
  7036.        loop time 0.170758008957
  7037.        callback_time 0.147303819656
  7038.  LocalOptGroup
  7039.  ---------------------
  7040.    time taken - times applied - times tried - name - node_created:
  7041.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  7042.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  7043.  
  7044.    0.118040s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 7639, 7639) - 0.000s
  7045.      TopoOptimizer  pre_local_IncSubtensor_serialize
  7046.        nb_node (start, end, changed) (7639, 7639, 9)
  7047.        init io_toposort 0.0757310390472
  7048.        loop time 0.0422530174255
  7049.        callback_time 0.00764083862305
  7050.    0.100773s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 4633, 4633) - 0.000s
  7051.      EquilibriumOptimizer      uncanonicalize
  7052.        time 0.101s for 1 passes
  7053.        nb nodes (start, end,  max) 4633 4633 4633
  7054.        time io_toposort 0.042s
  7055.        time in local optimizers 0.001s
  7056.        time in global optimizers 0.000s
  7057.        time in final optimizers 0.050s
  7058.        time in cleanup optimizers 0.000s
  7059.         0 - 0.101s 0 (0.050s in global opts, 0.042s io_toposort) - 4633 nodes -
  7060.      Global, final and clean up optimizers
  7061.      Iter 0
  7062.              TopoOptimizer  topo_constant_folding
  7063.                nb_node (start, end, changed) (4633, 4633, 0)
  7064.                init io_toposort 0.0444288253784
  7065.                loop time 0.00537204742432
  7066.                callback_time 0.0
  7067.    0.072703s - ('specialize_device', 'EquilibriumOptimizer', 17, 5796, 5796) - 0.000s
  7068.      EquilibriumOptimizer      specialize_device
  7069.        time 0.073s for 1 passes
  7070.        nb nodes (start, end,  max) 5796 5796 5796
  7071.        time io_toposort 0.062s
  7072.        time in local optimizers 0.000s
  7073.        time in global optimizers 0.000s
  7074.        time in final optimizers 0.000s
  7075.        time in cleanup optimizers 0.000s
  7076.         0 - 0.073s 0 (0.000s in global opts, 0.062s io_toposort) - 5796 nodes -
  7077.    0.068079s - ('AbstractConvCheck', 'TopoOptimizer', 18, 5796, 5796) - 0.000s
  7078.      TopoOptimizer  AbstractConvCheck
  7079.        nb_node (start, end, changed) (5796, 5796, 0)
  7080.        init io_toposort 0.0607590675354
  7081.        loop time 0.00725507736206
  7082.        callback_time 0.0
  7083.    0.062603s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 5272, 5272) - 0.000s
  7084.      FusionOptimizer
  7085.       nb_iter 1
  7086.       nb_replacement 0
  7087.       nb_inconsistency_replace 0
  7088.       validate_time 0.0
  7089.       callback_time 0.0
  7090.       time_toposort 0.0588040351868
  7091.    0.054372s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 2957, 2957) - 0.000s
  7092.    0.054163s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 2957, 2957) - 0.000s
  7093.    0.052785s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 4622, 4622) - 0.000s
  7094.      TopoOptimizer  local_fill_to_alloc
  7095.        nb_node (start, end, changed) (4622, 4622, 0)
  7096.        init io_toposort 0.0429711341858
  7097.        loop time 0.00975203514099
  7098.        callback_time 0.0
  7099.    0.049093s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 4633, 4633) - 0.000s
  7100.    0.046055s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 2956, 2956) - 0.000s
  7101.      TopoOptimizer  InplaceGpuBlasOpt
  7102.        nb_node (start, end, changed) (2956, 2956, 0)
  7103.        init io_toposort 0.0284509658813
  7104.        loop time 0.0174732208252
  7105.        callback_time 0.0
  7106.  LocalOptGroup
  7107.  ---------------------
  7108.  --- The Optimizer wasn't successful ---
  7109.  
  7110.   0.046001s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 2956, 2956) - 0.000s
  7111.     TopoOptimizer  local_dnn_conv_inplace
  7112.       nb_node (start, end, changed) (2956, 2956, 0)
  7113.       init io_toposort 0.0287458896637
  7114.       loop time 0.0171229839325
  7115.       callback_time 0.0
  7116. LocalOptGroup
  7117. ---------------------
  7118. --- The Optimizer wasn't successful ---
  7119.  
  7120.    0.045713s - ('blas_opt_inplace', 'TopoOptimizer', 34, 2956, 2956) - 0.000s
  7121.      TopoOptimizer  InplaceBlasOpt
  7122.        nb_node (start, end, changed) (2956, 2956, 0)
  7123.        init io_toposort 0.0284130573273
  7124.        loop time 0.0171570777893
  7125.        callback_time 0.0
  7126.  LocalOptGroup
  7127.  ---------------------
  7128.  --- The Optimizer wasn't successful ---
  7129.  
  7130.   0.044733s - ('c_blas_destructive', 'TopoOptimizer', 37, 2956, 2956) - 0.000s
  7131.     TopoOptimizer  c_blas_destructive
  7132.       nb_node (start, end, changed) (2956, 2956, 0)
  7133.       init io_toposort 0.0276620388031
  7134.       loop time 0.0170109272003
  7135.       callback_time 0.0
  7136. LocalOptGroup
  7137. ---------------------
  7138. --- The Optimizer wasn't successful ---
  7139.  
  7140.    0.033650s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 2956, 2956) - 0.000s
  7141.      TopoOptimizer  local_advincsub1_gpua_inplace
  7142.        nb_node (start, end, changed) (2956, 2956, 0)
  7143.        init io_toposort 0.0301179885864
  7144.        loop time 0.00347709655762
  7145.        callback_time 0.0
  7146.    0.032851s - ('make_ger_destructive', 'TopoOptimizer', 41, 2957, 2957) - 0.000s
  7147.      TopoOptimizer  make_scipy_blas_destructive
  7148.        nb_node (start, end, changed) (2957, 2957, 0)
  7149.        init io_toposort 0.0286540985107
  7150.        loop time 0.00413203239441
  7151.        callback_time 0.0
  7152.    0.032754s - ('cond_make_inplace', 'TopoOptimizer', 47, 2957, 2957) - 0.000s
  7153.      TopoOptimizer  cond_make_inplace
  7154.        nb_node (start, end, changed) (2957, 2957, 0)
  7155.        init io_toposort 0.0294511318207
  7156.        loop time 0.00323390960693
  7157.        callback_time 0.0
  7158.    0.032468s - ('local_destructive', 'TopoOptimizer', 48, 2957, 2957) - 0.000s
  7159.      TopoOptimizer  CURAND_destructive
  7160.        nb_node (start, end, changed) (2957, 2957, 0)
  7161.        init io_toposort 0.0291059017181
  7162.        loop time 0.00328993797302
  7163.        callback_time 0.0
  7164.    0.032315s - ('random_make_inplace', 'TopoOptimizer', 49, 2957, 2957) - 0.000s
  7165.      TopoOptimizer  random_make_inplace
  7166.        nb_node (start, end, changed) (2957, 2957, 0)
  7167.        init io_toposort 0.0288689136505
  7168.        loop time 0.00338506698608
  7169.        callback_time 0.0
  7170.    0.031910s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 2956, 2956) - 0.000s
  7171.      TopoOptimizer  local_inplace_incsubtensor1
  7172.        nb_node (start, end, changed) (2956, 2956, 0)
  7173.        init io_toposort 0.028608083725
  7174.        loop time 0.00320410728455
  7175.        callback_time 0.0
  7176.    0.031852s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 2956, 2956) - 0.000s
  7177.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  7178.        nb_node (start, end, changed) (2956, 2956, 0)
  7179.        init io_toposort 0.0288248062134
  7180.        loop time 0.00291800498962
  7181.        callback_time 0.0
  7182.    0.031726s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 2956, 2956) - 0.000s
  7183.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  7184.        nb_node (start, end, changed) (2956, 2956, 0)
  7185.        init io_toposort 0.0286209583282
  7186.        loop time 0.00300693511963
  7187.        callback_time 0.0
  7188.    0.031622s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 2957, 2957) - 0.000s
  7189.      TopoOptimizer  local_gemm16_inplace
  7190.        nb_node (start, end, changed) (2957, 2957, 0)
  7191.        init io_toposort 0.0288269519806
  7192.        loop time 0.00267696380615
  7193.        callback_time 0.0
  7194.    0.031494s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 2956, 2956) - 0.000s
  7195.      TopoOptimizer  local_inplace_sparseblockouter
  7196.        nb_node (start, end, changed) (2956, 2956, 0)
  7197.        init io_toposort 0.0281920433044
  7198.        loop time 0.00320482254028
  7199.        callback_time 0.0
  7200.    0.031309s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 2957, 2957) - 0.000s
  7201.    0.031218s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 2956, 2956) - 0.000s
  7202.      TopoOptimizer  local_inplace_sparseblockgemv
  7203.        nb_node (start, end, changed) (2956, 2956, 0)
  7204.        init io_toposort 0.0281801223755
  7205.        loop time 0.00294089317322
  7206.        callback_time 0.0
  7207.    0.031101s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 2956, 2956) - 0.000s
  7208.      TopoOptimizer  local_inplace_sparse_block_outer
  7209.        nb_node (start, end, changed) (2956, 2956, 0)
  7210.        init io_toposort 0.0280029773712
  7211.        loop time 0.00299906730652
  7212.        callback_time 0.0
  7213.    0.031038s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 2956, 2956) - 0.000s
  7214.      TopoOptimizer  local_inplace_sparse_block_gemv
  7215.        nb_node (start, end, changed) (2956, 2956, 0)
  7216.        init io_toposort 0.0279738903046
  7217.        loop time 0.00296902656555
  7218.        callback_time 0.0
  7219.    0.029802s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 2957, 2957) - 0.000s
  7220.    0.002216s - ('merge1.1', 'MergeOptimizer', 4, 7645, 7639) - 0.000s
  7221.      MergeOptimizer
  7222.        nb fail=    0 merged=   24 constant=   18
  7223.        time replace=0.00 validate=0.00 callback=0.00
  7224.    0.000065s - ('merge1.2', 'MergeOptimizer', 7, 4559, 4559) - 0.000s
  7225.      MergeOptimizer
  7226.        nb fail=    0 merged=    0 constant=    0
  7227.        time replace=0.00 validate=0.00 callback=0.00
  7228.  
  7229. Here are tips to potentially make your code run faster
  7230.                  (if you think of new ones, suggest them on the mailing list).
  7231.                  Test them first, as they are not guaranteed to always provide a speedup.
  7232.   Sorry, no tip for today.
  7233. Function profiling
  7234. ==================
  7235.   Message: sb/convnet/sb_resnet.py:341
  7236.   Time in 0 calls to Function.__call__: 0.000000e+00s
  7237.   Total compile time: 2.954339e+01s
  7238.     Number of Apply nodes: 313
  7239.     Theano Optimizer time: 2.045334e+00s
  7240.        Theano validate time: 2.595465e-01s
  7241.     Theano Linker time (includes C, CUDA code generation/compiling): 2.744651e+01s
  7242.        Import time 9.034133e-02s
  7243.        Node make_thunk time 2.742749e+01s
  7244.  
  7245. Time in all call to theano.grad() 2.656322e+00s
  7246. Time since theano import 477.958s
  7247. Optimizer Profile
  7248. -----------------
  7249.  SeqOptimizer  OPT_FAST_RUN  time 2.045s for 385/313 nodes before/after optimization
  7250.    0.470s for callback
  7251.        0.260s for fgraph.validate()
  7252.    time      - (name, class, index, nodes before, nodes after) - validate time
  7253.    0.348002s - ('elemwise_fusion', 'SeqOptimizer', 19, 574, 420) - 0.001s
  7254.      SeqOptimizer      elemwise_fusion  time 0.348s for 574/420 nodes before/after optimization
  7255.        0.011s for callback
  7256.            0.001s for fgraph.validate()
  7257.        0.327701s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 561, 420) - 0.000s
  7258.          FusionOptimizer
  7259.           nb_iter 3
  7260.           nb_replacement 37
  7261.           nb_inconsistency_replace 0
  7262.           validate_time 0.000482320785522
  7263.           callback_time 0.0088849067688
  7264.           time_toposort 0.0130350589752
  7265.        0.020121s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 574, 561) - 0.000s
  7266.          FusionOptimizer
  7267.           nb_iter 2
  7268.           nb_replacement 23
  7269.           nb_inconsistency_replace 0
  7270.           validate_time 0.000242233276367
  7271.           callback_time 0.00241637229919
  7272.           time_toposort 0.00995802879333
  7273.  
  7274.    0.292821s - ('gpuarray_opt', 'SeqOptimizer', 16, 339, 574) - 0.001s
  7275.      SeqOptimizer      gpuarray_opt  time 0.293s for 339/574 nodes before/after optimization
  7276.        0.097s for callback
  7277.            0.001s for fgraph.validate()
  7278.        0.168130s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 313, 603) - 0.001s
  7279.          EquilibriumOptimizer          gpuarray_local_optimizations
  7280.            time 0.168s for 3 passes
  7281.            nb nodes (start, end,  max) 313 603 638
  7282.            time io_toposort 0.016s
  7283.            time in local optimizers 0.142s
  7284.            time in global optimizers 0.000s
  7285.            time in final optimizers 0.000s
  7286.            time in cleanup optimizers 0.000s
  7287.             0 - 0.111s 55 (0.000s in global opts, 0.003s io_toposort) - 313 nodes - ('constant_folding', 28) ('local_gpua_elemwise', 13) ('local_abstractconv_cudnn', 13) ('local_gpua_dimshuffle', 1)
  7288.             1 - 0.042s 35 (0.000s in global opts, 0.008s io_toposort) - 638 nodes - ('constant_folding', 35)
  7289.             2 - 0.015s 0 (0.000s in global opts, 0.006s io_toposort) - 603 nodes -
  7290.            times - times applied - nb node created - name:
  7291.            0.059s - 63 - 0 - constant_folding
  7292.            0.056s - 13 - 328 - local_abstractconv_cudnn
  7293.            0.013s - 13 - 49 - local_gpua_elemwise
  7294.            0.001s - 1 - 3 - local_gpua_dimshuffle
  7295.            0.014s - in 66 optimization that were not used (display only those with a runtime > 0)
  7296.              0.002s - local_track_shape_i
  7297.              0.001s - local_dnn_conv_output_merge
  7298.              0.001s - local_gpua_gemm_alpha_merge
  7299.              0.001s - local_gpua_gemm_output_merge
  7300.              0.001s - local_dnn_conv_alpha_merge
  7301.              0.001s - local_gpua_gemmbatch_alpha_merge
  7302.              0.001s - local_dnn_convw_alpha_merge
  7303.              0.001s - local_gpua_gemmbatch_output_merge
  7304.              0.001s - local_gemm16_output_merge
  7305.              0.001s - local_dnn_convi_alpha_merge
  7306.              0.001s - local_gemm16_alpha_merge
  7307.              0.001s - local_dnn_convw_output_merge
  7308.              0.001s - local_dnn_convi_output_merge
  7309.              0.001s - local_log_softmax_dnn
  7310.              0.000s - local_gpua_shape
  7311.              0.000s - local_gpu_contiguous_gpu_contiguous
  7312.              0.000s - local_gpua_abstractconv2d
  7313.              0.000s - local_gpua_assert
  7314.              0.000s - local_gpu_elemwise_careduce
  7315.  
  7316.        0.106716s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 339, 313) - 0.000s
  7317.          GraphToGPUOptimizer          gpuarray_graph_optimization
  7318.            time io_toposort 0.004s
  7319.          Total time taken by local optimizers 0.012s
  7320.            times - times applied - Node created - name:
  7321.            0.006s - 3 - 3 - local_gpua_careduce
  7322.            0.003s - 120 - 154 - local_gpua_elemwise
  7323.            0.001s - 13 - 13 - local_gpua_lift_abstractconv2d_graph
  7324.            0.001s - 50 - 50 - local_gpua_dimshuffle
  7325.            0.000s - 3 - 6 - local_gpua_mrg_graph
  7326.            0.000s - 8 - 8 - local_gpua_subtensor_graph
  7327.            0.000s - 3 - 3 - local_gpua_assert_graph
  7328.            0.000s - 7 - 7 - local_gpua_reshape
  7329.            0.000s - 4 - 4 - local_gpua_dot22
  7330.            0.000s - 1 - 1 - local_gpua_crossentropysoftmaxargmax1hotwithbias
  7331.            0.000s - in 1 optimization that were not used (display only those with a runtime > 0)
  7332.  
  7333.        0.017739s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 603, 574) - 0.000s
  7334.          EquilibriumOptimizer          gpuarray_cut_transfers
  7335.            time 0.018s for 2 passes
  7336.            nb nodes (start, end,  max) 603 574 603
  7337.            time io_toposort 0.011s
  7338.            time in local optimizers 0.004s
  7339.            time in global optimizers 0.000s
  7340.            time in final optimizers 0.000s
  7341.            time in cleanup optimizers 0.000s
  7342.             0 - 0.011s 15 (0.000s in global opts, 0.006s io_toposort) - 603 nodes - ('local_cut_gpu_transfers', 15)
  7343.             1 - 0.007s 0 (0.000s in global opts, 0.005s io_toposort) - 574 nodes -
  7344.            times - times applied - nb node created - name:
  7345.            0.003s - 15 - 0 - local_cut_gpu_transfers
  7346.            0.001s - in 1 optimization that were not used (display only those with a runtime > 0)
  7347.              0.001s - constant_folding
  7348.  
  7349.        0.000043s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 313, 313) - 0.000s
  7350.  
  7351.    0.271405s - ('canonicalize', 'EquilibriumOptimizer', 6, 330, 340) - 0.002s
  7352.      EquilibriumOptimizer      canonicalize
  7353.        time 0.271s for 4 passes
  7354.        nb nodes (start, end,  max) 330 340 360
  7355.        time io_toposort 0.015s
  7356.        time in local optimizers 0.199s
  7357.        time in global optimizers 0.000s
  7358.        time in final optimizers 0.030s
  7359.        time in cleanup optimizers 0.014s
  7360.         0 - 0.137s 83 (0.016s in global opts, 0.004s io_toposort) - 330 nodes - ('MergeOptimizer', 31) ('local_add_canonizer', 10) ('local_mul_canonizer', 10) ('local_shape_to_shape_i', 6) ('local_subtensor_make_vector', 5) ...
  7361.         1 - 0.064s 46 (0.007s in global opts, 0.004s io_toposort) - 360 nodes - ('MergeOptimizer', 14) ('local_dimshuffle_lift', 7) ('local_subtensor_make_vector', 7) ('local_mul_canonizer', 6) ('local_upcast_elemwise_constant_inputs', 6) ...
  7362.         2 - 0.036s 2 (0.004s in global opts, 0.004s io_toposort) - 340 nodes - ('MergeOptimizer', 1) ('local_mul_canonizer', 1)
  7363.         3 - 0.034s 0 (0.004s in global opts, 0.003s io_toposort) - 340 nodes -
  7364.        times - times applied - nb node created - name:
  7365.        0.045s - 5 - 10 - local_subtensor_merge
  7366.        0.030s - 2 - 0 - topo_constant_folding
  7367.        0.026s - 17 - 23 - local_mul_canonizer
  7368.        0.022s - 11 - 22 - local_add_canonizer
  7369.        0.014s - 46 - 1 - MergeOptimizer
  7370.        0.013s - 3 - 6 - local_reshape_to_dimshuffle
  7371.        0.011s - 9 - 27 - local_upcast_elemwise_constant_inputs
  7372.        0.007s - 11 - 18 - local_dimshuffle_lift
  7373.        0.006s - 6 - 58 - local_shape_to_shape_i
  7374.        0.003s - 12 - 0 - local_subtensor_make_vector
  7375.        0.002s - 3 - 0 - local_intdiv_by_one
  7376.        0.002s - 1 - 0 - local_useless_switch
  7377.        0.001s - 2 - 4 - local_subtensor_lift
  7378.        0.001s - 2 - 0 - local_useless_fill
  7379.        0.000s - 1 - 1 - local_neg_to_mul
  7380.        0.061s - in 74 optimization that were not used (display only those with a runtime > 0)
  7381.          0.016s - local_greedy_distributor
  7382.          0.006s - local_mul_zero
  7383.          0.006s - local_fill_sink
  7384.          0.005s - local_func_inv
  7385.          0.004s - local_useless_elemwise
  7386.          0.003s - local_one_minus_erf2
  7387.          0.003s - local_useless_elemwise_comparison
  7388.          0.003s - local_merge_switch_same_cond
  7389.          0.002s - local_one_minus_erf
  7390.          0.002s - local_track_shape_i
  7391.          0.001s - local_fill_cut
  7392.          0.001s - local_mul_switch_sink
  7393.          0.001s - local_useless_subtensor
  7394.          0.001s - local_expm1
  7395.          0.001s - local_cast_cast
  7396.          0.001s - local_IncSubtensor_serialize
  7397.          0.001s - local_useless_slice
  7398.          0.001s - local_cut_gpu_transfers
  7399.          0.000s - f
  7400.          0.000s - local_zero_div
  7401.          0.000s - local_abs_lift
  7402.          0.000s - local_subtensor_remove_broadcastable_index
  7403.          0.000s - local_lift_transpose_through_dot
  7404.          0.000s - local_div_switch_sink
  7405.          0.000s - local_dimshuffle_no_inplace_at_canonicalize
  7406.          0.000s - local_pow_canonicalize
  7407.          0.000s - local_0_dot_x
  7408.          0.000s - local_canonicalize_alloc
  7409.          0.000s - local_useless_dimshuffle_in_reshape
  7410.          0.000s - local_useless_reshape
  7411.          0.000s - local_reshape_lift
  7412.          0.000s - local_sum_prod_div_dimshuffle
  7413.          0.000s - local_subtensor_of_alloc
  7414.          0.000s - local_subtensor_of_dot
  7415.          0.000s - local_useless_alloc
  7416.          0.000s - local_reduce_join
  7417.          0.000s - local_op_of_op
  7418.          0.000s - local_useless_reduce
  7419.          0.000s - local_merge_alloc
  7420.          0.000s - local_sum_prod_all_to_none
  7421.          0.000s - local_sumsqr2dot
  7422.          0.000s - local_scalar_tensor_scalar
  7423.  
  7424.      Global, final and clean up optimizers
  7425.      Iter 0
  7426.              TopoOptimizer  topo_constant_folding
  7427.                nb_node (start, end, changed) (375, 363, 12)
  7428.                init io_toposort 0.00319600105286
  7429.                loop time 0.0124809741974
  7430.                callback_time 0.00716590881348
  7431.              MergeOptimizer
  7432.                nb fail=    0 merged=   74 constant=   50
  7433.                time replace=0.01 validate=0.00 callback=0.01
  7434.      Iter 1
  7435.              TopoOptimizer  topo_constant_folding
  7436.                nb_node (start, end, changed) (343, 340, 3)
  7437.                init io_toposort 0.00289988517761
  7438.                loop time 0.00371909141541
  7439.                callback_time 0.00157237052917
  7440.              MergeOptimizer
  7441.                nb fail=    0 merged=   21 constant=    9
  7442.                time replace=0.00 validate=0.00 callback=0.00
  7443.      Iter 2
  7444.              TopoOptimizer  topo_constant_folding
  7445.                nb_node (start, end, changed) (340, 340, 0)
  7446.                init io_toposort 0.00361585617065
  7447.                loop time 0.000491142272949
  7448.                callback_time 0.0
  7449.              MergeOptimizer
  7450.                nb fail=    0 merged=    1 constant=    1
  7451.                time replace=0.00 validate=0.00 callback=0.00
  7452.      Iter 3
  7453.              TopoOptimizer  topo_constant_folding
  7454.                nb_node (start, end, changed) (340, 340, 0)
  7455.                init io_toposort 0.00327706336975
  7456.                loop time 0.00036096572876
  7457.                callback_time 0.0
  7458.              MergeOptimizer
  7459.                nb fail=    0 merged=    0 constant=    0
  7460.                time replace=0.00 validate=0.00 callback=0.00
  7461.    0.232062s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 313, 313) - 0.137s
  7462.    0.221204s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 420, 327) - 0.000s
  7463.      FusionOptimizer
  7464.       nb_iter 3
  7465.       nb_replacement 36
  7466.       nb_inconsistency_replace 0
  7467.       validate_time 0.000401258468628
  7468.       callback_time 0.00900220870972
  7469.       time_toposort 0.0129661560059
  7470.    0.144045s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 313, 313) - 0.068s
  7471.    0.079509s - ('specialize', 'EquilibriumOptimizer', 13, 343, 339) - 0.000s
  7472.      EquilibriumOptimizer      specialize
  7473.        time 0.079s for 3 passes
  7474.        nb nodes (start, end,  max) 343 339 343
  7475.        time io_toposort 0.010s
  7476.        time in local optimizers 0.040s
  7477.        time in global optimizers 0.012s
  7478.        time in final optimizers 0.009s
  7479.        time in cleanup optimizers 0.000s
  7480.         0 - 0.029s 11 (0.006s in global opts, 0.003s io_toposort) - 343 nodes - ('local_div_to_inv', 6) ('local_mul_specialize', 3) ('local_softmax_with_bias', 1) ('local_argmax_pushdown', 1)
  7481.         1 - 0.025s 1 (0.010s in global opts, 0.003s io_toposort) - 339 nodes - ('crossentropy_to_crossentropy_with_softmax_with_bias', 1)
  7482.         2 - 0.025s 0 (0.006s in global opts, 0.004s io_toposort) - 339 nodes -
  7483.        times - times applied - nb node created - name:
  7484.        0.012s - 1 - 1 - crossentropy_to_crossentropy_with_softmax_with_bias
  7485.        0.003s - 3 - 0 - local_mul_specialize
  7486.        0.003s - 6 - 6 - local_div_to_inv
  7487.        0.001s - 1 - 1 - local_argmax_pushdown
  7488.        0.000s - 1 - 1 - local_softmax_with_bias
  7489.        0.043s - in 70 optimization that were not used (display only those with a runtime > 0)
  7490.          0.009s - topo_constant_folding
  7491.          0.005s - local_add_specialize
  7492.          0.003s - local_func_inv
  7493.          0.003s - local_useless_elemwise
  7494.          0.003s - local_elemwise_alloc
  7495.          0.003s - local_one_minus_erf2
  7496.          0.002s - local_one_minus_erf
  7497.          0.002s - local_useless_elemwise_comparison
  7498.          0.001s - local_track_shape_i
  7499.          0.001s - local_abs_merge
  7500.          0.001s - local_mul_switch_sink
  7501.          0.001s - local_useless_switch
  7502.          0.001s - local_elemwise_sub_zeros
  7503.          0.001s - local_expm1
  7504.          0.001s - local_useless_subtensor
  7505.          0.001s - local_logsoftmax
  7506.          0.001s - local_cast_cast
  7507.          0.001s - local_alloc_unary
  7508.          0.000s - local_dimshuffle_lift
  7509.          0.000s - local_mul_to_sqr
  7510.          0.000s - local_remove_useless_assert
  7511.          0.000s - local_useless_slice
  7512.          0.000s - local_pow_specialize
  7513.          0.000s - local_subtensor_remove_broadcastable_index
  7514.          0.000s - local_sum_prod_mul_by_scalar
  7515.          0.000s - local_zero_div
  7516.          0.000s - local_sum_prod_div_dimshuffle
  7517.          0.000s - local_subtensor_make_vector
  7518.          0.000s - local_grad_log_erfc_neg
  7519.          0.000s - local_reduce_broadcastable
  7520.          0.000s - local_subtensor_merge
  7521.          0.000s - local_subtensor_of_dot
  7522.          0.000s - local_subtensor_of_alloc
  7523.          0.000s - local_sumsqr2dot
  7524.          0.000s - local_opt_alloc
  7525.          0.000s - local_scalar_tensor_scalar
  7526.  
  7527.      Global, final and clean up optimizers
  7528.      Iter 0
  7529.              TopoOptimizer  topo_constant_folding
  7530.                nb_node (start, end, changed) (340, 340, 0)
  7531.                init io_toposort 0.0026798248291
  7532.                loop time 0.000288963317871
  7533.                callback_time 0.0
  7534.      Iter 1
  7535.              TopoOptimizer  topo_constant_folding
  7536.                nb_node (start, end, changed) (339, 339, 0)
  7537.                init io_toposort 0.00300598144531
  7538.                loop time 0.000308036804199
  7539.                callback_time 0.0
  7540.      Iter 2
  7541.              TopoOptimizer  topo_constant_folding
  7542.                nb_node (start, end, changed) (339, 339, 0)
  7543.                init io_toposort 0.00263500213623
  7544.                loop time 0.000291109085083
  7545.                callback_time 0.0
  7546.    0.058268s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 346, 343) - 0.000s
  7547.      EquilibriumOptimizer      scan_eqopt2
  7548.        time 0.058s for 2 passes
  7549.        nb nodes (start, end,  max) 346 343 346
  7550.        time io_toposort 0.005s
  7551.        time in local optimizers 0.000s
  7552.        time in global optimizers 0.052s
  7553.        time in final optimizers 0.000s
  7554.        time in cleanup optimizers 0.000s
  7555.         0 - 0.031s 1 (0.027s in global opts, 0.003s io_toposort) - 343 nodes - ('constant_folding', 1)
  7556.         1 - 0.028s 0 (0.024s in global opts, 0.003s io_toposort) - 343 nodes -
  7557.        times - times applied - nb node created - name:
  7558.        0.009s - 1 - 0 - constant_folding
  7559.        0.043s - in 6 optimization that were not used (display only those with a runtime > 0)
  7560.          0.008s - <theano.scan_module.scan_opt.ScanMerge object at 0x10f69f750>
  7561.          0.008s - remove_constants_and_unused_inputs_scan
  7562.          0.007s - scan_merge_inouts
  7563.          0.007s - remove_constants_and_unused_inputs_scan
  7564.          0.006s - remove_constants_and_unused_inputs_scan
  7565.          0.006s - <theano.scan_module.scan_opt.ScanSaveMem object at 0x10f69fa10>
  7566.  
  7567.      Global, final and clean up optimizers
  7568.      Iter 0
  7569.              TopoOptimizer  constant_folding_for_scan2
  7570.                nb_node (start, end, changed) (346, 343, 3)
  7571.                init io_toposort 0.00319814682007
  7572.                loop time 0.00264716148376
  7573.                callback_time 0.000752687454224
  7574.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  7575.                nb_node (start, end, changed) (343, 343, 0)
  7576.                init io_toposort 0.00274419784546
  7577.                loop time 0.000353097915649
  7578.                callback_time 0.0
  7579.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  7580.                nb_node (start, end, changed) (343, 343, 0)
  7581.                init io_toposort 0.00378894805908
  7582.                loop time 0.000379085540771
  7583.                callback_time 0.0
  7584.              TopoOptimizer  scanOp_merge_inouts
  7585.                nb_node (start, end, changed) (343, 343, 0)
  7586.                init io_toposort 0.0033597946167
  7587.                loop time 0.000405073165894
  7588.                callback_time 0.0
  7589.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  7590.                nb_node (start, end, changed) (343, 343, 0)
  7591.                init io_toposort 0.00262212753296
  7592.                loop time 0.000285148620605
  7593.                callback_time 0.0
  7594.      Iter 1
  7595.              TopoOptimizer  constant_folding_for_scan2
  7596.                nb_node (start, end, changed) (343, 343, 0)
  7597.                init io_toposort 0.0027289390564
  7598.                loop time 0.000293016433716
  7599.                callback_time 0.0
  7600.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  7601.                nb_node (start, end, changed) (343, 343, 0)
  7602.                init io_toposort 0.00342202186584
  7603.                loop time 0.000372171401978
  7604.                callback_time 0.0
  7605.              TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  7606.                nb_node (start, end, changed) (343, 343, 0)
  7607.                init io_toposort 0.00345611572266
  7608.                loop time 0.000340938568115
  7609.                callback_time 0.0
  7610.              TopoOptimizer  scanOp_merge_inouts
  7611.                nb_node (start, end, changed) (343, 343, 0)
  7612.                init io_toposort 0.00299596786499
  7613.                loop time 0.000335931777954
  7614.                callback_time 0.0
  7615.              TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  7616.                nb_node (start, end, changed) (343, 343, 0)
  7617.                init io_toposort 0.0031099319458
  7618.                loop time 0.000379085540771
  7619.                callback_time 0.0
  7620.    0.054531s - ('ShapeOpt', 'ShapeOptimizer', 2, 330, 330) - 0.000s
  7621.    0.040722s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 313, 313) - 0.005s
  7622.      TopoOptimizer  dimshuffle_as_view
  7623.        nb_node (start, end, changed) (313, 313, 51)
  7624.        init io_toposort 0.00292992591858
  7625.        loop time 0.0377469062805
  7626.        callback_time 0.0264596939087
  7627.    0.038850s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 313, 313) - 0.025s
  7628.      TopoOptimizer  local_dnna_conv_inplace
  7629.        nb_node (start, end, changed) (313, 313, 13)
  7630.        init io_toposort 0.00262403488159
  7631.        loop time 0.0361239910126
  7632.        callback_time 0.0301699638367
  7633.  LocalOptGroup
  7634.  ---------------------
  7635.    time taken - times applied - times tried - name - node_created:
  7636.    -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  7637.    -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  7638.    -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  7639.    0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  7640.  
  7641.    0.024305s - ('BlasOpt', 'SeqOptimizer', 12, 343, 343) - 0.000s
  7642.      SeqOptimizer      BlasOpt  time 0.024s for 343/343 nodes before/after optimization
  7643.        0.000s for callback
  7644.            0.000s for fgraph.validate()
  7645.        0.004798s - ('gemm_optimizer', 'GemmOptimizer', 1, 343, 343) - 0.000s
  7646.          GemmOptimizer
  7647.           nb_iter 1
  7648.           nb_replacement 0
  7649.           nb_replacement_didn_t_remove 0
  7650.           nb_inconsistency_make 0
  7651.           nb_inconsistency_replace 0
  7652.           time_canonicalize 0.000886917114258
  7653.           time_factor_can 0
  7654.           time_factor_list 0
  7655.           time_toposort 0.00261306762695
  7656.           validate_time 0.0
  7657.           callback_time 0.0
  7658.        0.004567s - ('use_c_blas', 'TopoOptimizer', 4, 343, 343) - 0.000s
  7659.          TopoOptimizer  use_c_blas
  7660.            nb_node (start, end, changed) (343, 343, 0)
  7661.            init io_toposort 0.00274109840393
  7662.            loop time 0.00178909301758
  7663.            callback_time 0.0
  7664.  LocalOptGroup
  7665.  ---------------------
  7666.  --- The Optimizer wasn't successful ---
  7667.  
  7668.       0.004229s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 343, 343) - 0.000s
  7669.         TopoOptimizer  local_dot_to_dot22
  7670.           nb_node (start, end, changed) (343, 343, 4)
  7671.           init io_toposort 0.00279211997986
  7672.           loop time 0.00139617919922
  7673.           callback_time 0.000430583953857
  7674.       0.003866s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 343, 343) - 0.000s
  7675.         EquilibriumOptimizer          local_gemm_to_gemv
  7676.           time 0.004s for 1 passes
  7677.           nb nodes (start, end,  max) 343 343 343
  7678.           time io_toposort 0.003s
  7679.           time in local optimizers 0.000s
  7680.           time in global optimizers 0.000s
  7681.           time in final optimizers 0.000s
  7682.           time in cleanup optimizers 0.000s
  7683.            0 - 0.004s 0 (0.000s in global opts, 0.003s io_toposort) - 343 nodes -
  7684.       0.003740s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 343, 343) - 0.000s
  7685.         TopoOptimizer  local_dot22_to_dot22scalar
  7686.           nb_node (start, end, changed) (343, 343, 0)
  7687.           init io_toposort 0.00274991989136
  7688.           loop time 0.000944852828979
  7689.           callback_time 0.0
  7690.       0.002956s - ('use_scipy_ger', 'TopoOptimizer', 5, 343, 343) - 0.000s
  7691.         TopoOptimizer  scipy_blas
  7692.           nb_node (start, end, changed) (343, 343, 0)
  7693.           init io_toposort 0.00257515907288
  7694.           loop time 0.000352144241333
  7695.           callback_time 0.0
  7696.  
  7697.   0.021485s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 313, 313) - 0.015s
  7698.     TopoOptimizer  random_make_inplace_mrg
  7699.       nb_node (start, end, changed) (313, 313, 3)
  7700.       init io_toposort 0.00355195999146
  7701.       loop time 0.0178859233856
  7702.       callback_time 0.0168223381042
  7703.   0.021211s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 330, 330) - 0.000s
  7704.     EquilibriumOptimizer      scan_eqopt1
  7705.       time 0.021s for 1 passes
  7706.       nb nodes (start, end,  max) 330 330 330
  7707.       time io_toposort 0.003s
  7708.       time in local optimizers 0.000s
  7709.       time in global optimizers 0.018s
  7710.       time in final optimizers 0.000s
  7711.       time in cleanup optimizers 0.000s
  7712.        0 - 0.021s 0 (0.018s in global opts, 0.003s io_toposort) - 330 nodes -
  7713.     Global, final and clean up optimizers
  7714.     Iter 0
  7715.             SeqOptimizer              all_pushout_opt  time 0.018s for 330/330 nodes before/after optimization
  7716.               0.000s for callback
  7717.                   0.000s for fgraph.validate()
  7718.               0.004939s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 330, 330) - 0.000s
  7719.               0.004040s - ('scan_pushout_dot1', 'PushOutDot1', 3, 330, 330) - 0.000s
  7720.               0.003186s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 330, 330) - 0.000s
  7721.               0.002830s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 330, 330) - 0.000s
  7722.                 TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  7723.                   nb_node (start, end, changed) (330, 330, 0)
  7724.                   init io_toposort 0.00245404243469
  7725.                   loop time 0.000336885452271
  7726.                   callback_time 0.0
  7727.               0.002699s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 330, 330) - 0.000s
  7728.  
  7729.   0.018380s - ('stabilize', 'EquilibriumOptimizer', 8, 340, 340) - 0.000s
  7730.     EquilibriumOptimizer      stabilize
  7731.       time 0.018s for 1 passes
  7732.       nb nodes (start, end,  max) 340 340 340
  7733.       time io_toposort 0.003s
  7734.       time in local optimizers 0.008s
  7735.       time in global optimizers 0.003s
  7736.       time in final optimizers 0.003s
  7737.       time in cleanup optimizers 0.000s
  7738.        0 - 0.018s 0 (0.006s in global opts, 0.003s io_toposort) - 340 nodes -
  7739.     Global, final and clean up optimizers
  7740.     Iter 0
  7741.             TopoOptimizer  topo_constant_folding
  7742.               nb_node (start, end, changed) (340, 340, 0)
  7743.               init io_toposort 0.00290107727051
  7744.               loop time 0.000303030014038
  7745.               callback_time 0.0
  7746.   0.014653s - ('add_destroy_handler', 'AddDestroyHandler', 23, 313, 313) - 0.000s
  7747.   0.014369s - ('merge2', 'MergeOptimizer', 22, 327, 313) - 0.002s
  7748.     MergeOptimizer
  7749.       nb fail=    0 merged=  192 constant=  144
  7750.       time replace=0.01 validate=0.00 callback=0.01
  7751.   0.010946s - ('merge1', 'MergeOptimizer', 0, 385, 330) - 0.001s
  7752.     MergeOptimizer
  7753.       nb fail=    0 merged=  109 constant=   54
  7754.       time replace=0.01 validate=0.00 callback=0.00
  7755.   0.008074s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 339, 339) - 0.000s
  7756.     EquilibriumOptimizer      uncanonicalize
  7757.       time 0.008s for 1 passes
  7758.       nb nodes (start, end,  max) 339 339 339
  7759.       time io_toposort 0.003s
  7760.       time in local optimizers 0.000s
  7761.       time in global optimizers 0.000s
  7762.       time in final optimizers 0.005s
  7763.       time in cleanup optimizers 0.000s
  7764.        0 - 0.008s 0 (0.005s in global opts, 0.003s io_toposort) - 339 nodes -
  7765.     Global, final and clean up optimizers
  7766.     Iter 0
  7767.             TopoOptimizer  topo_constant_folding
  7768.               nb_node (start, end, changed) (339, 339, 0)
  7769.               init io_toposort 0.00420188903809
  7770.               loop time 0.000434875488281
  7771.               callback_time 0.0
  7772.   0.007642s - ('useless', 'TopoOptimizer', 3, 330, 330) - 0.000s
  7773.     TopoOptimizer  useless
  7774.       nb_node (start, end, changed) (330, 330, 0)
  7775.       init io_toposort 0.00319695472717
  7776.       loop time 0.00438714027405
  7777.       callback_time 0.0
  7778. LocalOptGroup
  7779. ---------------------
  7780.   time taken - times applied - times tried - name - node_created:
  7781.   -0.000s - 0 - 3 - local_merge_alloc - 0
  7782.   -0.000s - 0 - 3 - local_useless_alloc - 0
  7783.   -0.000s - 0 - 3 - local_useless_reduce - 0
  7784.   -0.000s - 0 - 7 - local_useless_reshape - 0
  7785.   -0.000s - 0 - 22 - local_subtensor_of_alloc - 0
  7786.   -0.000s - 0 - 22 - local_subtensor_make_vector - 0
  7787.   -0.000s - 0 - 137 - local_useless_switch - 0
  7788.   -0.000s - 0 - 22 - local_useless_slice - 0
  7789.   -0.000s - 0 - 137 - local_useless_elemwise_comparison - 0
  7790.   -0.001s - 0 - 137 - local_useless_elemwise - 0
  7791.   0.000s - in 9 optimization that were not used (display those with runtime greater than 0)
  7792.  
  7793.   0.007514s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 420, 420) - 0.000s
  7794.     FusionOptimizer
  7795.      nb_iter 1
  7796.      nb_replacement 0
  7797.      nb_inconsistency_replace 0
  7798.      validate_time 0.0
  7799.      callback_time 0.0
  7800.      time_toposort 0.00684809684753
  7801.   0.006168s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 313, 313) - 0.000s
  7802.     TopoOptimizer  InplaceGpuaBlasOpt
  7803.       nb_node (start, end, changed) (313, 313, 0)
  7804.       init io_toposort 0.0039050579071
  7805.       loop time 0.00214195251465
  7806.       callback_time 0.0
  7807. LocalOptGroup
  7808. ---------------------
  7809.   time taken - times applied - times tried - name - node_created:
  7810.   -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  7811.   0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  7812.  
  7813.   0.006007s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 340, 346) - 0.000s
  7814.     TopoOptimizer  local_elemwise_alloc
  7815.       nb_node (start, end, changed) (340, 346, 3)
  7816.       init io_toposort 0.00270009040833
  7817.       loop time 0.00326800346375
  7818.       callback_time 0.000940084457397
  7819.   0.005795s - ('specialize_device', 'EquilibriumOptimizer', 17, 574, 574) - 0.000s
  7820.     EquilibriumOptimizer      specialize_device
  7821.       time 0.006s for 1 passes
  7822.       nb nodes (start, end,  max) 574 574 574
  7823.       time io_toposort 0.005s
  7824.       time in local optimizers 0.000s
  7825.       time in global optimizers 0.000s
  7826.       time in final optimizers 0.000s
  7827.       time in cleanup optimizers 0.000s
  7828.        0 - 0.006s 0 (0.000s in global opts, 0.005s io_toposort) - 574 nodes -
  7829.   0.005605s - ('AbstractConvCheck', 'TopoOptimizer', 18, 574, 574) - 0.000s
  7830.     TopoOptimizer  AbstractConvCheck
  7831.       nb_node (start, end, changed) (574, 574, 0)
  7832.       init io_toposort 0.00474500656128
  7833.       loop time 0.000811100006104
  7834.       callback_time 0.0
  7835.   0.005336s - ('c_blas_destructive', 'TopoOptimizer', 37, 313, 313) - 0.000s
  7836.     TopoOptimizer  c_blas_destructive
  7837.       nb_node (start, end, changed) (313, 313, 0)
  7838.       init io_toposort 0.00340914726257
  7839.       loop time 0.00188398361206
  7840.       callback_time 0.0
  7841. LocalOptGroup
  7842. ---------------------
  7843. --- The Optimizer wasn't successful ---
  7844.  
  7845.    0.005236s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 313, 313) - 0.000s
  7846.      TopoOptimizer  InplaceGpuBlasOpt
  7847.        nb_node (start, end, changed) (313, 313, 0)
  7848.        init io_toposort 0.00278186798096
  7849.        loop time 0.00232291221619
  7850.        callback_time 0.0
  7851.  LocalOptGroup
  7852.  ---------------------
  7853.  --- The Optimizer wasn't successful ---
  7854.  
  7855.   0.005101s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 313, 313) - 0.000s
  7856.   0.004823s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 313, 313) - 0.000s
  7857.   0.004742s - ('blas_opt_inplace', 'TopoOptimizer', 34, 313, 313) - 0.000s
  7858.     TopoOptimizer  InplaceBlasOpt
  7859.       nb_node (start, end, changed) (313, 313, 0)
  7860.       init io_toposort 0.00268507003784
  7861.       loop time 0.00196695327759
  7862.       callback_time 0.0
  7863. LocalOptGroup
  7864. ---------------------
  7865. --- The Optimizer wasn't successful ---
  7866.  
  7867.    0.004588s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 313, 313) - 0.000s
  7868.      TopoOptimizer  local_gemm16_inplace
  7869.        nb_node (start, end, changed) (313, 313, 0)
  7870.        init io_toposort 0.00413489341736
  7871.        loop time 0.000356912612915
  7872.        callback_time 0.0
  7873.    0.004424s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 313, 313) - 0.000s
  7874.      TopoOptimizer  local_dnn_conv_inplace
  7875.        nb_node (start, end, changed) (313, 313, 0)
  7876.        init io_toposort 0.00266098976135
  7877.        loop time 0.00166201591492
  7878.        callback_time 0.0
  7879.  LocalOptGroup
  7880.  ---------------------
  7881.  --- The Optimizer wasn't successful ---
  7882.  
  7883.   0.004422s - ('random_make_inplace', 'TopoOptimizer', 49, 313, 313) - 0.000s
  7884.     TopoOptimizer  random_make_inplace
  7885.       nb_node (start, end, changed) (313, 313, 0)
  7886.       init io_toposort 0.00393199920654
  7887.       loop time 0.000427007675171
  7888.       callback_time 0.0
  7889.   0.004200s - ('make_ger_destructive', 'TopoOptimizer', 41, 313, 313) - 0.000s
  7890.     TopoOptimizer  make_scipy_blas_destructive
  7891.       nb_node (start, end, changed) (313, 313, 0)
  7892.       init io_toposort 0.00368809700012
  7893.       loop time 0.000468015670776
  7894.       callback_time 0.0
  7895.   0.004082s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 313, 313) - 0.000s
  7896.     TopoOptimizer  local_advincsub1_gpua_inplace
  7897.       nb_node (start, end, changed) (313, 313, 0)
  7898.       init io_toposort 0.00373101234436
  7899.       loop time 0.000315189361572
  7900.       callback_time 0.0
  7901.   0.003584s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 313, 313) - 0.000s
  7902.     TopoOptimizer  local_inplace_incsubtensor1
  7903.       nb_node (start, end, changed) (313, 313, 0)
  7904.       init io_toposort 0.00316691398621
  7905.       loop time 0.000317811965942
  7906.       callback_time 0.0
  7907.   0.003561s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 330, 330) - 0.000s
  7908.     TopoOptimizer  pre_local_IncSubtensor_serialize
  7909.       nb_node (start, end, changed) (330, 330, 0)
  7910.       init io_toposort 0.00282597541809
  7911.       loop time 0.000701904296875
  7912.       callback_time 0.0
  7913.   0.003460s - ('local_destructive', 'TopoOptimizer', 48, 313, 313) - 0.000s
  7914.     TopoOptimizer  CURAND_destructive
  7915.       nb_node (start, end, changed) (313, 313, 0)
  7916.       init io_toposort 0.00305008888245
  7917.       loop time 0.000365972518921
  7918.       callback_time 0.0
  7919.   0.003311s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 340, 340) - 0.000s
  7920.     TopoOptimizer  local_fill_to_alloc
  7921.       nb_node (start, end, changed) (340, 340, 0)
  7922.       init io_toposort 0.00266814231873
  7923.       loop time 0.000611066818237
  7924.       callback_time 0.0
  7925.   0.003242s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 313, 313) - 0.000s
  7926.   0.003147s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 313, 313) - 0.000s
  7927.     TopoOptimizer  local_inplace_setsubtensor
  7928.       nb_node (start, end, changed) (313, 313, 0)
  7929.       init io_toposort 0.00282001495361
  7930.       loop time 0.000253915786743
  7931.       callback_time 0.0
  7932.   0.003041s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 313, 313) - 0.000s
  7933.     TopoOptimizer  local_inplace_gpu_sparse_block_outer
  7934.       nb_node (start, end, changed) (313, 313, 0)
  7935.       init io_toposort 0.00271677970886
  7936.       loop time 0.00026798248291
  7937.       callback_time 0.0
  7938.   0.003027s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 313, 313) - 0.000s
  7939.     TopoOptimizer  local_inplace_sparseblockgemv
  7940.       nb_node (start, end, changed) (313, 313, 0)
  7941.       init io_toposort 0.00271415710449
  7942.       loop time 0.000258922576904
  7943.       callback_time 0.0
  7944.   0.002971s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 313, 313) - 0.000s
  7945.     TopoOptimizer  local_inplace_sparseblockouter
  7946.       nb_node (start, end, changed) (313, 313, 0)
  7947.       init io_toposort 0.00265693664551
  7948.       loop time 0.000259160995483
  7949.       callback_time 0.0
  7950.   0.002969s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 313, 313) - 0.000s
  7951.     TopoOptimizer  local_inplace_sparse_block_outer
  7952.       nb_node (start, end, changed) (313, 313, 0)
  7953.       init io_toposort 0.00266313552856
  7954.       loop time 0.000251054763794
  7955.       callback_time 0.0
  7956.   0.002954s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 313, 313) - 0.000s
  7957.     TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  7958.       nb_node (start, end, changed) (313, 313, 0)
  7959.       init io_toposort 0.00263500213623
  7960.       loop time 0.000256061553955
  7961.       callback_time 0.0
  7962.   0.002952s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 313, 313) - 0.000s
  7963.     TopoOptimizer  local_inplace_sparse_block_gemv
  7964.       nb_node (start, end, changed) (313, 313, 0)
  7965.       init io_toposort 0.00263714790344
  7966.       loop time 0.000253200531006
  7967.       callback_time 0.0
  7968.   0.002938s - ('cond_make_inplace', 'TopoOptimizer', 47, 313, 313) - 0.000s
  7969.     TopoOptimizer  cond_make_inplace
  7970.       nb_node (start, end, changed) (313, 313, 0)
  7971.       init io_toposort 0.00264000892639
  7972.       loop time 0.000258207321167
  7973.       callback_time 0.0
  7974.   0.002860s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 339, 339) - 0.000s
  7975.   0.000331s - ('merge3', 'MergeOptimizer', 51, 313, 313) - 0.000s
  7976.     MergeOptimizer
  7977.       nb fail=    0 merged=    0 constant=    0
  7978.       time replace=0.00 validate=0.00 callback=0.00
  7979.   0.000049s - ('merge1.2', 'MergeOptimizer', 7, 340, 340) - 0.000s
  7980.     MergeOptimizer
  7981.       nb fail=    0 merged=    0 constant=    0
  7982.       time replace=0.00 validate=0.00 callback=0.00
  7983.   0.000028s - ('merge1.1', 'MergeOptimizer', 4, 330, 330) - 0.000s
  7984.     MergeOptimizer
  7985.       nb fail=    0 merged=    0 constant=    0
  7986.       time replace=0.00 validate=0.00 callback=0.00
  7987.  
  7988. Here are tips to potentially make your code run faster
  7989.                 (if you think of new ones, suggest them on the mailing list).
  7990.                 Test them first, as they are not guaranteed to always provide a speedup.
  7991.  Sorry, no tip for today.
  7992. Function profiling
  7993. ==================
  7994.  Message: sb/convnet/sb_resnet.py:349
  7995.  Time in 0 calls to Function.__call__: 0.000000e+00s
  7996.  Total compile time: 3.226146e+00s
  7997.    Number of Apply nodes: 313
  7998.    Theano Optimizer time: 2.992740e+00s
  7999.       Theano validate time: 2.675121e-01s
  8000.    Theano Linker time (includes C, CUDA code generation/compiling): 1.833398e-01s
  8001.       Import time 0.000000e+00s
  8002.       Node make_thunk time 1.669791e-01s
  8003.  
  8004. Time in all call to theano.grad() 2.656322e+00s
  8005. Time since theano import 477.967s
  8006. Optimizer Profile
  8007. -----------------
  8008. SeqOptimizer  OPT_FAST_RUN  time 2.992s for 385/313 nodes before/after optimization
  8009.   0.486s for callback
  8010.       0.268s for fgraph.validate()
  8011.   time      - (name, class, index, nodes before, nodes after) - validate time
  8012.   1.219774s - ('canonicalize', 'EquilibriumOptimizer', 6, 330, 340) - 0.003s
  8013.     EquilibriumOptimizer      canonicalize
  8014.       time 1.219s for 4 passes
  8015.       nb nodes (start, end,  max) 330 340 360
  8016.       time io_toposort 0.011s
  8017.       time in local optimizers 1.152s
  8018.       time in global optimizers 0.000s
  8019.       time in final optimizers 0.030s
  8020.       time in cleanup optimizers 0.013s
  8021.        0 - 0.138s 83 (0.015s in global opts, 0.003s io_toposort) - 330 nodes - ('MergeOptimizer', 31) ('local_add_canonizer', 10) ('local_mul_canonizer', 10) ('local_shape_to_shape_i', 6) ('local_subtensor_make_vector', 5) ...
  8022.        1 - 1.012s 46 (0.007s in global opts, 0.003s io_toposort) - 360 nodes - ('MergeOptimizer', 14) ('local_dimshuffle_lift', 7) ('local_subtensor_make_vector', 7) ('local_mul_canonizer', 6) ('local_upcast_elemwise_constant_inputs', 6) ...
  8023.        2 - 0.036s 2 (0.004s in global opts, 0.003s io_toposort) - 340 nodes - ('MergeOptimizer', 1) ('local_mul_canonizer', 1)
  8024.        3 - 0.033s 0 (0.004s in global opts, 0.003s io_toposort) - 340 nodes -
  8025.       times - times applied - nb node created - name:
  8026.       0.963s - 3 - 6 - local_reshape_to_dimshuffle
  8027.       0.044s - 5 - 10 - local_subtensor_merge
  8028.       0.030s - 2 - 0 - topo_constant_folding
  8029.       0.026s - 17 - 23 - local_mul_canonizer
  8030.       0.022s - 11 - 22 - local_add_canonizer
  8031.       0.013s - 46 - 1 - MergeOptimizer
  8032.       0.011s - 9 - 27 - local_upcast_elemwise_constant_inputs
  8033.       0.007s - 6 - 58 - local_shape_to_shape_i
  8034.       0.007s - 11 - 18 - local_dimshuffle_lift
  8035.       0.003s - 12 - 0 - local_subtensor_make_vector
  8036.       0.002s - 3 - 0 - local_intdiv_by_one
  8037.       0.002s - 1 - 0 - local_useless_switch
  8038.       0.001s - 2 - 4 - local_subtensor_lift
  8039.       0.001s - 2 - 0 - local_useless_fill
  8040.       0.000s - 1 - 1 - local_neg_to_mul
  8041.       0.064s - in 74 optimization that were not used (display only those with a runtime > 0)
  8042.         0.016s - local_greedy_distributor
  8043.         0.007s - local_mul_zero
  8044.         0.006s - local_fill_sink
  8045.         0.005s - local_func_inv
  8046.         0.004s - local_useless_elemwise
  8047.         0.003s - local_one_minus_erf2
  8048.         0.003s - local_useless_elemwise_comparison
  8049.         0.003s - local_merge_switch_same_cond
  8050.         0.002s - local_one_minus_erf
  8051.         0.002s - local_track_shape_i
  8052.         0.001s - local_fill_cut
  8053.         0.001s - local_mul_switch_sink
  8054.         0.001s - local_useless_subtensor
  8055.         0.001s - local_expm1
  8056.         0.001s - local_cast_cast
  8057.         0.001s - local_IncSubtensor_serialize
  8058.         0.001s - local_cut_gpu_transfers
  8059.         0.001s - local_useless_slice
  8060.         0.000s - local_zero_div
  8061.         0.000s - f
  8062.         0.000s - local_lift_transpose_through_dot
  8063.         0.000s - local_abs_lift
  8064.         0.000s - local_subtensor_remove_broadcastable_index
  8065.         0.000s - local_dimshuffle_no_inplace_at_canonicalize
  8066.         0.000s - local_div_switch_sink
  8067.         0.000s - local_pow_canonicalize
  8068.         0.000s - local_0_dot_x
  8069.         0.000s - local_canonicalize_alloc
  8070.         0.000s - local_useless_reshape
  8071.         0.000s - local_reshape_lift
  8072.         0.000s - local_useless_dimshuffle_in_reshape
  8073.         0.000s - local_subtensor_of_alloc
  8074.         0.000s - local_subtensor_of_dot
  8075.         0.000s - local_sum_prod_div_dimshuffle
  8076.         0.000s - local_useless_alloc
  8077.         0.000s - local_scalar_tensor_scalar
  8078.         0.000s - local_op_of_op
  8079.         0.000s - local_merge_alloc
  8080.         0.000s - local_sum_prod_all_to_none
  8081.         0.000s - local_useless_reduce
  8082.         0.000s - local_sumsqr2dot
  8083.         0.000s - local_reduce_join
  8084.  
  8085.     Global, final and clean up optimizers
  8086.     Iter 0
  8087.             TopoOptimizer  topo_constant_folding
  8088.               nb_node (start, end, changed) (375, 363, 12)
  8089.               init io_toposort 0.00306487083435
  8090.               loop time 0.012228012085
  8091.               callback_time 0.00723528862
  8092.             MergeOptimizer
  8093.               nb fail=    0 merged=   74 constant=   50
  8094.               time replace=0.01 validate=0.00 callback=0.01
  8095.     Iter 1
  8096.             TopoOptimizer  topo_constant_folding
  8097.               nb_node (start, end, changed) (343, 340, 3)
  8098.               init io_toposort 0.00357699394226
  8099.               loop time 0.00341296195984
  8100.               callback_time 0.00148725509644
  8101.             MergeOptimizer
  8102.               nb fail=    0 merged=   21 constant=    9
  8103.               time replace=0.00 validate=0.00 callback=0.00
  8104.     Iter 2
  8105.             TopoOptimizer  topo_constant_folding
  8106.               nb_node (start, end, changed) (340, 340, 0)
  8107.               init io_toposort 0.00325512886047
  8108.               loop time 0.000423908233643
  8109.               callback_time 0.0
  8110.             MergeOptimizer
  8111.               nb fail=    0 merged=    1 constant=    1
  8112.               time replace=0.00 validate=0.00 callback=0.00
  8113.     Iter 3
  8114.             TopoOptimizer  topo_constant_folding
  8115.               nb_node (start, end, changed) (340, 340, 0)
  8116.               init io_toposort 0.00319790840149
  8117.               loop time 0.000303983688354
  8118.               callback_time 0.0
  8119.             MergeOptimizer
  8120.               nb fail=    0 merged=    0 constant=    0
  8121.               time replace=0.00 validate=0.00 callback=0.00
  8122.   0.340164s - ('elemwise_fusion', 'SeqOptimizer', 19, 574, 420) - 0.001s
  8123.     SeqOptimizer      elemwise_fusion  time 0.340s for 574/420 nodes before/after optimization
  8124.       0.011s for callback
  8125.           0.001s for fgraph.validate()
  8126.       0.316551s - ('composite_elemwise_fusion', 'FusionOptimizer', 1, 561, 420) - 0.000s
  8127.         FusionOptimizer
  8128.          nb_iter 3
  8129.          nb_replacement 37
  8130.          nb_inconsistency_replace 0
  8131.          validate_time 0.00044584274292
  8132.          callback_time 0.00821328163147
  8133.          time_toposort 0.0146968364716
  8134.       0.023436s - ('local_add_mul_fusion', 'FusionOptimizer', 0, 574, 561) - 0.000s
  8135.         FusionOptimizer
  8136.          nb_iter 2
  8137.          nb_replacement 23
  8138.          nb_inconsistency_replace 0
  8139.          validate_time 0.000319242477417
  8140.          callback_time 0.00327849388123
  8141.          time_toposort 0.0101230144501
  8142.  
  8143.   0.290148s - ('gpuarray_opt', 'SeqOptimizer', 16, 339, 574) - 0.001s
  8144.     SeqOptimizer      gpuarray_opt  time 0.290s for 339/574 nodes before/after optimization
  8145.       0.100s for callback
  8146.           0.001s for fgraph.validate()
  8147.       0.164832s - ('gpuarray_local_optimizations', 'EquilibriumOptimizer', 2, 313, 603) - 0.001s
  8148.         EquilibriumOptimizer          gpuarray_local_optimizations
  8149.           time 0.164s for 3 passes
  8150.           nb nodes (start, end,  max) 313 603 638
  8151.           time io_toposort 0.016s
  8152.           time in local optimizers 0.138s
  8153.           time in global optimizers 0.000s
  8154.           time in final optimizers 0.000s
  8155.           time in cleanup optimizers 0.000s
  8156.            0 - 0.108s 55 (0.000s in global opts, 0.003s io_toposort) - 313 nodes - ('constant_folding', 28) ('local_gpua_elemwise', 13) ('local_abstractconv_cudnn', 13) ('local_gpua_dimshuffle', 1)
  8157.            1 - 0.037s 35 (0.000s in global opts, 0.006s io_toposort) - 638 nodes - ('constant_folding', 35)
  8158.            2 - 0.020s 0 (0.000s in global opts, 0.008s io_toposort) - 603 nodes -
  8159.           times - times applied - nb node created - name:
  8160.           0.054s - 13 - 328 - local_abstractconv_cudnn
  8161.           0.054s - 63 - 0 - constant_folding
  8162.           0.014s - 13 - 49 - local_gpua_elemwise
  8163.           0.001s - 1 - 3 - local_gpua_dimshuffle
  8164.           0.015s - in 66 optimization that were not used (display only those with a runtime > 0)
  8165.             0.002s - local_track_shape_i
  8166.             0.001s - local_gpua_gemm_alpha_merge
  8167.             0.001s - local_dnn_conv_output_merge
  8168.             0.001s - local_gpua_gemm_output_merge
  8169.             0.001s - local_gpua_gemmbatch_alpha_merge
  8170.             0.001s - local_gemm16_alpha_merge
  8171.             0.001s - local_dnn_conv_alpha_merge
  8172.             0.001s - local_dnn_convi_alpha_merge
  8173.             0.001s - local_dnn_convw_alpha_merge
  8174.             0.001s - local_dnn_convw_output_merge
  8175.             0.001s - local_gpua_gemmbatch_output_merge
  8176.             0.001s - local_dnn_convi_output_merge
  8177.             0.001s - local_gemm16_output_merge
  8178.             0.001s - local_log_softmax_dnn
  8179.             0.000s - local_gpua_shape
  8180.             0.000s - local_gpu_contiguous_gpu_contiguous
  8181.             0.000s - local_gpua_abstractconv2d
  8182.             0.000s - local_gpua_assert
  8183.             0.000s - local_gpu_elemwise_careduce
  8184.  
  8185.       0.105904s - ('gpuarray_graph_optimization', 'GraphToGPU', 0, 339, 313) - 0.000s
  8186.         GraphToGPUOptimizer          gpuarray_graph_optimization
  8187.           time io_toposort 0.004s
  8188.         Total time taken by local optimizers 0.010s
  8189.           times - times applied - Node created - name:
  8190.           0.006s - 3 - 3 - local_gpua_careduce
  8191.           0.003s - 120 - 154 - local_gpua_elemwise
  8192.           0.001s - 13 - 13 - local_gpua_lift_abstractconv2d_graph
  8193.           0.001s - 50 - 50 - local_gpua_dimshuffle
  8194.           0.000s - 3 - 6 - local_gpua_mrg_graph
  8195.           0.000s - 8 - 8 - local_gpua_subtensor_graph
  8196.           0.000s - 3 - 3 - local_gpua_assert_graph
  8197.           0.000s - 7 - 7 - local_gpua_reshape
  8198.           0.000s - 4 - 4 - local_gpua_dot22
  8199.           0.000s - 1 - 1 - local_gpua_crossentropysoftmaxargmax1hotwithbias
  8200.           0.000s - in 1 optimization that were not used (display only those with a runtime > 0)
  8201.  
  8202.       0.019174s - ('gpuarray_cut_transfers', 'EquilibriumOptimizer', 3, 603, 574) - 0.000s
  8203.         EquilibriumOptimizer          gpuarray_cut_transfers
  8204.           time 0.019s for 2 passes
  8205.           nb nodes (start, end,  max) 603 574 603
  8206.           time io_toposort 0.012s
  8207.           time in local optimizers 0.004s
  8208.           time in global optimizers 0.000s
  8209.           time in final optimizers 0.000s
  8210.           time in cleanup optimizers 0.000s
  8211.            0 - 0.012s 15 (0.000s in global opts, 0.007s io_toposort) - 603 nodes - ('local_cut_gpu_transfers', 15)
  8212.            1 - 0.007s 0 (0.000s in global opts, 0.005s io_toposort) - 574 nodes -
  8213.           times - times applied - nb node created - name:
  8214.           0.003s - 15 - 0 - local_cut_gpu_transfers
  8215.           0.001s - in 1 optimization that were not used (display only those with a runtime > 0)
  8216.             0.001s - constant_folding
  8217.  
  8218.       0.000047s - ('InputToGpuArrayOptimizer', 'InputToGpuOptimizer', 1, 313, 313) - 0.000s
  8219.  
  8220.   0.230584s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 45, 313, 313) - 0.137s
  8221.   0.229299s - ('gpua_elemwise_fusion', 'FusionOptimizer', 21, 420, 327) - 0.000s
  8222.     FusionOptimizer
  8223.      nb_iter 3
  8224.      nb_replacement 36
  8225.      nb_inconsistency_replace 0
  8226.      validate_time 0.000416994094849
  8227.      callback_time 0.00961184501648
  8228.      time_toposort 0.0111560821533
  8229.   0.156492s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 43, 313, 313) - 0.076s
  8230.   0.071392s - ('specialize', 'EquilibriumOptimizer', 13, 343, 339) - 0.000s
  8231.     EquilibriumOptimizer      specialize
  8232.       time 0.071s for 3 passes
  8233.       nb nodes (start, end,  max) 343 339 343
  8234.       time io_toposort 0.008s
  8235.       time in local optimizers 0.035s
  8236.       time in global optimizers 0.012s
  8237.       time in final optimizers 0.009s
  8238.       time in cleanup optimizers 0.000s
  8239.        0 - 0.024s 11 (0.006s in global opts, 0.003s io_toposort) - 343 nodes - ('local_div_to_inv', 6) ('local_mul_specialize', 3) ('local_softmax_with_bias', 1) ('local_argmax_pushdown', 1)
  8240.        1 - 0.025s 1 (0.010s in global opts, 0.003s io_toposort) - 339 nodes - ('crossentropy_to_crossentropy_with_softmax_with_bias', 1)
  8241.        2 - 0.022s 0 (0.006s in global opts, 0.003s io_toposort) - 339 nodes -
  8242.       times - times applied - nb node created - name:
  8243.       0.012s - 1 - 1 - crossentropy_to_crossentropy_with_softmax_with_bias
  8244.       0.003s - 3 - 0 - local_mul_specialize
  8245.       0.002s - 6 - 6 - local_div_to_inv
  8246.       0.001s - 1 - 1 - local_argmax_pushdown
  8247.       0.000s - 1 - 1 - local_softmax_with_bias
  8248.       0.039s - in 70 optimization that were not used (display only those with a runtime > 0)
  8249.         0.009s - topo_constant_folding
  8250.         0.005s - local_add_specialize
  8251.         0.003s - local_func_inv
  8252.         0.003s - local_elemwise_alloc
  8253.         0.003s - local_useless_elemwise
  8254.         0.002s - local_one_minus_erf2
  8255.         0.002s - local_one_minus_erf
  8256.         0.002s - local_useless_elemwise_comparison
  8257.         0.001s - local_track_shape_i
  8258.         0.001s - local_abs_merge
  8259.         0.001s - local_mul_switch_sink
  8260.         0.001s - local_useless_switch
  8261.         0.001s - local_expm1
  8262.         0.001s - local_elemwise_sub_zeros
  8263.         0.001s - local_logsoftmax
  8264.         0.001s - local_useless_subtensor
  8265.         0.001s - local_cast_cast
  8266.         0.001s - local_dimshuffle_lift
  8267.         0.001s - local_alloc_unary
  8268.         0.000s - local_mul_to_sqr
  8269.         0.000s - local_remove_useless_assert
  8270.         0.000s - local_useless_slice
  8271.         0.000s - local_pow_specialize
  8272.         0.000s - local_subtensor_remove_broadcastable_index
  8273.         0.000s - local_subtensor_make_vector
  8274.         0.000s - local_zero_div
  8275.         0.000s - local_sum_prod_mul_by_scalar
  8276.         0.000s - local_subtensor_merge
  8277.         0.000s - local_sum_prod_div_dimshuffle
  8278.         0.000s - local_subtensor_of_alloc
  8279.         0.000s - local_grad_log_erfc_neg
  8280.         0.000s - local_subtensor_of_dot
  8281.         0.000s - local_reduce_broadcastable
  8282.         0.000s - local_sumsqr2dot
  8283.         0.000s - local_opt_alloc
  8284.         0.000s - local_scalar_tensor_scalar
  8285.  
  8286.     Global, final and clean up optimizers
  8287.     Iter 0
  8288.             TopoOptimizer  topo_constant_folding
  8289.               nb_node (start, end, changed) (340, 340, 0)
  8290.               init io_toposort 0.00276398658752
  8291.               loop time 0.000319004058838
  8292.               callback_time 0.0
  8293.     Iter 1
  8294.             TopoOptimizer  topo_constant_folding
  8295.               nb_node (start, end, changed) (339, 339, 0)
  8296.               init io_toposort 0.00275802612305
  8297.               loop time 0.000292062759399
  8298.               callback_time 0.0
  8299.     Iter 2
  8300.             TopoOptimizer  topo_constant_folding
  8301.               nb_node (start, end, changed) (339, 339, 0)
  8302.               init io_toposort 0.00278401374817
  8303.               loop time 0.000298976898193
  8304.               callback_time 0.0
  8305.   0.055384s - ('scan_eqopt2', 'EquilibriumOptimizer', 11, 346, 343) - 0.000s
  8306.     EquilibriumOptimizer      scan_eqopt2
  8307.       time 0.055s for 2 passes
  8308.       nb nodes (start, end,  max) 346 343 346
  8309.       time io_toposort 0.006s
  8310.       time in local optimizers 0.000s
  8311.       time in global optimizers 0.048s
  8312.       time in final optimizers 0.000s
  8313.       time in cleanup optimizers 0.000s
  8314.        0 - 0.031s 1 (0.027s in global opts, 0.003s io_toposort) - 343 nodes - ('constant_folding', 1)
  8315.        1 - 0.024s 0 (0.021s in global opts, 0.003s io_toposort) - 343 nodes -
  8316.       times - times applied - nb node created - name:
  8317.       0.009s - 1 - 0 - constant_folding
  8318.       0.039s - in 6 optimization that were not used (display only those with a runtime > 0)
  8319.         0.007s - scan_merge_inouts
  8320.         0.007s - <theano.scan_module.scan_opt.ScanSaveMem object at 0x10f69fa10>
  8321.         0.007s - remove_constants_and_unused_inputs_scan
  8322.         0.006s - remove_constants_and_unused_inputs_scan
  8323.         0.006s - remove_constants_and_unused_inputs_scan
  8324.         0.006s - <theano.scan_module.scan_opt.ScanMerge object at 0x10f69f750>
  8325.  
  8326.     Global, final and clean up optimizers
  8327.     Iter 0
  8328.             TopoOptimizer  constant_folding_for_scan2
  8329.               nb_node (start, end, changed) (346, 343, 3)
  8330.               init io_toposort 0.00302219390869
  8331.               loop time 0.00299000740051
  8332.               callback_time 0.000831604003906
  8333.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  8334.               nb_node (start, end, changed) (343, 343, 0)
  8335.               init io_toposort 0.0029149055481
  8336.               loop time 0.000319004058838
  8337.               callback_time 0.0
  8338.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  8339.               nb_node (start, end, changed) (343, 343, 0)
  8340.               init io_toposort 0.00260496139526
  8341.               loop time 0.000286817550659
  8342.               callback_time 0.0
  8343.             TopoOptimizer  scanOp_merge_inouts
  8344.               nb_node (start, end, changed) (343, 343, 0)
  8345.               init io_toposort 0.0032901763916
  8346.               loop time 0.000686883926392
  8347.               callback_time 0.0
  8348.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  8349.               nb_node (start, end, changed) (343, 343, 0)
  8350.               init io_toposort 0.00342392921448
  8351.               loop time 0.000342845916748
  8352.               callback_time 0.0
  8353.     Iter 1
  8354.             TopoOptimizer  constant_folding_for_scan2
  8355.               nb_node (start, end, changed) (343, 343, 0)
  8356.               init io_toposort 0.00266885757446
  8357.               loop time 0.000299215316772
  8358.               callback_time 0.0
  8359.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs1
  8360.               nb_node (start, end, changed) (343, 343, 0)
  8361.               init io_toposort 0.002690076828
  8362.               loop time 0.000297069549561
  8363.               callback_time 0.0
  8364.             TopoOptimizer  scanop_remove_constants_and_unused_inputs2
  8365.               nb_node (start, end, changed) (343, 343, 0)
  8366.               init io_toposort 0.00262594223022
  8367.               loop time 0.000285863876343
  8368.               callback_time 0.0
  8369.             TopoOptimizer  scanOp_merge_inouts
  8370.               nb_node (start, end, changed) (343, 343, 0)
  8371.               init io_toposort 0.00301003456116
  8372.               loop time 0.000369071960449
  8373.               callback_time 0.0
  8374.             TopoOptimizer  scanOp_remove_constants_and_unused_inputs3
  8375.               nb_node (start, end, changed) (343, 343, 0)
  8376.               init io_toposort 0.00259804725647
  8377.               loop time 0.000284910202026
  8378.               callback_time 0.0
  8379.   0.050506s - ('ShapeOpt', 'ShapeOptimizer', 2, 330, 330) - 0.000s
  8380.   0.041654s - ('dimshuffle_as_view', 'TopoOptimizer', 24, 313, 313) - 0.005s
  8381.     TopoOptimizer  dimshuffle_as_view
  8382.       nb_node (start, end, changed) (313, 313, 51)
  8383.       init io_toposort 0.00297808647156
  8384.       loop time 0.0386300086975
  8385.       callback_time 0.0270938873291
  8386.   0.039686s - ('local_dnna_conv_inplace', 'TopoOptimizer', 39, 313, 313) - 0.026s
  8387.     TopoOptimizer  local_dnna_conv_inplace
  8388.       nb_node (start, end, changed) (313, 313, 13)
  8389.       init io_toposort 0.00298881530762
  8390.       loop time 0.0365929603577
  8391.       callback_time 0.0307540893555
  8392. LocalOptGroup
  8393. ---------------------
  8394.   time taken - times applied - times tried - name - node_created:
  8395.   -0.004s - 36 - 72 - local_dnn_convgi_inplace - 36
  8396.   -0.004s - 43 - 86 - local_dnn_convgw_inplace - 43
  8397.   -0.005s - 55 - 110 - local_dnn_conv_inplace - 56
  8398.   0.000s - in 0 optimization that were not used (display those with runtime greater than 0)
  8399.  
  8400.   0.024131s - ('BlasOpt', 'SeqOptimizer', 12, 343, 343) - 0.000s
  8401.     SeqOptimizer      BlasOpt  time 0.024s for 343/343 nodes before/after optimization
  8402.       0.000s for callback
  8403.           0.000s for fgraph.validate()
  8404.       0.004802s - ('gemm_optimizer', 'GemmOptimizer', 1, 343, 343) - 0.000s
  8405.         GemmOptimizer
  8406.          nb_iter 1
  8407.          nb_replacement 0
  8408.          nb_replacement_didn_t_remove 0
  8409.          nb_inconsistency_make 0
  8410.          nb_inconsistency_replace 0
  8411.          time_canonicalize 0.000934362411499
  8412.          time_factor_can 0
  8413.          time_factor_list 0
  8414.          time_toposort 0.00260996818542
  8415.          validate_time 0.0
  8416.          callback_time 0.0
  8417.       0.004657s - ('use_c_blas', 'TopoOptimizer', 4, 343, 343) - 0.000s
  8418.         TopoOptimizer  use_c_blas
  8419.           nb_node (start, end, changed) (343, 343, 0)
  8420.           init io_toposort 0.00270009040833
  8421.           loop time 0.00190901756287
  8422.           callback_time 0.0
  8423. LocalOptGroup
  8424. ---------------------
  8425. --- The Optimizer wasn't successful ---
  8426.  
  8427.        0.004111s - ('local_dot_to_dot22', 'TopoOptimizer', 0, 343, 343) - 0.000s
  8428.          TopoOptimizer  local_dot_to_dot22
  8429.            nb_node (start, end, changed) (343, 343, 4)
  8430.            init io_toposort 0.00284814834595
  8431.            loop time 0.0012309551239
  8432.            callback_time 0.000388622283936
  8433.        0.003664s - ('local_gemm_to_gemv', 'EquilibriumOptimizer', 3, 343, 343) - 0.000s
  8434.          EquilibriumOptimizer          local_gemm_to_gemv
  8435.            time 0.004s for 1 passes
  8436.            nb nodes (start, end,  max) 343 343 343
  8437.            time io_toposort 0.003s
  8438.            time in local optimizers 0.000s
  8439.            time in global optimizers 0.000s
  8440.            time in final optimizers 0.000s
  8441.            time in cleanup optimizers 0.000s
  8442.             0 - 0.004s 0 (0.000s in global opts, 0.003s io_toposort) - 343 nodes -
  8443.        0.003414s - ('local_dot22_to_dot22scalar', 'TopoOptimizer', 2, 343, 343) - 0.000s
  8444.          TopoOptimizer  local_dot22_to_dot22scalar
  8445.            nb_node (start, end, changed) (343, 343, 0)
  8446.            init io_toposort 0.00264620780945
  8447.            loop time 0.000739097595215
  8448.            callback_time 0.0
  8449.        0.003338s - ('use_scipy_ger', 'TopoOptimizer', 5, 343, 343) - 0.000s
  8450.          TopoOptimizer  scipy_blas
  8451.            nb_node (start, end, changed) (343, 343, 0)
  8452.            init io_toposort 0.00294184684753
  8453.            loop time 0.000363111495972
  8454.            callback_time 0.0
  8455.  
  8456.    0.021946s - ('stabilize', 'EquilibriumOptimizer', 8, 340, 340) - 0.000s
  8457.      EquilibriumOptimizer      stabilize
  8458.        time 0.022s for 1 passes
  8459.        nb nodes (start, end,  max) 340 340 340
  8460.        time io_toposort 0.004s
  8461.        time in local optimizers 0.008s
  8462.        time in global optimizers 0.006s
  8463.        time in final optimizers 0.003s
  8464.        time in cleanup optimizers 0.000s
  8465.         0 - 0.022s 0 (0.009s in global opts, 0.004s io_toposort) - 340 nodes -
  8466.      Global, final and clean up optimizers
  8467.      Iter 0
  8468.              TopoOptimizer  topo_constant_folding
  8469.                nb_node (start, end, changed) (340, 340, 0)
  8470.                init io_toposort 0.00267601013184
  8471.                loop time 0.000298976898193
  8472.                callback_time 0.0
  8473.    0.020503s - ('mrg_random_make_inplace', 'TopoOptimizer', 50, 313, 313) - 0.015s
  8474.      TopoOptimizer  random_make_inplace_mrg
  8475.        nb_node (start, end, changed) (313, 313, 3)
  8476.        init io_toposort 0.00290107727051
  8477.        loop time 0.0175631046295
  8478.        callback_time 0.0165855884552
  8479.    0.018693s - ('merge2', 'MergeOptimizer', 22, 327, 313) - 0.002s
  8480.      MergeOptimizer
  8481.        nb fail=    0 merged=  192 constant=  144
  8482.        time replace=0.02 validate=0.00 callback=0.01
  8483.    0.017984s - ('scan_eqopt1', 'EquilibriumOptimizer', 1, 330, 330) - 0.000s
  8484.      EquilibriumOptimizer      scan_eqopt1
  8485.        time 0.018s for 1 passes
  8486.        nb nodes (start, end,  max) 330 330 330
  8487.        time io_toposort 0.003s
  8488.        time in local optimizers 0.000s
  8489.        time in global optimizers 0.015s
  8490.        time in final optimizers 0.000s
  8491.        time in cleanup optimizers 0.000s
  8492.         0 - 0.018s 0 (0.015s in global opts, 0.003s io_toposort) - 330 nodes -
  8493.      Global, final and clean up optimizers
  8494.      Iter 0
  8495.              SeqOptimizer              all_pushout_opt  time 0.015s for 330/330 nodes before/after optimization
  8496.                0.000s for callback
  8497.                    0.000s for fgraph.validate()
  8498.                0.003534s - ('remove_constants_and_unused_inputs_scan', 'TopoOptimizer', 0, 330, 330) - 0.000s
  8499.                  TopoOptimizer  scanOp_remove_constants_and_unused_inputs0
  8500.                    nb_node (start, end, changed) (330, 330, 0)
  8501.                    init io_toposort 0.00314712524414
  8502.                    loop time 0.000341892242432
  8503.                    callback_time 0.0
  8504.                0.003078s - ('scanOp_pushout_nonseqs_ops', 'PushOutNonSeqScan', 1, 330, 330) - 0.000s
  8505.                0.002669s - ('scanOp_pushout_seqs_ops', 'PushOutSeqScan', 2, 330, 330) - 0.000s
  8506.                0.002653s - ('scanOp_pushout_output', 'PushOutScanOutput', 4, 330, 330) - 0.000s
  8507.                0.002631s - ('scan_pushout_dot1', 'PushOutDot1', 3, 330, 330) - 0.000s
  8508.  
  8509.    0.014251s - ('add_destroy_handler', 'AddDestroyHandler', 23, 313, 313) - 0.000s
  8510.    0.010826s - ('merge1', 'MergeOptimizer', 0, 385, 330) - 0.001s
  8511.      MergeOptimizer
  8512.        nb fail=    0 merged=  109 constant=   54
  8513.        time replace=0.01 validate=0.00 callback=0.00
  8514.    0.007993s - ('local_elemwise_alloc', 'TopoOptimizer', 10, 340, 346) - 0.000s
  8515.      TopoOptimizer  local_elemwise_alloc
  8516.        nb_node (start, end, changed) (340, 346, 3)
  8517.        init io_toposort 0.00397491455078
  8518.        loop time 0.0039701461792
  8519.        callback_time 0.00116753578186
  8520.    0.006738s - ('useless', 'TopoOptimizer', 3, 330, 330) - 0.000s
  8521.      TopoOptimizer  useless
  8522.        nb_node (start, end, changed) (330, 330, 0)
  8523.        init io_toposort 0.00299000740051
  8524.        loop time 0.00368285179138
  8525.        callback_time 0.0
  8526.  LocalOptGroup
  8527.  ---------------------
  8528.    time taken - times applied - times tried - name - node_created:
  8529.    -0.000s - 0 - 3 - local_merge_alloc - 0
  8530.    -0.000s - 0 - 3 - local_useless_reduce - 0
  8531.    -0.000s - 0 - 3 - local_useless_alloc - 0
  8532.    -0.000s - 0 - 7 - local_useless_reshape - 0
  8533.    -0.000s - 0 - 22 - local_subtensor_of_alloc - 0
  8534.    -0.000s - 0 - 22 - local_subtensor_make_vector - 0
  8535.    -0.000s - 0 - 137 - local_useless_switch - 0
  8536.    -0.000s - 0 - 22 - local_useless_slice - 0
  8537.    -0.000s - 0 - 137 - local_useless_elemwise_comparison - 0
  8538.    -0.001s - 0 - 137 - local_useless_elemwise - 0
  8539.    0.000s - in 9 optimization that were not used (display those with runtime greater than 0)
  8540.  
  8541.    0.006556s - ('uncanonicalize', 'EquilibriumOptimizer', 15, 339, 339) - 0.000s
  8542.      EquilibriumOptimizer      uncanonicalize
  8543.        time 0.006s for 1 passes
  8544.        nb nodes (start, end,  max) 339 339 339
  8545.        time io_toposort 0.003s
  8546.        time in local optimizers 0.000s
  8547.        time in global optimizers 0.000s
  8548.        time in final optimizers 0.003s
  8549.        time in cleanup optimizers 0.000s
  8550.         0 - 0.006s 0 (0.003s in global opts, 0.003s io_toposort) - 339 nodes -
  8551.      Global, final and clean up optimizers
  8552.      Iter 0
  8553.              TopoOptimizer  topo_constant_folding
  8554.                nb_node (start, end, changed) (339, 339, 0)
  8555.                init io_toposort 0.00266122817993
  8556.                loop time 0.000303983688354
  8557.                callback_time 0.0
  8558.    0.005904s - ('local_inplace_sparse_block_outer', 'TopoOptimizer', 31, 313, 313) - 0.000s
  8559.      TopoOptimizer  local_inplace_sparse_block_outer
  8560.        nb_node (start, end, changed) (313, 313, 0)
  8561.        init io_toposort 0.0051109790802
  8562.        loop time 0.000679016113281
  8563.        callback_time 0.0
  8564.    0.005808s - ('specialize_device', 'EquilibriumOptimizer', 17, 574, 574) - 0.000s
  8565.      EquilibriumOptimizer      specialize_device
  8566.        time 0.006s for 1 passes
  8567.        nb nodes (start, end,  max) 574 574 574
  8568.        time io_toposort 0.005s
  8569.        time in local optimizers 0.000s
  8570.        time in global optimizers 0.000s
  8571.        time in final optimizers 0.000s
  8572.        time in cleanup optimizers 0.000s
  8573.         0 - 0.006s 0 (0.000s in global opts, 0.005s io_toposort) - 574 nodes -
  8574.    0.005567s - ('scanOp_make_inplace', 'ScanInplaceOptimizer', 46, 313, 313) - 0.000s
  8575.    0.005320s - ('AbstractConvCheck', 'TopoOptimizer', 18, 574, 574) - 0.000s
  8576.      TopoOptimizer  AbstractConvCheck
  8577.        nb_node (start, end, changed) (574, 574, 0)
  8578.        init io_toposort 0.00467395782471
  8579.        loop time 0.000613927841187
  8580.        callback_time 0.0
  8581.    0.005308s - ('blas_opt_inplace', 'TopoOptimizer', 34, 313, 313) - 0.000s
  8582.      TopoOptimizer  InplaceBlasOpt
  8583.        nb_node (start, end, changed) (313, 313, 0)
  8584.        init io_toposort 0.00336408615112
  8585.        loop time 0.00183486938477
  8586.        callback_time 0.0
  8587.  LocalOptGroup
  8588.  ---------------------
  8589.  --- The Optimizer wasn't successful ---
  8590.  
  8591.   0.004798s - ('cond_make_inplace', 'TopoOptimizer', 47, 313, 313) - 0.000s
  8592.     TopoOptimizer  cond_make_inplace
  8593.       nb_node (start, end, changed) (313, 313, 0)
  8594.       init io_toposort 0.0043408870697
  8595.       loop time 0.000392913818359
  8596.       callback_time 0.0
  8597.   0.004715s - ('local_advincsub1_gpua_inplace', 'TopoOptimizer', 25, 313, 313) - 0.000s
  8598.     TopoOptimizer  local_advincsub1_gpua_inplace
  8599.       nb_node (start, end, changed) (313, 313, 0)
  8600.       init io_toposort 0.00419616699219
  8601.       loop time 0.000470161437988
  8602.       callback_time 0.0
  8603.   0.004631s - ('local_dnn_conv_inplace', 'TopoOptimizer', 38, 313, 313) - 0.000s
  8604.     TopoOptimizer  local_dnn_conv_inplace
  8605.       nb_node (start, end, changed) (313, 313, 0)
  8606.       init io_toposort 0.00288701057434
  8607.       loop time 0.00163793563843
  8608.       callback_time 0.0
  8609. LocalOptGroup
  8610. ---------------------
  8611. --- The Optimizer wasn't successful ---
  8612.  
  8613.    0.004615s - ('InplaceGpuBlasOpt', 'TopoOptimizer', 35, 313, 313) - 0.000s
  8614.      TopoOptimizer  InplaceGpuBlasOpt
  8615.        nb_node (start, end, changed) (313, 313, 0)
  8616.        init io_toposort 0.00274896621704
  8617.        loop time 0.00176906585693
  8618.        callback_time 0.0
  8619.  LocalOptGroup
  8620.  ---------------------
  8621.  --- The Optimizer wasn't successful ---
  8622.  
  8623.   0.004609s - ('local_inplace_sparseblockgemv', 'TopoOptimizer', 32, 313, 313) - 0.000s
  8624.     TopoOptimizer  local_inplace_sparseblockgemv
  8625.       nb_node (start, end, changed) (313, 313, 0)
  8626.       init io_toposort 0.0040340423584
  8627.       loop time 0.000416040420532
  8628.       callback_time 0.0
  8629.   0.004513s - ('gpu_elemwise_fusion', 'FusionOptimizer', 20, 420, 420) - 0.000s
  8630.     FusionOptimizer
  8631.      nb_iter 1
  8632.      nb_replacement 0
  8633.      nb_inconsistency_replace 0
  8634.      validate_time 0.0
  8635.      callback_time 0.0
  8636.      time_toposort 0.00397276878357
  8637.   0.004498s - ('c_blas_destructive', 'TopoOptimizer', 37, 313, 313) - 0.000s
  8638.     TopoOptimizer  c_blas_destructive
  8639.       nb_node (start, end, changed) (313, 313, 0)
  8640.       init io_toposort 0.00261497497559
  8641.       loop time 0.00184512138367
  8642.       callback_time 0.0
  8643. LocalOptGroup
  8644. ---------------------
  8645. --- The Optimizer wasn't successful ---
  8646.  
  8647.    0.004486s - ('gpua_scanOp_make_inplace', 'ScanInplaceOptimizer', 44, 313, 313) - 0.000s
  8648.    0.004417s - ('gpuablas_opt_inplace', 'TopoOptimizer', 36, 313, 313) - 0.000s
  8649.      TopoOptimizer  InplaceGpuaBlasOpt
  8650.        nb_node (start, end, changed) (313, 313, 0)
  8651.        init io_toposort 0.00274109840393
  8652.        loop time 0.00157904624939
  8653.        callback_time 0.0
  8654.  LocalOptGroup
  8655.  ---------------------
  8656.    time taken - times applied - times tried - name - node_created:
  8657.    -0.001s - 10 - 20 - local_inplace_gpuagemm - 10
  8658.    0.000s - in 2 optimization that were not used (display those with runtime greater than 0)
  8659.  
  8660.    0.004336s - ('local_fill_to_alloc', 'TopoOptimizer', 9, 340, 340) - 0.000s
  8661.      TopoOptimizer  local_fill_to_alloc
  8662.        nb_node (start, end, changed) (340, 340, 0)
  8663.        init io_toposort 0.0031270980835
  8664.        loop time 0.00114893913269
  8665.        callback_time 0.0
  8666.    0.004204s - ('local_destructive', 'TopoOptimizer', 48, 313, 313) - 0.000s
  8667.      TopoOptimizer  CURAND_destructive
  8668.        nb_node (start, end, changed) (313, 313, 0)
  8669.        init io_toposort 0.00382399559021
  8670.        loop time 0.000334978103638
  8671.        callback_time 0.0
  8672.    0.004054s - ('local_inplace_gpu_sparse_block_outer', 'TopoOptimizer', 27, 313, 313) - 0.000s
  8673.      TopoOptimizer  local_inplace_gpu_sparse_block_outer
  8674.        nb_node (start, end, changed) (313, 313, 0)
  8675.        init io_toposort 0.00367403030396
  8676.        loop time 0.000312089920044
  8677.        callback_time 0.0
  8678.    0.003904s - ('local_inplace_gpu_sparse_block_gemv', 'TopoOptimizer', 26, 313, 313) - 0.000s
  8679.      TopoOptimizer  local_inplace_gpu_sparse_block_gemv
  8680.        nb_node (start, end, changed) (313, 313, 0)
  8681.        init io_toposort 0.00352501869202
  8682.        loop time 0.000298976898193
  8683.        callback_time 0.0
  8684.    0.003778s - ('random_make_inplace', 'TopoOptimizer', 49, 313, 313) - 0.000s
  8685.      TopoOptimizer  random_make_inplace
  8686.        nb_node (start, end, changed) (313, 313, 0)
  8687.        init io_toposort 0.00336599349976
  8688.        loop time 0.000365018844604
  8689.        callback_time 0.0
  8690.    0.003698s - ('local_inplace_sparseblockouter', 'TopoOptimizer', 33, 313, 313) - 0.000s
  8691.      TopoOptimizer  local_inplace_sparseblockouter
  8692.        nb_node (start, end, changed) (313, 313, 0)
  8693.        init io_toposort 0.00329613685608
  8694.        loop time 0.000308990478516
  8695.        callback_time 0.0
  8696.    0.003278s - ('local_IncSubtensor_serialize', 'TopoOptimizer', 5, 330, 330) - 0.000s
  8697.      TopoOptimizer  pre_local_IncSubtensor_serialize
  8698.        nb_node (start, end, changed) (330, 330, 0)
  8699.        init io_toposort 0.00257587432861
  8700.        loop time 0.000670909881592
  8701.        callback_time 0.0
  8702.    0.003210s - ('make_ger_destructive', 'TopoOptimizer', 41, 313, 313) - 0.000s
  8703.      TopoOptimizer  make_scipy_blas_destructive
  8704.        nb_node (start, end, changed) (313, 313, 0)
  8705.        init io_toposort 0.00282907485962
  8706.        loop time 0.000349044799805
  8707.        callback_time 0.0
  8708.    0.003062s - ('local_gemm16_inplace', 'TopoOptimizer', 40, 313, 313) - 0.000s
  8709.      TopoOptimizer  local_gemm16_inplace
  8710.        nb_node (start, end, changed) (313, 313, 0)
  8711.        init io_toposort 0.00276303291321
  8712.        loop time 0.000236988067627
  8713.        callback_time 0.0
  8714.    0.003057s - ('local_inplace_incsubtensor1', 'TopoOptimizer', 28, 313, 313) - 0.000s
  8715.      TopoOptimizer  local_inplace_incsubtensor1
  8716.        nb_node (start, end, changed) (313, 313, 0)
  8717.        init io_toposort 0.00274205207825
  8718.        loop time 0.00025200843811
  8719.        callback_time 0.0
  8720.    0.002935s - ('local_inplace_sparse_block_gemv', 'TopoOptimizer', 30, 313, 313) - 0.000s
  8721.      TopoOptimizer  local_inplace_sparse_block_gemv
  8722.        nb_node (start, end, changed) (313, 313, 0)
  8723.        init io_toposort 0.00262904167175
  8724.        loop time 0.000252962112427
  8725.        callback_time 0.0
  8726.    0.002929s - ('local_inplace_setsubtensor', 'TopoOptimizer', 29, 313, 313) - 0.000s
  8727.      TopoOptimizer  local_inplace_setsubtensor
  8728.        nb_node (start, end, changed) (313, 313, 0)
  8729.        init io_toposort 0.00262403488159
  8730.        loop time 0.00025200843811
  8731.        callback_time 0.0
  8732.    0.002910s - ('crossentropy_to_crossentropy_with_softmax', 'FromFunctionOptimizer', 14, 339, 339) - 0.000s
  8733.    0.002738s - ('inplace_elemwise_optimizer', 'FromFunctionOptimizer', 42, 313, 313) - 0.000s
  8734.    0.000303s - ('merge3', 'MergeOptimizer', 51, 313, 313) - 0.000s
  8735.      MergeOptimizer
  8736.        nb fail=    0 merged=    0 constant=    0
  8737.        time replace=0.00 validate=0.00 callback=0.00
  8738.    0.000046s - ('merge1.2', 'MergeOptimizer', 7, 340, 340) - 0.000s
  8739.      MergeOptimizer
  8740.        nb fail=    0 merged=    0 constant=    0
  8741.        time replace=0.00 validate=0.00 callback=0.00
  8742.    0.000027s - ('merge1.1', 'MergeOptimizer', 4, 330, 330) - 0.000s
  8743.      MergeOptimizer
  8744.        nb fail=    0 merged=    0 constant=    0
  8745.        time replace=0.00 validate=0.00 callback=0.00
  8746.  
  8747. Here are tips to potentially make your code run faster
  8748.                  (if you think of new ones, suggest them on the mailing list).
  8749.                  Test them first, as they are not guaranteed to always provide a speedup.
  8750.   Sorry, no tip for today.
  8751. Function profiling
  8752. ==================
  8753.   Message: Sum of all(14) printed profiles at exit excluding Scan op profile.
  8754.   Time in 11 calls to Function.__call__: 1.068115e-03s
  8755.   Time in Function.fn.__call__: 9.267330e-04s (86.763%)
  8756.   Time in thunks: 8.780956e-04s (82.210%)
  8757.   Total compile time: 4.684355e+02s
  8758.     Number of Apply nodes: 1
  8759.     Theano Optimizer time: 7.320596e+01s
  8760.        Theano validate time: 1.033737e+01s
  8761.     Theano Linker time (includes C, CUDA code generation/compiling): 3.916804e+02s
  8762.        Import time 1.612234e+00s
  8763.        Node make_thunk time 3.911215e+02s
  8764.  
  8765. Time in all call to theano.grad() 2.656322e+00s
  8766. Time since theano import 478.217s
  8767. Class
  8768. ---
  8769. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Class name>
  8770.   94.8%    94.8%       0.001s       4.16e-04s     C        2       2   theano.gpuarray.basic_ops.HostFromGpu
  8771.    2.3%    97.1%       0.000s       3.38e-06s     C        6       6   theano.compile.ops.DeepCopyOp
  8772.    2.3%    99.3%       0.000s       9.89e-06s     C        2       2   theano.gpuarray.subtensor.GpuSubtensor
  8773.    0.7%   100.0%       0.000s       1.99e-06s     C        3       3   theano.compile.ops.Shape_i
  8774.    ... (remaining 0 Classes account for   0.00%(0.00s) of the runtime)
  8775.  
  8776. Ops
  8777. ---
  8778. <% time> <sum %> <apply time> <time per call> <type> <#call> <#apply> <Op name>
  8779.   94.8%    94.8%       0.001s       4.16e-04s     C        2        2   HostFromGpu(gpuarray)
  8780.    2.3%    97.1%       0.000s       3.38e-06s     C        6        6   DeepCopyOp
  8781.    2.3%    99.3%       0.000s       9.89e-06s     C        2        2   GpuSubtensor{:int64:}
  8782.    0.7%   100.0%       0.000s       1.99e-06s     C        3        3   Shape_i{0}
  8783.    ... (remaining 0 Ops account for   0.00%(0.00s) of the runtime)
  8784.  
  8785. Apply
  8786. ------
  8787. <% time> <sum %> <apply time> <time per call> <#call> <id> <Apply name>
  8788.   88.1%    88.1%       0.001s       7.74e-04s      1     1   HostFromGpu(gpuarray)(GpuSubtensor{:int64:}.0)
  8789.    6.6%    94.8%       0.000s       5.82e-05s      1     1   HostFromGpu(gpuarray)(GpuSubtensor{:int64:}.0)
  8790.    1.8%    96.6%       0.000s       1.60e-05s      1     0   GpuSubtensor{:int64:}(<GpuArrayType<None>(float32, (False, False, False, False))>, Constant{128})
  8791.    0.6%    97.1%       0.000s       5.01e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  8792.    0.5%    97.6%       0.000s       4.05e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  8793.    0.5%    98.1%       0.000s       4.05e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  8794.    0.4%    98.5%       0.000s       3.81e-06s      1     0   GpuSubtensor{:int64:}(<GpuArrayType<None>(float32, (False,))>, Constant{128})
  8795.    0.4%    98.9%       0.000s       3.10e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  8796.    0.2%    99.1%       0.000s       2.15e-06s      1     0   Shape_i{0}(<GpuArrayType<None>(float32, (False,))>)
  8797.    0.2%    99.3%       0.000s       2.15e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  8798.    0.2%    99.6%       0.000s       1.91e-06s      1     0   Shape_i{0}(<GpuArrayType<None>(float32, (False,))>)
  8799.    0.2%    99.8%       0.000s       1.91e-06s      1     0   Shape_i{0}(<GpuArrayType<None>(float32, (False,))>)
  8800.    0.2%   100.0%       0.000s       1.91e-06s      1     0   DeepCopyOp(TensorConstant{-0.577215671539})
  8801.    ... (remaining 0 Apply instances account for 0.00%(0.00s) of the runtime)
  8802.  
  8803. Optimizer Profile
  8804. -----------------
  8805.  SeqOptimizer  time 73.202s for -1/-1 nodes before/after optimization
  8806.    23.222s for callback
  8807.        10.337s for fgraph.validate()
  8808.    callbacks_time
  8809.         <theano.gof.destroyhandler.DestroyHandler object at 0x1291a3f10> , 7.663418293
  8810.         <theano.tensor.opt.ShapeFeature object at 0x121562590> , 4.24110126495
  8811.         Updater{canonicalize} , 3.05698680878
  8812.         <theano.compile.function_module.Supervisor instance at 0x1227d3f38> , 2.56386876106
  8813.         <theano.gof.opt.MergeFeature object at 0x1227e7190> , 1.74654364586
  8814.         Updater{canonicalize} , 1.31303954124
  8815.         Updater{gpuarray_local_optimizations} , 0.261435270309
  8816.         Updater{gpuarray_cut_transfers} , 0.261384248734
  8817.         Updater{canonicalize} , 0.194210767746
  8818.         <theano.gof.destroyhandler.DestroyHandler object at 0x130d80590> , 0.18234872818
  8819.         <theano.gof.destroyhandler.DestroyHandler object at 0x12feccdd0> , 0.177083730698
  8820.         Updater{specialize} , 0.130860090256
  8821.         <theano.gof.toolbox.ReplaceValidate object at 0x12087ee50> , 0.11890411377
  8822.         <theano.tensor.opt.ShapeFeature object at 0x1306d5510> , 0.118015050888
  8823.         <theano.tensor.opt.ShapeFeature object at 0x12e307fd0> , 0.111938238144
  8824.         <theano.compile.function_module.Supervisor instance at 0x130748f38> , 0.0925514698029
  8825.         <theano.compile.function_module.Supervisor instance at 0x12e3986c8> , 0.0896117687225
  8826.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x122195850> , 0.0774388313293
  8827.         Updater{canonicalize} , 0.0659112930298
  8828.         <theano.gof.opt.ChangeTracker instance at 0x11f7d8f38> , 0.0626258850098
  8829.         <theano.gof.opt.MergeFeature object at 0x12e376410> , 0.0525453090668
  8830.         <theano.gof.opt.MergeFeature object at 0x130723d10> , 0.0524771213531
  8831.         Updater{canonicalize} , 0.0376682281494
  8832.         Updater{canonicalize} , 0.0349521636963
  8833.         Updater{gpuarray_local_optimizations} , 0.0294797420502
  8834.         Updater{specialize} , 0.026261806488
  8835.         <theano.gof.opt.ChangeTracker instance at 0x125a3e1b8> , 0.012256860733
  8836.         Updater{local_elemwise_alloc} , 0.00888752937317
  8837.         <theano.gof.opt.ChangeTracker instance at 0x124d8a098> , 0.00651884078979
  8838.         Updater{pre_local_IncSubtensor_serialize} , 0.00562787055969
  8839.         Updater{specialize} , 0.00509214401245
  8840.         <theano.gof.toolbox.ReplaceValidate object at 0x130665490> , 0.00423550605774
  8841.         <theano.gof.toolbox.ReplaceValidate object at 0x12aea9610> , 0.00402593612671
  8842.         Updater{dimshuffle_as_view} , 0.00395131111145
  8843.         Updater{topo_constant_folding} , 0.00292682647705
  8844.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x130723c10> , 0.00266361236572
  8845.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x12e376310> , 0.00266003608704
  8846.         Updater{useless} , 0.00235295295715
  8847.         <theano.gof.opt.ChangeTracker instance at 0x127399f38> , 0.00211262702942
  8848.         Updater{canonicalize} , 0.00179743766785
  8849.         Updater{canonicalize} , 0.00169730186462
  8850.         Updater{canonicalize} , 0.00153088569641
  8851.         Updater{canonicalize} , 0.00149822235107
  8852.         Updater{local_inplace_setsubtensor} , 0.0013701915741
  8853.         <theano.gof.opt.ChangeTracker instance at 0x12f961908> , 0.00130605697632
  8854.         <theano.gof.opt.ChangeTracker instance at 0x1308fe518> , 0.00124955177307
  8855.         <theano.gof.opt.ChangeTracker instance at 0x130782518> , 0.00123405456543
  8856.         Updater{local_dnna_conv_inplace} , 0.00122761726379
  8857.         Updater{specialize} , 0.00121068954468
  8858.         <theano.gof.opt.ChangeTracker instance at 0x12e7055f0> , 0.00116181373596
  8859.         Updater{gpuarray_local_optimizations} , 0.00108671188354
  8860.         Updater{gpuarray_local_optimizations} , 0.00101017951965
  8861.         Updater{gpuarray_cut_transfers} , 0.000955104827881
  8862.         Updater{constant_folding_for_scan2} , 0.000903367996216
  8863.         Updater{gpuarray_cut_transfers} , 0.000853061676025
  8864.         Updater{stabilize} , 0.000802278518677
  8865.         Updater{specialize} , 0.000703573226929
  8866.         Updater{topo_constant_folding} , 0.000659465789795
  8867.         <theano.gof.opt.ChangeTracker instance at 0x124bccfc8> , 0.000591039657593
  8868.         Updater{specialize} , 0.000572443008423
  8869.         <theano.tensor.opt.ShapeFeature object at 0x11b82cf10> , 0.000552177429199
  8870.         <theano.tensor.opt.ShapeFeature object at 0x11b873150> , 0.000530242919922
  8871.         Updater{dimshuffle_as_view} , 0.000411033630371
  8872.         Updater{dimshuffle_as_view} , 0.000401735305786
  8873.         Updater{GemmOptimizer} , 0.000359296798706
  8874.         Updater{local_dot_to_dot22} , 0.000351905822754
  8875.         Updater{topo_constant_folding} , 0.000331401824951
  8876.         <theano.gof.opt.MergeFeature object at 0x11c2d6cd0> , 0.000263452529907
  8877.         <theano.gof.opt.MergeFeature object at 0x11bf9ae10> , 0.000262498855591
  8878.         <theano.gof.opt.ChangeTracker instance at 0x1226b5560> , 0.000254154205322
  8879.         <theano.gof.opt.MergeFeature object at 0x11c315290> , 0.000249147415161
  8880.         <theano.gof.opt.MergeFeature object at 0x11bf8db90> , 0.000229120254517
  8881.         <theano.gof.opt.MergeFeature object at 0x11c787510> , 0.000221729278564
  8882.         <theano.gof.opt.MergeFeature object at 0x11c805b50> , 0.000217199325562
  8883.         Updater{random_make_inplace_mrg} , 0.000191688537598
  8884.         <theano.tensor.opt.ShapeFeature object at 0x11b747410> , 0.000166416168213
  8885.         Updater{gpuarray_local_optimizations} , 0.000160932540894
  8886.         Updater{gpuarray_local_optimizations} , 0.000154256820679
  8887.         <theano.tensor.opt.ShapeFeature object at 0x11c2d6210> , 0.000149726867676
  8888.         <theano.tensor.opt.ShapeFeature object at 0x11b7b4110> , 0.000149488449097
  8889.         <theano.tensor.opt.ShapeFeature object at 0x11bfcd150> , 0.000147342681885
  8890.         Updater{topo_constant_folding} , 0.000140428543091
  8891.         Updater{topo_constant_folding} , 0.000139713287354
  8892.         Updater{topo_constant_folding} , 0.000137805938721
  8893.         <theano.tensor.opt.ShapeFeature object at 0x11c315550> , 0.000135898590088
  8894.         Updater{local_dnna_conv_inplace} , 0.000133037567139
  8895.         <theano.tensor.opt.ShapeFeature object at 0x11b7ed110> , 0.00013279914856
  8896.         <theano.tensor.opt.ShapeFeature object at 0x11bf8d3d0> , 0.00013279914856
  8897.         Updater{local_dnna_conv_inplace} , 0.00013279914856
  8898.         <theano.tensor.opt.ShapeFeature object at 0x11c835d50> , 0.000131607055664
  8899.         <theano.tensor.opt.ShapeFeature object at 0x11bfcd4d0> , 0.000126838684082
  8900.         Updater{local_elemwise_alloc} , 0.000117063522339
  8901.         <theano.gof.destroyhandler.DestroyHandler object at 0x11bf6eed0> , 0.000104904174805
  8902.         Updater{local_elemwise_alloc} , 0.00010085105896
  8903.         Updater{InplaceGpuaBlasOpt} , 9.91821289062e-05
  8904.         <theano.gof.destroyhandler.DestroyHandler object at 0x11b747c90> , 9.58442687988e-05
  8905.         <theano.gof.destroyhandler.DestroyHandler object at 0x11b842050> , 9.48905944824e-05
  8906.         <theano.gof.destroyhandler.DestroyHandler object at 0x11c2b9e90> , 7.89165496826e-05
  8907.         <theano.gof.destroyhandler.DestroyHandler object at 0x11bfcda10> , 7.72476196289e-05
  8908.         Updater{specialize} , 7.67707824707e-05
  8909.         <theano.gof.destroyhandler.DestroyHandler object at 0x11c835bd0> , 7.39097595215e-05
  8910.         <theano.gof.destroyhandler.DestroyHandler object at 0x11b82c650> , 6.69956207275e-05
  8911.         <theano.gof.destroyhandler.DestroyHandler object at 0x11c315690> , 6.67572021484e-05
  8912.         Updater{topo_constant_folding} , 6.60419464111e-05
  8913.         <theano.gof.opt.MergeFeature object at 0x11b72fe10> , 6.55651092529e-05
  8914.         <theano.gof.opt.ChangeTracker instance at 0x12f7506c8> , 6.48498535156e-05
  8915.         <theano.gof.destroyhandler.DestroyHandler object at 0x11b842790> , 6.27040863037e-05
  8916.         <theano.gof.destroyhandler.DestroyHandler object at 0x11b7b4990> , 6.27040863037e-05
  8917.         Updater{specialize} , 6.24656677246e-05
  8918.         <theano.gof.destroyhandler.DestroyHandler object at 0x11b7ed990> , 6.103515625e-05
  8919.         <theano.gof.opt.MergeFeature object at 0x11b79dd10> , 5.98430633545e-05
  8920.         <theano.gof.opt.MergeFeature object at 0x11b7d6d10> , 5.26905059814e-05
  8921.         <theano.gof.opt.ChangeTracker instance at 0x12fd22830> , 5.22136688232e-05
  8922.         Updater{specialize} , 4.02927398682e-05
  8923.         <theano.gof.toolbox.ReplaceValidate object at 0x11bf8de10> , 3.95774841309e-05
  8924.         Updater{random_make_inplace_mrg} , 3.67164611816e-05
  8925.         Updater{canonicalize} , 3.60012054443e-05
  8926.         Updater{random_make_inplace_mrg} , 3.40938568115e-05
  8927.         <theano.gof.opt.ChangeTracker instance at 0x130a69170> , 3.38554382324e-05
  8928.         <theano.gof.toolbox.ReplaceValidate object at 0x11c670310> , 3.24249267578e-05
  8929.         <theano.gof.toolbox.ReplaceValidate object at 0x11bf9ad10> , 3.17096710205e-05
  8930.         <theano.gof.toolbox.ReplaceValidate object at 0x11c824a50> , 3.14712524414e-05
  8931.         <theano.gof.toolbox.ReplaceValidate object at 0x11c2e45d0> , 3.12328338623e-05
  8932.         <theano.gof.opt.ChangeTracker instance at 0x12fad9290> , 3.09944152832e-05
  8933.         Updater{local_dot_to_dot22} , 3.09944152832e-05
  8934.         Updater{canonicalize} , 3.0517578125e-05
  8935.         <theano.gof.opt.ChangeTracker instance at 0x11bfd3950> , 3.00407409668e-05
  8936.         Updater{local_dot_to_dot22} , 2.98023223877e-05
  8937.         <theano.gof.toolbox.ReplaceValidate object at 0x11b79db90> , 2.93254852295e-05
  8938.         Updater{specialize} , 2.93254852295e-05
  8939.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11bf8da50> , 2.93254852295e-05
  8940.         Updater{canonicalize} , 2.93254852295e-05
  8941.         <theano.gof.toolbox.ReplaceValidate object at 0x11b633b50> , 2.86102294922e-05
  8942.         <theano.gof.opt.ChangeTracker instance at 0x11c2f63b0> , 2.83718109131e-05
  8943.         Updater{topo_constant_folding} , 2.8133392334e-05
  8944.         <theano.gof.toolbox.ReplaceValidate object at 0x11c315150> , 2.76565551758e-05
  8945.         Updater{specialize} , 2.76565551758e-05
  8946.         Updater{canonicalize} , 2.67028808594e-05
  8947.         Updater{canonicalize} , 2.64644622803e-05
  8948.         Updater{topo_constant_folding} , 2.55107879639e-05
  8949.         Updater{specialize} , 2.55107879639e-05
  8950.         Updater{canonicalize} , 2.45571136475e-05
  8951.         <theano.gof.opt.ChangeTracker instance at 0x11bfa3518> , 2.43186950684e-05
  8952.         <theano.gof.toolbox.ReplaceValidate object at 0x11b7d6b50> , 2.40802764893e-05
  8953.         <theano.gof.opt.MergeFeature object at 0x11b859d50> , 2.31266021729e-05
  8954.         <theano.gof.opt.ChangeTracker instance at 0x11c867b48> , 2.31266021729e-05
  8955.         <theano.gof.opt.MergeFeature object at 0x11b82cb50> , 2.24113464355e-05
  8956.         <theano.gof.opt.ChangeTracker instance at 0x11c848908> , 2.14576721191e-05
  8957.         Updater{local_dot22_to_dot22scalar} , 2.121925354e-05
  8958.         <theano.gof.opt.ChangeTracker instance at 0x11c31bc68> , 2.02655792236e-05
  8959.         <theano.compile.function_module.Supervisor instance at 0x11c3177e8> , 2.00271606445e-05
  8960.         <theano.compile.function_module.Supervisor instance at 0x11bfc8320> , 1.97887420654e-05
  8961.         <theano.compile.function_module.Supervisor instance at 0x11c2e1ab8> , 1.93119049072e-05
  8962.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11bf9acd0> , 1.93119049072e-05
  8963.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11c805a90> , 1.90734863281e-05
  8964.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11c2d6fd0> , 1.8835067749e-05
  8965.         <theano.compile.function_module.Supervisor instance at 0x11c846680> , 1.83582305908e-05
  8966.         <theano.compile.function_module.Supervisor instance at 0x11c847710> , 1.74045562744e-05
  8967.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11c315110> , 1.74045562744e-05
  8968.         Updater{constant_folding_for_scan2} , 1.71661376953e-05
  8969.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11b72fcd0> , 1.71661376953e-05
  8970.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11c787d90> , 1.6450881958e-05
  8971.         <theano.compile.function_module.Supervisor instance at 0x11bfa1c68> , 1.6450881958e-05
  8972.         Updater{topo_constant_folding} , 1.52587890625e-05
  8973.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11b7d6cd0> , 1.45435333252e-05
  8974.         <theano.gof.toolbox.ReplaceValidate object at 0x11b859b90> , 1.43051147461e-05
  8975.         Updater{topo_constant_folding} , 1.43051147461e-05
  8976.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11b79dcd0> , 1.4066696167e-05
  8977.         Updater{topo_constant_folding} , 1.38282775879e-05
  8978.         <theano.compile.function_module.Supervisor instance at 0x11b794fc8> , 1.31130218506e-05
  8979.         Updater{topo_constant_folding} , 1.28746032715e-05
  8980.         <theano.gof.toolbox.ReplaceValidate object at 0x11b82c990> , 1.21593475342e-05
  8981.         Updater{constant_folding_for_scan2} , 1.21593475342e-05
  8982.         <theano.gof.opt.ChangeTracker instance at 0x12f9d21b8> , 1.21593475342e-05
  8983.         <theano.compile.function_module.Supervisor instance at 0x11b73bef0> , 1.21593475342e-05
  8984.         Updater{canonicalize} , 1.07288360596e-05
  8985.         Updater{topo_constant_folding} , 1.07288360596e-05
  8986.         <theano.gof.opt.ChangeTracker instance at 0x11b7b29e0> , 1.07288360596e-05
  8987.         Updater{canonicalize} , 1.02519989014e-05
  8988.         Updater{canonicalize} , 1.02519989014e-05
  8989.         <theano.compile.function_module.Supervisor instance at 0x11b7e0128> , 1.00135803223e-05
  8990.         Updater{topo_constant_folding} , 9.77516174316e-06
  8991.         Updater{topo_constant_folding} , 9.29832458496e-06
  8992.         <theano.gof.opt.ChangeTracker instance at 0x11b7f0ab8> , 9.05990600586e-06
  8993.         Updater{canonicalize} , 9.05990600586e-06
  8994.         <theano.gof.opt.ChangeTracker instance at 0x11b75a5a8> , 9.05990600586e-06
  8995.         Updater{topo_constant_folding} , 9.05990600586e-06
  8996.         Updater{topo_constant_folding} , 8.82148742676e-06
  8997.         Updater{canonicalize} , 8.10623168945e-06
  8998.         <theano.gof.opt.ChangeTracker instance at 0x12f745e18> , 8.10623168945e-06
  8999.         Updater{topo_constant_folding} , 8.10623168945e-06
  9000.         Updater{canonicalize} , 7.86781311035e-06
  9001.         Updater{canonicalize} , 7.15255737305e-06
  9002.         Updater{topo_constant_folding} , 6.91413879395e-06
  9003.         Updater{topo_constant_folding} , 6.67572021484e-06
  9004.         Updater{canonicalize} , 6.67572021484e-06
  9005.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11b859d10> , 6.19888305664e-06
  9006.         <theano.gof.toolbox.PreserveVariableAttributes object at 0x11b82cb10> , 5.96046447754e-06
  9007.         Updater{gpuarray_local_optimizations} , 5.72204589844e-06
  9008.         <theano.compile.function_module.Supervisor instance at 0x11b82ac20> , 4.05311584473e-06
  9009.         <theano.compile.function_module.Supervisor instance at 0x11b86f4d0> , 4.05311584473e-06
  9010.    time      - (name, class, index, nodes before, nodes after) - validate time
  9011.    21.451680s - ('canonicalize', 'EquilibriumOptimizer', 43, 3590, 3590)
  9012.      EquilibriumOptimizer      
  9013.        time 21.446s for 7 passes
  9014.        nb nodes (start, end,  max) 7639 4559 7639
  9015.        time io_toposort 1.053s
  9016.        time in local optimizers 14.061s
  9017.        time in global optimizers 0.000s
  9018.        time in final optimizers 1.266s
  9019.        time in cleanup optimizers 4.601s
  9020.         0 - 10.811s 4886 (0.457s in global opts, 0.388s io_toposort) - 8311 nodes - ('MergeOptimizer', 1791) ('local_useless_fill', 647) ('local_mul_canonizer', 378) ('local_fill_sink', 315) ('local_neg_to_mul', 308) ...
  9021.         1 - 5.488s 1785 (0.150s in global opts, 0.434s io_toposort) - 7026 nodes - ('MergeOptimizer', 658) ('local_dimshuffle_lift', 230) ('local_mul_canonizer', 218) ('local_fill_sink', 203) ('local_upcast_elemwise_constant_inputs', 143) ...
  9022.         2 - 1.574s 501 (0.064s in global opts, 0.051s io_toposort) - 5477 nodes - ('MergeOptimizer', 144) ('local_fill_sink', 114) ('local_useless_fill', 57) ('local_zero_div', 57) ('local_sum_prod_div_dimshuffle', 56) ...
  9023.         3 - 0.911s 123 (0.056s in global opts, 0.049s io_toposort) - 5248 nodes - ('MergeOptimizer', 60) ('local_dimshuffle_lift', 56) ('local_useless_fill', 3) ('local_mul_zero', 3) ('topo_constant_folding', 1)
  9024.         4 - 0.762s 67 (0.048s in global opts, 0.041s io_toposort) - 4568 nodes - ('MergeOptimizer', 32) ('local_sum_prod_div_dimshuffle', 28) ('local_zero_div', 3) ('local_fill_sink', 3) ('topo_constant_folding', 1)
  9025.         5 - 1.199s 56 (0.441s in global opts, 0.047s io_toposort) - 4559 nodes - ('local_dimshuffle_lift', 28) ('MergeOptimizer', 28)
  9026.         6 - 0.702s 0 (0.051s in global opts, 0.044s io_toposort) - 4559 nodes -
  9027.        times - times applied - nb node created - name:
  9028.        4.601s - 2713 - 25 - MergeOptimizer
  9029.        2.269s - 599 - 1568 - local_mul_canonizer
  9030.        2.156s - 186 - 975 - local_greedy_distributor
  9031.        1.684s - 635 - 790 - local_fill_sink
  9032.        1.536s - 572 - 1780 - local_dimshuffle_lift
  9033.        1.266s - 15 - 0 - topo_constant_folding
  9034.        1.118s - 18 - 36 - local_reshape_to_dimshuffle
  9035.        0.817s - 260 - 1317 - local_mul_zero
  9036.        0.817s - 273 - 445 - local_add_canonizer
  9037.        0.378s - 803 - 0 - local_useless_fill
  9038.        0.360s - 161 - 483 - local_upcast_elemwise_constant_inputs
  9039.        0.330s - 308 - 608 - local_neg_to_mul
  9040.        0.327s - 168 - 504 - local_sum_prod_div_dimshuffle
  9041.        0.230s - 24 - 48 - local_subtensor_merge
  9042.        0.171s - 134 - 0 - local_cut_gpu_transfers
  9043.        0.168s - 3 - 3 - local_useless_elemwise
  9044.        0.152s - 161 - 571 - local_shape_to_shape_i
  9045.        0.136s - 90 - 180 - local_zero_div
  9046.        0.120s - 36 - 108 - local_mul_switch_sink
  9047.        0.089s - 33 - 99 - local_div_switch_sink
  9048.        0.077s - 18 - 9 - local_useless_switch
  9049.        0.065s - 9 - 0 - local_join_1
  9050.        0.053s - 108 - 32 - local_subtensor_make_vector
  9051.        0.049s - 9 - 9 - local_useless_dimshuffle_in_reshape
  9052.        0.035s - 31 - 62 - local_inv_canon
  9053.        0.021s - 19 - 0 - local_pow_canonicalize
  9054.        0.017s - 14 - 28 - local_subtensor_lift
  9055.        0.015s - 18 - 0 - local_intdiv_by_one
  9056.        0.874s - in 61 optimization that were not used (display only those with a runtime > 0)
  9057.          0.189s - local_func_inv
  9058.          0.119s - local_one_minus_erf2
  9059.          0.099s - local_merge_switch_same_cond
  9060.          0.094s - local_useless_elemwise_comparison
  9061.          0.067s - local_track_shape_i
  9062.          0.056s - local_fill_cut
  9063.          0.046s - local_expm1
  9064.          0.041s - local_cast_cast
  9065.          0.034s - local_one_minus_erf
  9066.          0.034s - local_IncSubtensor_serialize
  9067.          0.018s - local_useless_subtensor
  9068.          0.010s - local_sum_prod_all_to_none
  9069.          0.008s - local_lift_transpose_through_dot
  9070.          0.007s - local_useless_slice
  9071.          0.007s - local_op_of_op
  9072.          0.006s - local_useless_reduce
  9073.          0.005s - local_sumsqr2dot
  9074.          0.005s - local_dimshuffle_no_inplace_at_canonicalize
  9075.          0.005s - f
  9076.          0.005s - local_reduce_join
  9077.          0.004s - local_subtensor_remove_broadcastable_index
  9078.          0.003s - local_0_dot_x
  9079.          0.002s - local_abs_lift
  9080.          0.001s - local_useless_reshape
  9081.          0.001s - local_incsubtensor_of_zeros
  9082.          0.001s - local_subtensor_of_dot
  9083.          0.001s - local_subtensor_of_alloc
  9084.          0.001s - local_reshape_lift
  9085.          0.001s - local_canonicalize_alloc
  9086.          0.001s - local_useless_inc_subtensor
  9087.          0.000s - local_useless_inc_subtensor_alloc
  9088.          0.000s - local_useless_alloc
  9089.          0.000s - local_merge_alloc
  9090.          0.000s - local_setsubtensor_of_constants
  9091.          0.000s - local_scalar_tensor_scalar
  9092.  
  9093.      Global, final and clean up optimizers
  9094.      Iter 0
  9095.              TopoOptimizer  topo_constant_folding
  9096.                nb_node (start, end, changed) (2, 0, 2)
  9097.                init io_toposort 3.48091125488e-05
  9098.                loop time 0.000529050827026
  9099.                callback_time 0.000237703323364
  9100. Error in atexit._run_exitfuncs:
  9101. Traceback (most recent call last):
  9102.   File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/atexit.py", line 24, in _run_exitfuncs
  9103.     func(*targs, **kargs)
  9104.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/compile/profiling.py", line 103, in _atexit_print_fn
  9105.     n_apply_to_print=config.profiling.n_apply)
  9106.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/compile/profiling.py", line 1256, in summary
  9107.     self.optimizer_profile[1])
  9108.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/gof/opt.py", line 337, in print_profile
  9109.     level=level + 1)
  9110.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/gof/opt.py", line 2588, in print_profile
  9111.     o.print_profile(stream, prof, level + 2)
  9112.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/gof/opt.py", line 892, in print_profile
  9113.     callback_time, callbacks_time, nb_merged, nb_constant) = prof
  9114. ValueError: too many values to unpack
  9115. Error in sys.exitfunc:
  9116. Traceback (most recent call last):
  9117.   File "/System/Library/Frameworks/Python.framework/Versions/2.7/lib/python2.7/atexit.py", line 24, in _run_exitfuncs
  9118.     func(*targs, **kargs)
  9119.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/compile/profiling.py", line 103, in _atexit_print_fn
  9120.     n_apply_to_print=config.profiling.n_apply)
  9121.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/compile/profiling.py", line 1256, in summary
  9122.     self.optimizer_profile[1])
  9123.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/gof/opt.py", line 337, in print_profile
  9124.     level=level + 1)
  9125.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/gof/opt.py", line 2588, in print_profile
  9126.     o.print_profile(stream, prof, level + 2)
  9127.   File "/Users/Ramana/projects/macvnev/lib/python2.7/site-packages/theano/gof/opt.py", line 892, in print_profile
  9128.     callback_time, callbacks_time, nb_merged, nb_constant) = prof
  9129. ValueError: too many values to unpack
  9130. Segmentation fault: 11
Advertisement
Add Comment
Please, Sign In to add comment
Advertisement