diff --git "a/WhisperEncoder.mlmodelc/model.espresso.net" "b/WhisperEncoder.mlmodelc/model.espresso.net"
new file mode 100644--- /dev/null
+++ "b/WhisperEncoder.mlmodelc/model.espresso.net"
@@ -0,0 +1,6883 @@
+{
+  "storage" : "model.espresso.weights",
+  "analyses" : {
+
+  },
+  "properties" : {
+
+  },
+  "format_version" : 200,
+  "metadata_in_weights" : [
+
+  ],
+  "layers" : [
+    {
+      "size_of_axes" : 1,
+      "bottom" : "input_features",
+      "axes_0" : -2,
+      "weights" : {
+
+      },
+      "nd_axis" : 0,
+      "debug_info" : "expand_dims_0",
+      "top" : "expand_dims_0",
+      "type" : "expand_dims",
+      "name" : "expand_dims_0"
+    },
+    {
+      "pad_r" : 1,
+      "fused_relu" : 0,
+      "fused_tanh" : 0,
+      "debug_info" : "conv_0",
+      "pad_fill_mode" : 0,
+      "pad_b" : 0,
+      "pad_l" : 1,
+      "top" : "conv_0",
+      "blob_weights" : 3,
+      "K" : 80,
+      "blob_biases" : 1,
+      "name" : "conv_0",
+      "has_batch_norm" : 0,
+      "type" : "convolution",
+      "n_groups" : 1,
+      "pad_t" : 0,
+      "has_biases" : 1,
+      "C" : 768,
+      "bottom" : "expand_dims_0",
+      "weights" : {
+
+      },
+      "Nx" : 3,
+      "pad_mode" : 0,
+      "pad_value" : 0,
+      "Ny" : 1,
+      "n_parallel" : 1
+    },
+    {
+      "size_of_axes" : 1,
+      "bottom" : "conv_0",
+      "axes_0" : -2,
+      "weights" : {
+
+      },
+      "nd_axis" : 0,
+      "debug_info" : "52",
+      "version" : 1,
+      "top" : "52",
+      "type" : "squeeze",
+      "name" : "52"
+    },
+    {
+      "bottom" : "52",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.1",
+      "top" : "input.1",
+      "type" : "activation",
+      "name" : "input.1"
+    },
+    {
+      "size_of_axes" : 1,
+      "bottom" : "input.1",
+      "axes_0" : -2,
+      "weights" : {
+
+      },
+      "nd_axis" : 0,
+      "debug_info" : "expand_dims_2",
+      "top" : "expand_dims_2",
+      "type" : "expand_dims",
+      "name" : "expand_dims_2"
+    },
+    {
+      "pad_r" : 1,
+      "fused_relu" : 0,
+      "fused_tanh" : 0,
+      "debug_info" : "conv_1",
+      "pad_fill_mode" : 0,
+      "pad_b" : 0,
+      "pad_l" : 1,
+      "top" : "conv_1",
+      "blob_weights" : 7,
+      "K" : 768,
+      "blob_biases" : 5,
+      "stride_x" : 2,
+      "name" : "conv_1",
+      "has_batch_norm" : 0,
+      "type" : "convolution",
+      "n_groups" : 1,
+      "pad_t" : 0,
+      "has_biases" : 1,
+      "C" : 768,
+      "bottom" : "expand_dims_2",
+      "weights" : {
+
+      },
+      "Nx" : 3,
+      "pad_mode" : 0,
+      "pad_value" : 0,
+      "Ny" : 1,
+      "n_parallel" : 1
+    },
+    {
+      "size_of_axes" : 1,
+      "bottom" : "conv_1",
+      "axes_0" : -2,
+      "weights" : {
+
+      },
+      "nd_axis" : 0,
+      "debug_info" : "60",
+      "version" : 1,
+      "top" : "60",
+      "type" : "squeeze",
+      "name" : "60"
+    },
+    {
+      "bottom" : "60",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "inputs_embeds.1",
+      "top" : "inputs_embeds.1",
+      "type" : "activation",
+      "name" : "inputs_embeds.1"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 1,
+      "bottom" : "inputs_embeds.1",
+      "axis_k" : 2,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_108",
+      "top" : "transpose_108",
+      "type" : "transpose",
+      "name" : "transpose_108"
+    },
+    {
+      "top" : "encoder.embed_positions.weight",
+      "w" : 768,
+      "h" : 1500,
+      "name" : "encoder.embed_positions.weight",
+      "nd_rank" : 2,
+      "type" : "load_constant",
+      "k" : 1,
+      "bottom" : "",
+      "debug_info" : "encoder.embed_positions.weight",
+      "n" : 1,
+      "weights" : {
+
+      },
+      "constant_blob" : 9
+    },
+    {
+      "bottom" : "encoder.embed_positions.weight,transpose_108",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.3",
+      "top" : "input.3",
+      "type" : "elementwise",
+      "name" : "input.3",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.1_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.3",
+      "debug_info" : "hidden_states.1_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.1_reshape"
+    },
+    {
+      "bottom" : "hidden_states.1_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.1_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.1_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.1_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.1_scale",
+      "constant_blob" : 11,
+      "top" : "hidden_states.1_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.1_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.1_scale",
+      "constant_blob" : 13,
+      "top" : "hidden_states.1_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.1_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.1_mvn,hidden_states.1_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.1_scale",
+      "top" : "hidden_states.1_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.1_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.1_scale_mul_out,hidden_states.1_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.1_scale",
+      "top" : "hidden_states.1_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.1_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.1",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.1_scale",
+      "debug_info" : "hidden_states.1",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.1"
+    },
+    {
+      "nB" : 768,
+      "top" : "83",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 17,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.1",
+      "blob_biases" : 15,
+      "has_tanh" : 0,
+      "debug_info" : "83",
+      "name" : "83",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "83",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.1",
+      "top" : "tensor.1",
+      "type" : "elementwise",
+      "name" : "tensor.1",
+      "beta" : 0
+    },
+    {
+      "name" : "87",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.1",
+      "debug_info" : "87",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "87"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.3",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 21,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.1",
+      "blob_biases" : 19,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.3",
+      "name" : "tensor.3",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "93",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.3",
+      "debug_info" : "93",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "93"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.5",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 25,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.1",
+      "blob_biases" : 23,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.5",
+      "name" : "tensor.5",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "100",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.5",
+      "debug_info" : "100",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "100"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "93",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_105",
+      "top" : "transpose_105",
+      "type" : "transpose",
+      "name" : "transpose_105"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "87",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_106",
+      "top" : "transpose_106",
+      "type" : "transpose",
+      "name" : "transpose_106"
+    },
+    {
+      "bottom" : "transpose_106,transpose_105",
+      "weights" : {
+
+      },
+      "debug_info" : "input.7",
+      "top" : "input.7",
+      "type" : "batch_matmul",
+      "name" : "input.7",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.7",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.9",
+      "top" : "input.9",
+      "type" : "softmax_nd",
+      "name" : "input.9"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "100",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_107",
+      "top" : "transpose_107",
+      "type" : "transpose",
+      "name" : "transpose_107"
+    },
+    {
+      "bottom" : "input.9,transpose_107",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.1",
+      "top" : "attn_output.1",
+      "type" : "batch_matmul",
+      "name" : "attn_output.1",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.1",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_104",
+      "top" : "transpose_104",
+      "type" : "transpose",
+      "name" : "transpose_104"
+    },
+    {
+      "name" : "input.11",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_104",
+      "debug_info" : "input.11",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.11"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.13",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 29,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.11",
+      "blob_biases" : 27,
+      "has_tanh" : 0,
+      "debug_info" : "input.13",
+      "name" : "input.13",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.3,input.13",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.15",
+      "top" : "input.15",
+      "type" : "elementwise",
+      "name" : "input.15",
+      "beta" : 0
+    },
+    {
+      "name" : "input.17_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.15",
+      "debug_info" : "input.17_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.17_reshape"
+    },
+    {
+      "bottom" : "input.17_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.17_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.17_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.17_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.17_scale",
+      "constant_blob" : 31,
+      "top" : "input.17_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.17_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.17_scale",
+      "constant_blob" : 33,
+      "top" : "input.17_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.17_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.17_mvn,input.17_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.17_scale",
+      "top" : "input.17_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.17_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.17_scale_mul_out,input.17_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.17_scale",
+      "top" : "input.17_scale",
+      "type" : "elementwise",
+      "name" : "input.17_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.17",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.17_scale",
+      "debug_info" : "input.17",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.17"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.19",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 37,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.17",
+      "blob_biases" : 35,
+      "has_tanh" : 0,
+      "debug_info" : "input.19",
+      "name" : "input.19",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.19",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.21",
+      "top" : "input.21",
+      "type" : "activation",
+      "name" : "input.21"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.25",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 41,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.21",
+      "blob_biases" : 39,
+      "has_tanh" : 0,
+      "debug_info" : "input.25",
+      "name" : "input.25",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.15,input.25",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.27",
+      "top" : "input.27",
+      "type" : "elementwise",
+      "name" : "input.27",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.7_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.27",
+      "debug_info" : "hidden_states.7_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.7_reshape"
+    },
+    {
+      "bottom" : "hidden_states.7_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.7_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.7_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.7_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.7_scale",
+      "constant_blob" : 43,
+      "top" : "hidden_states.7_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.7_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.7_scale",
+      "constant_blob" : 45,
+      "top" : "hidden_states.7_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.7_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.7_mvn,hidden_states.7_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.7_scale",
+      "top" : "hidden_states.7_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.7_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.7_scale_mul_out,hidden_states.7_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.7_scale",
+      "top" : "hidden_states.7_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.7_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.7",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.7_scale",
+      "debug_info" : "hidden_states.7",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.7"
+    },
+    {
+      "nB" : 768,
+      "top" : "147",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 49,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.7",
+      "blob_biases" : 47,
+      "has_tanh" : 0,
+      "debug_info" : "147",
+      "name" : "147",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "147",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.7",
+      "top" : "tensor.7",
+      "type" : "elementwise",
+      "name" : "tensor.7",
+      "beta" : 0
+    },
+    {
+      "name" : "151",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.7",
+      "debug_info" : "151",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "151"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.9",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 53,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.7",
+      "blob_biases" : 51,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.9",
+      "name" : "tensor.9",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "157",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.9",
+      "debug_info" : "157",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "157"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.11",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 57,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.7",
+      "blob_biases" : 55,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.11",
+      "name" : "tensor.11",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "164",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.11",
+      "debug_info" : "164",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "164"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "157",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_101",
+      "top" : "transpose_101",
+      "type" : "transpose",
+      "name" : "transpose_101"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "151",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_102",
+      "top" : "transpose_102",
+      "type" : "transpose",
+      "name" : "transpose_102"
+    },
+    {
+      "bottom" : "transpose_102,transpose_101",
+      "weights" : {
+
+      },
+      "debug_info" : "input.29",
+      "top" : "input.29",
+      "type" : "batch_matmul",
+      "name" : "input.29",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.29",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.31",
+      "top" : "input.31",
+      "type" : "softmax_nd",
+      "name" : "input.31"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "164",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_103",
+      "top" : "transpose_103",
+      "type" : "transpose",
+      "name" : "transpose_103"
+    },
+    {
+      "bottom" : "input.31,transpose_103",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.5",
+      "top" : "attn_output.5",
+      "type" : "batch_matmul",
+      "name" : "attn_output.5",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.5",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_100",
+      "top" : "transpose_100",
+      "type" : "transpose",
+      "name" : "transpose_100"
+    },
+    {
+      "name" : "input.33",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_100",
+      "debug_info" : "input.33",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.33"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.35",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 61,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.33",
+      "blob_biases" : 59,
+      "has_tanh" : 0,
+      "debug_info" : "input.35",
+      "name" : "input.35",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.27,input.35",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.37",
+      "top" : "input.37",
+      "type" : "elementwise",
+      "name" : "input.37",
+      "beta" : 0
+    },
+    {
+      "name" : "input.39_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.37",
+      "debug_info" : "input.39_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.39_reshape"
+    },
+    {
+      "bottom" : "input.39_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.39_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.39_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.39_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.39_scale",
+      "constant_blob" : 63,
+      "top" : "input.39_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.39_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.39_scale",
+      "constant_blob" : 65,
+      "top" : "input.39_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.39_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.39_mvn,input.39_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.39_scale",
+      "top" : "input.39_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.39_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.39_scale_mul_out,input.39_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.39_scale",
+      "top" : "input.39_scale",
+      "type" : "elementwise",
+      "name" : "input.39_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.39",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.39_scale",
+      "debug_info" : "input.39",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.39"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.41",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 69,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.39",
+      "blob_biases" : 67,
+      "has_tanh" : 0,
+      "debug_info" : "input.41",
+      "name" : "input.41",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.41",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.43",
+      "top" : "input.43",
+      "type" : "activation",
+      "name" : "input.43"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.47",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 73,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.43",
+      "blob_biases" : 71,
+      "has_tanh" : 0,
+      "debug_info" : "input.47",
+      "name" : "input.47",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.37,input.47",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.49",
+      "top" : "input.49",
+      "type" : "elementwise",
+      "name" : "input.49",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.13_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.49",
+      "debug_info" : "hidden_states.13_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.13_reshape"
+    },
+    {
+      "bottom" : "hidden_states.13_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.13_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.13_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.13_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.13_scale",
+      "constant_blob" : 75,
+      "top" : "hidden_states.13_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.13_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.13_scale",
+      "constant_blob" : 77,
+      "top" : "hidden_states.13_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.13_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.13_mvn,hidden_states.13_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.13_scale",
+      "top" : "hidden_states.13_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.13_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.13_scale_mul_out,hidden_states.13_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.13_scale",
+      "top" : "hidden_states.13_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.13_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.13",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.13_scale",
+      "debug_info" : "hidden_states.13",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.13"
+    },
+    {
+      "nB" : 768,
+      "top" : "211",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 81,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.13",
+      "blob_biases" : 79,
+      "has_tanh" : 0,
+      "debug_info" : "211",
+      "name" : "211",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "211",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.13",
+      "top" : "tensor.13",
+      "type" : "elementwise",
+      "name" : "tensor.13",
+      "beta" : 0
+    },
+    {
+      "name" : "215",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.13",
+      "debug_info" : "215",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "215"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.15",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 85,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.13",
+      "blob_biases" : 83,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.15",
+      "name" : "tensor.15",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "221",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.15",
+      "debug_info" : "221",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "221"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.17",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 89,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.13",
+      "blob_biases" : 87,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.17",
+      "name" : "tensor.17",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "228",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.17",
+      "debug_info" : "228",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "228"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "221",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_97",
+      "top" : "transpose_97",
+      "type" : "transpose",
+      "name" : "transpose_97"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "215",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_98",
+      "top" : "transpose_98",
+      "type" : "transpose",
+      "name" : "transpose_98"
+    },
+    {
+      "bottom" : "transpose_98,transpose_97",
+      "weights" : {
+
+      },
+      "debug_info" : "input.51",
+      "top" : "input.51",
+      "type" : "batch_matmul",
+      "name" : "input.51",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.51",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.53",
+      "top" : "input.53",
+      "type" : "softmax_nd",
+      "name" : "input.53"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "228",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_99",
+      "top" : "transpose_99",
+      "type" : "transpose",
+      "name" : "transpose_99"
+    },
+    {
+      "bottom" : "input.53,transpose_99",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.9",
+      "top" : "attn_output.9",
+      "type" : "batch_matmul",
+      "name" : "attn_output.9",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.9",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_96",
+      "top" : "transpose_96",
+      "type" : "transpose",
+      "name" : "transpose_96"
+    },
+    {
+      "name" : "input.55",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_96",
+      "debug_info" : "input.55",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.55"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.57",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 93,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.55",
+      "blob_biases" : 91,
+      "has_tanh" : 0,
+      "debug_info" : "input.57",
+      "name" : "input.57",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.49,input.57",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.59",
+      "top" : "input.59",
+      "type" : "elementwise",
+      "name" : "input.59",
+      "beta" : 0
+    },
+    {
+      "name" : "input.61_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.59",
+      "debug_info" : "input.61_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.61_reshape"
+    },
+    {
+      "bottom" : "input.61_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.61_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.61_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.61_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.61_scale",
+      "constant_blob" : 95,
+      "top" : "input.61_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.61_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.61_scale",
+      "constant_blob" : 97,
+      "top" : "input.61_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.61_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.61_mvn,input.61_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.61_scale",
+      "top" : "input.61_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.61_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.61_scale_mul_out,input.61_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.61_scale",
+      "top" : "input.61_scale",
+      "type" : "elementwise",
+      "name" : "input.61_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.61",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.61_scale",
+      "debug_info" : "input.61",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.61"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.63",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 101,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.61",
+      "blob_biases" : 99,
+      "has_tanh" : 0,
+      "debug_info" : "input.63",
+      "name" : "input.63",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.63",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.65",
+      "top" : "input.65",
+      "type" : "activation",
+      "name" : "input.65"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.69",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 105,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.65",
+      "blob_biases" : 103,
+      "has_tanh" : 0,
+      "debug_info" : "input.69",
+      "name" : "input.69",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.59,input.69",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.71",
+      "top" : "input.71",
+      "type" : "elementwise",
+      "name" : "input.71",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.19_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.71",
+      "debug_info" : "hidden_states.19_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.19_reshape"
+    },
+    {
+      "bottom" : "hidden_states.19_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.19_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.19_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.19_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.19_scale",
+      "constant_blob" : 107,
+      "top" : "hidden_states.19_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.19_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.19_scale",
+      "constant_blob" : 109,
+      "top" : "hidden_states.19_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.19_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.19_mvn,hidden_states.19_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.19_scale",
+      "top" : "hidden_states.19_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.19_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.19_scale_mul_out,hidden_states.19_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.19_scale",
+      "top" : "hidden_states.19_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.19_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.19",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.19_scale",
+      "debug_info" : "hidden_states.19",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.19"
+    },
+    {
+      "nB" : 768,
+      "top" : "275",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 113,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.19",
+      "blob_biases" : 111,
+      "has_tanh" : 0,
+      "debug_info" : "275",
+      "name" : "275",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "275",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.19",
+      "top" : "tensor.19",
+      "type" : "elementwise",
+      "name" : "tensor.19",
+      "beta" : 0
+    },
+    {
+      "name" : "279",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.19",
+      "debug_info" : "279",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "279"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.21",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 117,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.19",
+      "blob_biases" : 115,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.21",
+      "name" : "tensor.21",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "285",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.21",
+      "debug_info" : "285",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "285"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.23",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 121,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.19",
+      "blob_biases" : 119,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.23",
+      "name" : "tensor.23",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "292",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.23",
+      "debug_info" : "292",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "292"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "285",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_93",
+      "top" : "transpose_93",
+      "type" : "transpose",
+      "name" : "transpose_93"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "279",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_94",
+      "top" : "transpose_94",
+      "type" : "transpose",
+      "name" : "transpose_94"
+    },
+    {
+      "bottom" : "transpose_94,transpose_93",
+      "weights" : {
+
+      },
+      "debug_info" : "input.73",
+      "top" : "input.73",
+      "type" : "batch_matmul",
+      "name" : "input.73",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.73",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.75",
+      "top" : "input.75",
+      "type" : "softmax_nd",
+      "name" : "input.75"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "292",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_95",
+      "top" : "transpose_95",
+      "type" : "transpose",
+      "name" : "transpose_95"
+    },
+    {
+      "bottom" : "input.75,transpose_95",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.13",
+      "top" : "attn_output.13",
+      "type" : "batch_matmul",
+      "name" : "attn_output.13",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.13",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_92",
+      "top" : "transpose_92",
+      "type" : "transpose",
+      "name" : "transpose_92"
+    },
+    {
+      "name" : "input.77",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_92",
+      "debug_info" : "input.77",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.77"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.79",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 125,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.77",
+      "blob_biases" : 123,
+      "has_tanh" : 0,
+      "debug_info" : "input.79",
+      "name" : "input.79",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.71,input.79",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.81",
+      "top" : "input.81",
+      "type" : "elementwise",
+      "name" : "input.81",
+      "beta" : 0
+    },
+    {
+      "name" : "input.83_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.81",
+      "debug_info" : "input.83_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.83_reshape"
+    },
+    {
+      "bottom" : "input.83_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.83_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.83_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.83_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.83_scale",
+      "constant_blob" : 127,
+      "top" : "input.83_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.83_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.83_scale",
+      "constant_blob" : 129,
+      "top" : "input.83_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.83_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.83_mvn,input.83_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.83_scale",
+      "top" : "input.83_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.83_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.83_scale_mul_out,input.83_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.83_scale",
+      "top" : "input.83_scale",
+      "type" : "elementwise",
+      "name" : "input.83_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.83",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.83_scale",
+      "debug_info" : "input.83",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.83"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.85",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 133,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.83",
+      "blob_biases" : 131,
+      "has_tanh" : 0,
+      "debug_info" : "input.85",
+      "name" : "input.85",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.85",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.87",
+      "top" : "input.87",
+      "type" : "activation",
+      "name" : "input.87"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.91",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 137,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.87",
+      "blob_biases" : 135,
+      "has_tanh" : 0,
+      "debug_info" : "input.91",
+      "name" : "input.91",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.81,input.91",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.93",
+      "top" : "input.93",
+      "type" : "elementwise",
+      "name" : "input.93",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.25_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.93",
+      "debug_info" : "hidden_states.25_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.25_reshape"
+    },
+    {
+      "bottom" : "hidden_states.25_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.25_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.25_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.25_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.25_scale",
+      "constant_blob" : 139,
+      "top" : "hidden_states.25_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.25_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.25_scale",
+      "constant_blob" : 141,
+      "top" : "hidden_states.25_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.25_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.25_mvn,hidden_states.25_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.25_scale",
+      "top" : "hidden_states.25_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.25_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.25_scale_mul_out,hidden_states.25_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.25_scale",
+      "top" : "hidden_states.25_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.25_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.25",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.25_scale",
+      "debug_info" : "hidden_states.25",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.25"
+    },
+    {
+      "nB" : 768,
+      "top" : "339",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 145,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.25",
+      "blob_biases" : 143,
+      "has_tanh" : 0,
+      "debug_info" : "339",
+      "name" : "339",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "339",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.25",
+      "top" : "tensor.25",
+      "type" : "elementwise",
+      "name" : "tensor.25",
+      "beta" : 0
+    },
+    {
+      "name" : "343",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.25",
+      "debug_info" : "343",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "343"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.27",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 149,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.25",
+      "blob_biases" : 147,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.27",
+      "name" : "tensor.27",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "349",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.27",
+      "debug_info" : "349",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "349"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.29",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 153,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.25",
+      "blob_biases" : 151,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.29",
+      "name" : "tensor.29",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "356",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.29",
+      "debug_info" : "356",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "356"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "349",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_89",
+      "top" : "transpose_89",
+      "type" : "transpose",
+      "name" : "transpose_89"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "343",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_90",
+      "top" : "transpose_90",
+      "type" : "transpose",
+      "name" : "transpose_90"
+    },
+    {
+      "bottom" : "transpose_90,transpose_89",
+      "weights" : {
+
+      },
+      "debug_info" : "input.95",
+      "top" : "input.95",
+      "type" : "batch_matmul",
+      "name" : "input.95",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.95",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.97",
+      "top" : "input.97",
+      "type" : "softmax_nd",
+      "name" : "input.97"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "356",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_91",
+      "top" : "transpose_91",
+      "type" : "transpose",
+      "name" : "transpose_91"
+    },
+    {
+      "bottom" : "input.97,transpose_91",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.17",
+      "top" : "attn_output.17",
+      "type" : "batch_matmul",
+      "name" : "attn_output.17",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.17",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_88",
+      "top" : "transpose_88",
+      "type" : "transpose",
+      "name" : "transpose_88"
+    },
+    {
+      "name" : "input.99",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_88",
+      "debug_info" : "input.99",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.99"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.101",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 157,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.99",
+      "blob_biases" : 155,
+      "has_tanh" : 0,
+      "debug_info" : "input.101",
+      "name" : "input.101",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.93,input.101",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.103",
+      "top" : "input.103",
+      "type" : "elementwise",
+      "name" : "input.103",
+      "beta" : 0
+    },
+    {
+      "name" : "input.105_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.103",
+      "debug_info" : "input.105_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.105_reshape"
+    },
+    {
+      "bottom" : "input.105_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.105_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.105_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.105_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.105_scale",
+      "constant_blob" : 159,
+      "top" : "input.105_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.105_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.105_scale",
+      "constant_blob" : 161,
+      "top" : "input.105_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.105_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.105_mvn,input.105_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.105_scale",
+      "top" : "input.105_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.105_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.105_scale_mul_out,input.105_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.105_scale",
+      "top" : "input.105_scale",
+      "type" : "elementwise",
+      "name" : "input.105_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.105",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.105_scale",
+      "debug_info" : "input.105",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.105"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.107",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 165,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.105",
+      "blob_biases" : 163,
+      "has_tanh" : 0,
+      "debug_info" : "input.107",
+      "name" : "input.107",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.107",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.109",
+      "top" : "input.109",
+      "type" : "activation",
+      "name" : "input.109"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.113",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 169,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.109",
+      "blob_biases" : 167,
+      "has_tanh" : 0,
+      "debug_info" : "input.113",
+      "name" : "input.113",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.103,input.113",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.115",
+      "top" : "input.115",
+      "type" : "elementwise",
+      "name" : "input.115",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.31_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.115",
+      "debug_info" : "hidden_states.31_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.31_reshape"
+    },
+    {
+      "bottom" : "hidden_states.31_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.31_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.31_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.31_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.31_scale",
+      "constant_blob" : 171,
+      "top" : "hidden_states.31_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.31_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.31_scale",
+      "constant_blob" : 173,
+      "top" : "hidden_states.31_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.31_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.31_mvn,hidden_states.31_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.31_scale",
+      "top" : "hidden_states.31_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.31_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.31_scale_mul_out,hidden_states.31_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.31_scale",
+      "top" : "hidden_states.31_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.31_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.31",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.31_scale",
+      "debug_info" : "hidden_states.31",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.31"
+    },
+    {
+      "nB" : 768,
+      "top" : "403",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 177,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.31",
+      "blob_biases" : 175,
+      "has_tanh" : 0,
+      "debug_info" : "403",
+      "name" : "403",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "403",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.31",
+      "top" : "tensor.31",
+      "type" : "elementwise",
+      "name" : "tensor.31",
+      "beta" : 0
+    },
+    {
+      "name" : "407",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.31",
+      "debug_info" : "407",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "407"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.33",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 181,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.31",
+      "blob_biases" : 179,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.33",
+      "name" : "tensor.33",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "413",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.33",
+      "debug_info" : "413",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "413"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.35",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 185,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.31",
+      "blob_biases" : 183,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.35",
+      "name" : "tensor.35",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "420",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.35",
+      "debug_info" : "420",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "420"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "413",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_85",
+      "top" : "transpose_85",
+      "type" : "transpose",
+      "name" : "transpose_85"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "407",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_86",
+      "top" : "transpose_86",
+      "type" : "transpose",
+      "name" : "transpose_86"
+    },
+    {
+      "bottom" : "transpose_86,transpose_85",
+      "weights" : {
+
+      },
+      "debug_info" : "input.117",
+      "top" : "input.117",
+      "type" : "batch_matmul",
+      "name" : "input.117",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.117",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.119",
+      "top" : "input.119",
+      "type" : "softmax_nd",
+      "name" : "input.119"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "420",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_87",
+      "top" : "transpose_87",
+      "type" : "transpose",
+      "name" : "transpose_87"
+    },
+    {
+      "bottom" : "input.119,transpose_87",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.21",
+      "top" : "attn_output.21",
+      "type" : "batch_matmul",
+      "name" : "attn_output.21",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.21",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_84",
+      "top" : "transpose_84",
+      "type" : "transpose",
+      "name" : "transpose_84"
+    },
+    {
+      "name" : "input.121",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_84",
+      "debug_info" : "input.121",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.121"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.123",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 189,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.121",
+      "blob_biases" : 187,
+      "has_tanh" : 0,
+      "debug_info" : "input.123",
+      "name" : "input.123",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.115,input.123",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.125",
+      "top" : "input.125",
+      "type" : "elementwise",
+      "name" : "input.125",
+      "beta" : 0
+    },
+    {
+      "name" : "input.127_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.125",
+      "debug_info" : "input.127_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.127_reshape"
+    },
+    {
+      "bottom" : "input.127_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.127_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.127_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.127_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.127_scale",
+      "constant_blob" : 191,
+      "top" : "input.127_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.127_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.127_scale",
+      "constant_blob" : 193,
+      "top" : "input.127_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.127_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.127_mvn,input.127_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.127_scale",
+      "top" : "input.127_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.127_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.127_scale_mul_out,input.127_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.127_scale",
+      "top" : "input.127_scale",
+      "type" : "elementwise",
+      "name" : "input.127_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.127",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.127_scale",
+      "debug_info" : "input.127",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.127"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.129",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 197,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.127",
+      "blob_biases" : 195,
+      "has_tanh" : 0,
+      "debug_info" : "input.129",
+      "name" : "input.129",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.129",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.131",
+      "top" : "input.131",
+      "type" : "activation",
+      "name" : "input.131"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.135",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 201,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.131",
+      "blob_biases" : 199,
+      "has_tanh" : 0,
+      "debug_info" : "input.135",
+      "name" : "input.135",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.125,input.135",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.137",
+      "top" : "input.137",
+      "type" : "elementwise",
+      "name" : "input.137",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.37_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.137",
+      "debug_info" : "hidden_states.37_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.37_reshape"
+    },
+    {
+      "bottom" : "hidden_states.37_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.37_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.37_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.37_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.37_scale",
+      "constant_blob" : 203,
+      "top" : "hidden_states.37_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.37_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.37_scale",
+      "constant_blob" : 205,
+      "top" : "hidden_states.37_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.37_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.37_mvn,hidden_states.37_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.37_scale",
+      "top" : "hidden_states.37_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.37_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.37_scale_mul_out,hidden_states.37_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.37_scale",
+      "top" : "hidden_states.37_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.37_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.37",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.37_scale",
+      "debug_info" : "hidden_states.37",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.37"
+    },
+    {
+      "nB" : 768,
+      "top" : "467",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 209,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.37",
+      "blob_biases" : 207,
+      "has_tanh" : 0,
+      "debug_info" : "467",
+      "name" : "467",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "467",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.37",
+      "top" : "tensor.37",
+      "type" : "elementwise",
+      "name" : "tensor.37",
+      "beta" : 0
+    },
+    {
+      "name" : "471",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.37",
+      "debug_info" : "471",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "471"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.39",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 213,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.37",
+      "blob_biases" : 211,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.39",
+      "name" : "tensor.39",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "477",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.39",
+      "debug_info" : "477",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "477"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.41",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 217,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.37",
+      "blob_biases" : 215,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.41",
+      "name" : "tensor.41",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "484",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.41",
+      "debug_info" : "484",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "484"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "477",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_81",
+      "top" : "transpose_81",
+      "type" : "transpose",
+      "name" : "transpose_81"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "471",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_82",
+      "top" : "transpose_82",
+      "type" : "transpose",
+      "name" : "transpose_82"
+    },
+    {
+      "bottom" : "transpose_82,transpose_81",
+      "weights" : {
+
+      },
+      "debug_info" : "input.139",
+      "top" : "input.139",
+      "type" : "batch_matmul",
+      "name" : "input.139",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.139",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.141",
+      "top" : "input.141",
+      "type" : "softmax_nd",
+      "name" : "input.141"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "484",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_83",
+      "top" : "transpose_83",
+      "type" : "transpose",
+      "name" : "transpose_83"
+    },
+    {
+      "bottom" : "input.141,transpose_83",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.25",
+      "top" : "attn_output.25",
+      "type" : "batch_matmul",
+      "name" : "attn_output.25",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.25",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_80",
+      "top" : "transpose_80",
+      "type" : "transpose",
+      "name" : "transpose_80"
+    },
+    {
+      "name" : "input.143",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_80",
+      "debug_info" : "input.143",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.143"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.145",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 221,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.143",
+      "blob_biases" : 219,
+      "has_tanh" : 0,
+      "debug_info" : "input.145",
+      "name" : "input.145",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.137,input.145",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.147",
+      "top" : "input.147",
+      "type" : "elementwise",
+      "name" : "input.147",
+      "beta" : 0
+    },
+    {
+      "name" : "input.149_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.147",
+      "debug_info" : "input.149_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.149_reshape"
+    },
+    {
+      "bottom" : "input.149_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.149_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.149_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.149_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.149_scale",
+      "constant_blob" : 223,
+      "top" : "input.149_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.149_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.149_scale",
+      "constant_blob" : 225,
+      "top" : "input.149_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.149_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.149_mvn,input.149_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.149_scale",
+      "top" : "input.149_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.149_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.149_scale_mul_out,input.149_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.149_scale",
+      "top" : "input.149_scale",
+      "type" : "elementwise",
+      "name" : "input.149_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.149",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.149_scale",
+      "debug_info" : "input.149",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.149"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.151",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 229,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.149",
+      "blob_biases" : 227,
+      "has_tanh" : 0,
+      "debug_info" : "input.151",
+      "name" : "input.151",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.151",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.153",
+      "top" : "input.153",
+      "type" : "activation",
+      "name" : "input.153"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.157",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 233,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.153",
+      "blob_biases" : 231,
+      "has_tanh" : 0,
+      "debug_info" : "input.157",
+      "name" : "input.157",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.147,input.157",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.159",
+      "top" : "input.159",
+      "type" : "elementwise",
+      "name" : "input.159",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.43_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.159",
+      "debug_info" : "hidden_states.43_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.43_reshape"
+    },
+    {
+      "bottom" : "hidden_states.43_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.43_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.43_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.43_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.43_scale",
+      "constant_blob" : 235,
+      "top" : "hidden_states.43_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.43_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.43_scale",
+      "constant_blob" : 237,
+      "top" : "hidden_states.43_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.43_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.43_mvn,hidden_states.43_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.43_scale",
+      "top" : "hidden_states.43_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.43_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.43_scale_mul_out,hidden_states.43_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.43_scale",
+      "top" : "hidden_states.43_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.43_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.43",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.43_scale",
+      "debug_info" : "hidden_states.43",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.43"
+    },
+    {
+      "nB" : 768,
+      "top" : "531",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 241,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.43",
+      "blob_biases" : 239,
+      "has_tanh" : 0,
+      "debug_info" : "531",
+      "name" : "531",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "531",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.43",
+      "top" : "tensor.43",
+      "type" : "elementwise",
+      "name" : "tensor.43",
+      "beta" : 0
+    },
+    {
+      "name" : "535",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.43",
+      "debug_info" : "535",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "535"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.45",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 245,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.43",
+      "blob_biases" : 243,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.45",
+      "name" : "tensor.45",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "541",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.45",
+      "debug_info" : "541",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "541"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.47",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 249,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.43",
+      "blob_biases" : 247,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.47",
+      "name" : "tensor.47",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "548",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.47",
+      "debug_info" : "548",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "548"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "541",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_77",
+      "top" : "transpose_77",
+      "type" : "transpose",
+      "name" : "transpose_77"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "535",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_78",
+      "top" : "transpose_78",
+      "type" : "transpose",
+      "name" : "transpose_78"
+    },
+    {
+      "bottom" : "transpose_78,transpose_77",
+      "weights" : {
+
+      },
+      "debug_info" : "input.161",
+      "top" : "input.161",
+      "type" : "batch_matmul",
+      "name" : "input.161",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.161",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.163",
+      "top" : "input.163",
+      "type" : "softmax_nd",
+      "name" : "input.163"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "548",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_79",
+      "top" : "transpose_79",
+      "type" : "transpose",
+      "name" : "transpose_79"
+    },
+    {
+      "bottom" : "input.163,transpose_79",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.29",
+      "top" : "attn_output.29",
+      "type" : "batch_matmul",
+      "name" : "attn_output.29",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.29",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_76",
+      "top" : "transpose_76",
+      "type" : "transpose",
+      "name" : "transpose_76"
+    },
+    {
+      "name" : "input.165",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_76",
+      "debug_info" : "input.165",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.165"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.167",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 253,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.165",
+      "blob_biases" : 251,
+      "has_tanh" : 0,
+      "debug_info" : "input.167",
+      "name" : "input.167",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.159,input.167",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.169",
+      "top" : "input.169",
+      "type" : "elementwise",
+      "name" : "input.169",
+      "beta" : 0
+    },
+    {
+      "name" : "input.171_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.169",
+      "debug_info" : "input.171_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.171_reshape"
+    },
+    {
+      "bottom" : "input.171_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.171_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.171_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.171_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.171_scale",
+      "constant_blob" : 255,
+      "top" : "input.171_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.171_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.171_scale",
+      "constant_blob" : 257,
+      "top" : "input.171_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.171_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.171_mvn,input.171_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.171_scale",
+      "top" : "input.171_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.171_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.171_scale_mul_out,input.171_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.171_scale",
+      "top" : "input.171_scale",
+      "type" : "elementwise",
+      "name" : "input.171_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.171",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.171_scale",
+      "debug_info" : "input.171",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.171"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.173",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 261,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.171",
+      "blob_biases" : 259,
+      "has_tanh" : 0,
+      "debug_info" : "input.173",
+      "name" : "input.173",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.173",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.175",
+      "top" : "input.175",
+      "type" : "activation",
+      "name" : "input.175"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.179",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 265,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.175",
+      "blob_biases" : 263,
+      "has_tanh" : 0,
+      "debug_info" : "input.179",
+      "name" : "input.179",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.169,input.179",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.181",
+      "top" : "input.181",
+      "type" : "elementwise",
+      "name" : "input.181",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.49_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.181",
+      "debug_info" : "hidden_states.49_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.49_reshape"
+    },
+    {
+      "bottom" : "hidden_states.49_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.49_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.49_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.49_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.49_scale",
+      "constant_blob" : 267,
+      "top" : "hidden_states.49_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.49_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.49_scale",
+      "constant_blob" : 269,
+      "top" : "hidden_states.49_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.49_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.49_mvn,hidden_states.49_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.49_scale",
+      "top" : "hidden_states.49_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.49_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.49_scale_mul_out,hidden_states.49_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.49_scale",
+      "top" : "hidden_states.49_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.49_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.49",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.49_scale",
+      "debug_info" : "hidden_states.49",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.49"
+    },
+    {
+      "nB" : 768,
+      "top" : "595",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 273,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.49",
+      "blob_biases" : 271,
+      "has_tanh" : 0,
+      "debug_info" : "595",
+      "name" : "595",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "595",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.49",
+      "top" : "tensor.49",
+      "type" : "elementwise",
+      "name" : "tensor.49",
+      "beta" : 0
+    },
+    {
+      "name" : "599",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.49",
+      "debug_info" : "599",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "599"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.51",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 277,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.49",
+      "blob_biases" : 275,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.51",
+      "name" : "tensor.51",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "605",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.51",
+      "debug_info" : "605",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "605"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.53",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 281,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.49",
+      "blob_biases" : 279,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.53",
+      "name" : "tensor.53",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "612",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.53",
+      "debug_info" : "612",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "612"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "605",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_73",
+      "top" : "transpose_73",
+      "type" : "transpose",
+      "name" : "transpose_73"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "599",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_74",
+      "top" : "transpose_74",
+      "type" : "transpose",
+      "name" : "transpose_74"
+    },
+    {
+      "bottom" : "transpose_74,transpose_73",
+      "weights" : {
+
+      },
+      "debug_info" : "input.183",
+      "top" : "input.183",
+      "type" : "batch_matmul",
+      "name" : "input.183",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.183",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.185",
+      "top" : "input.185",
+      "type" : "softmax_nd",
+      "name" : "input.185"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "612",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_75",
+      "top" : "transpose_75",
+      "type" : "transpose",
+      "name" : "transpose_75"
+    },
+    {
+      "bottom" : "input.185,transpose_75",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.33",
+      "top" : "attn_output.33",
+      "type" : "batch_matmul",
+      "name" : "attn_output.33",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.33",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_72",
+      "top" : "transpose_72",
+      "type" : "transpose",
+      "name" : "transpose_72"
+    },
+    {
+      "name" : "input.187",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_72",
+      "debug_info" : "input.187",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.187"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.189",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 285,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.187",
+      "blob_biases" : 283,
+      "has_tanh" : 0,
+      "debug_info" : "input.189",
+      "name" : "input.189",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.181,input.189",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.191",
+      "top" : "input.191",
+      "type" : "elementwise",
+      "name" : "input.191",
+      "beta" : 0
+    },
+    {
+      "name" : "input.193_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.191",
+      "debug_info" : "input.193_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.193_reshape"
+    },
+    {
+      "bottom" : "input.193_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.193_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.193_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.193_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.193_scale",
+      "constant_blob" : 287,
+      "top" : "input.193_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.193_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.193_scale",
+      "constant_blob" : 289,
+      "top" : "input.193_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.193_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.193_mvn,input.193_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.193_scale",
+      "top" : "input.193_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.193_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.193_scale_mul_out,input.193_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.193_scale",
+      "top" : "input.193_scale",
+      "type" : "elementwise",
+      "name" : "input.193_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.193",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.193_scale",
+      "debug_info" : "input.193",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.193"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.195",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 293,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.193",
+      "blob_biases" : 291,
+      "has_tanh" : 0,
+      "debug_info" : "input.195",
+      "name" : "input.195",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.195",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.197",
+      "top" : "input.197",
+      "type" : "activation",
+      "name" : "input.197"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.201",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 297,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.197",
+      "blob_biases" : 295,
+      "has_tanh" : 0,
+      "debug_info" : "input.201",
+      "name" : "input.201",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.191,input.201",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.203",
+      "top" : "input.203",
+      "type" : "elementwise",
+      "name" : "input.203",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.55_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.203",
+      "debug_info" : "hidden_states.55_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.55_reshape"
+    },
+    {
+      "bottom" : "hidden_states.55_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.55_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.55_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.55_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.55_scale",
+      "constant_blob" : 299,
+      "top" : "hidden_states.55_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.55_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.55_scale",
+      "constant_blob" : 301,
+      "top" : "hidden_states.55_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.55_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.55_mvn,hidden_states.55_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.55_scale",
+      "top" : "hidden_states.55_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.55_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.55_scale_mul_out,hidden_states.55_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.55_scale",
+      "top" : "hidden_states.55_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.55_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.55",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.55_scale",
+      "debug_info" : "hidden_states.55",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.55"
+    },
+    {
+      "nB" : 768,
+      "top" : "659",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 305,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.55",
+      "blob_biases" : 303,
+      "has_tanh" : 0,
+      "debug_info" : "659",
+      "name" : "659",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "659",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.55",
+      "top" : "tensor.55",
+      "type" : "elementwise",
+      "name" : "tensor.55",
+      "beta" : 0
+    },
+    {
+      "name" : "663",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.55",
+      "debug_info" : "663",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "663"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.57",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 309,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.55",
+      "blob_biases" : 307,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.57",
+      "name" : "tensor.57",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "669",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.57",
+      "debug_info" : "669",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "669"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.59",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 313,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.55",
+      "blob_biases" : 311,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.59",
+      "name" : "tensor.59",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "676",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.59",
+      "debug_info" : "676",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "676"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "669",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_69",
+      "top" : "transpose_69",
+      "type" : "transpose",
+      "name" : "transpose_69"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "663",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_70",
+      "top" : "transpose_70",
+      "type" : "transpose",
+      "name" : "transpose_70"
+    },
+    {
+      "bottom" : "transpose_70,transpose_69",
+      "weights" : {
+
+      },
+      "debug_info" : "input.205",
+      "top" : "input.205",
+      "type" : "batch_matmul",
+      "name" : "input.205",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.205",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.207",
+      "top" : "input.207",
+      "type" : "softmax_nd",
+      "name" : "input.207"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "676",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_71",
+      "top" : "transpose_71",
+      "type" : "transpose",
+      "name" : "transpose_71"
+    },
+    {
+      "bottom" : "input.207,transpose_71",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.37",
+      "top" : "attn_output.37",
+      "type" : "batch_matmul",
+      "name" : "attn_output.37",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.37",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_68",
+      "top" : "transpose_68",
+      "type" : "transpose",
+      "name" : "transpose_68"
+    },
+    {
+      "name" : "input.209",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_68",
+      "debug_info" : "input.209",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.209"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.211",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 317,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.209",
+      "blob_biases" : 315,
+      "has_tanh" : 0,
+      "debug_info" : "input.211",
+      "name" : "input.211",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.203,input.211",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.213",
+      "top" : "input.213",
+      "type" : "elementwise",
+      "name" : "input.213",
+      "beta" : 0
+    },
+    {
+      "name" : "input.215_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.213",
+      "debug_info" : "input.215_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.215_reshape"
+    },
+    {
+      "bottom" : "input.215_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.215_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.215_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.215_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.215_scale",
+      "constant_blob" : 319,
+      "top" : "input.215_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.215_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.215_scale",
+      "constant_blob" : 321,
+      "top" : "input.215_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.215_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.215_mvn,input.215_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.215_scale",
+      "top" : "input.215_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.215_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.215_scale_mul_out,input.215_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.215_scale",
+      "top" : "input.215_scale",
+      "type" : "elementwise",
+      "name" : "input.215_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.215",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.215_scale",
+      "debug_info" : "input.215",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.215"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.217",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 325,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.215",
+      "blob_biases" : 323,
+      "has_tanh" : 0,
+      "debug_info" : "input.217",
+      "name" : "input.217",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.217",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.219",
+      "top" : "input.219",
+      "type" : "activation",
+      "name" : "input.219"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.223",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 329,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.219",
+      "blob_biases" : 327,
+      "has_tanh" : 0,
+      "debug_info" : "input.223",
+      "name" : "input.223",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.213,input.223",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.225",
+      "top" : "input.225",
+      "type" : "elementwise",
+      "name" : "input.225",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.61_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.225",
+      "debug_info" : "hidden_states.61_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.61_reshape"
+    },
+    {
+      "bottom" : "hidden_states.61_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.61_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.61_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.61_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.61_scale",
+      "constant_blob" : 331,
+      "top" : "hidden_states.61_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.61_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.61_scale",
+      "constant_blob" : 333,
+      "top" : "hidden_states.61_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.61_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.61_mvn,hidden_states.61_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.61_scale",
+      "top" : "hidden_states.61_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.61_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.61_scale_mul_out,hidden_states.61_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.61_scale",
+      "top" : "hidden_states.61_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.61_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.61",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.61_scale",
+      "debug_info" : "hidden_states.61",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.61"
+    },
+    {
+      "nB" : 768,
+      "top" : "723",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 337,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.61",
+      "blob_biases" : 335,
+      "has_tanh" : 0,
+      "debug_info" : "723",
+      "name" : "723",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "723",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.61",
+      "top" : "tensor.61",
+      "type" : "elementwise",
+      "name" : "tensor.61",
+      "beta" : 0
+    },
+    {
+      "name" : "727",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.61",
+      "debug_info" : "727",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "727"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.63",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 341,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.61",
+      "blob_biases" : 339,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.63",
+      "name" : "tensor.63",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "733",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.63",
+      "debug_info" : "733",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "733"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.65",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 345,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.61",
+      "blob_biases" : 343,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.65",
+      "name" : "tensor.65",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "740",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.65",
+      "debug_info" : "740",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "740"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "733",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_65",
+      "top" : "transpose_65",
+      "type" : "transpose",
+      "name" : "transpose_65"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "727",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_66",
+      "top" : "transpose_66",
+      "type" : "transpose",
+      "name" : "transpose_66"
+    },
+    {
+      "bottom" : "transpose_66,transpose_65",
+      "weights" : {
+
+      },
+      "debug_info" : "input.227",
+      "top" : "input.227",
+      "type" : "batch_matmul",
+      "name" : "input.227",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.227",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.229",
+      "top" : "input.229",
+      "type" : "softmax_nd",
+      "name" : "input.229"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "740",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_67",
+      "top" : "transpose_67",
+      "type" : "transpose",
+      "name" : "transpose_67"
+    },
+    {
+      "bottom" : "input.229,transpose_67",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.41",
+      "top" : "attn_output.41",
+      "type" : "batch_matmul",
+      "name" : "attn_output.41",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.41",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_64",
+      "top" : "transpose_64",
+      "type" : "transpose",
+      "name" : "transpose_64"
+    },
+    {
+      "name" : "input.231",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_64",
+      "debug_info" : "input.231",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.231"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.233",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 349,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.231",
+      "blob_biases" : 347,
+      "has_tanh" : 0,
+      "debug_info" : "input.233",
+      "name" : "input.233",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.225,input.233",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.235",
+      "top" : "input.235",
+      "type" : "elementwise",
+      "name" : "input.235",
+      "beta" : 0
+    },
+    {
+      "name" : "input.237_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.235",
+      "debug_info" : "input.237_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.237_reshape"
+    },
+    {
+      "bottom" : "input.237_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.237_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.237_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.237_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.237_scale",
+      "constant_blob" : 351,
+      "top" : "input.237_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.237_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.237_scale",
+      "constant_blob" : 353,
+      "top" : "input.237_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.237_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.237_mvn,input.237_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.237_scale",
+      "top" : "input.237_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.237_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.237_scale_mul_out,input.237_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.237_scale",
+      "top" : "input.237_scale",
+      "type" : "elementwise",
+      "name" : "input.237_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.237",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.237_scale",
+      "debug_info" : "input.237",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.237"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.239",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 357,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.237",
+      "blob_biases" : 355,
+      "has_tanh" : 0,
+      "debug_info" : "input.239",
+      "name" : "input.239",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.239",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.241",
+      "top" : "input.241",
+      "type" : "activation",
+      "name" : "input.241"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.245",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 361,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.241",
+      "blob_biases" : 359,
+      "has_tanh" : 0,
+      "debug_info" : "input.245",
+      "name" : "input.245",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.235,input.245",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.247",
+      "top" : "input.247",
+      "type" : "elementwise",
+      "name" : "input.247",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.67_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.247",
+      "debug_info" : "hidden_states.67_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "hidden_states.67_reshape"
+    },
+    {
+      "bottom" : "hidden_states.67_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "hidden_states.67_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "hidden_states.67_mvn",
+      "type" : "l2_normalize",
+      "name" : "hidden_states.67_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.67_scale",
+      "constant_blob" : 363,
+      "top" : "hidden_states.67_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.67_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "hidden_states.67_scale",
+      "constant_blob" : 365,
+      "top" : "hidden_states.67_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "hidden_states.67_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "hidden_states.67_mvn,hidden_states.67_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.67_scale",
+      "top" : "hidden_states.67_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "hidden_states.67_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "hidden_states.67_scale_mul_out,hidden_states.67_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "hidden_states.67_scale",
+      "top" : "hidden_states.67_scale",
+      "type" : "elementwise",
+      "name" : "hidden_states.67_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "hidden_states.67",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "hidden_states.67_scale",
+      "debug_info" : "hidden_states.67",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "hidden_states.67"
+    },
+    {
+      "nB" : 768,
+      "top" : "787",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 369,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.67",
+      "blob_biases" : 367,
+      "has_tanh" : 0,
+      "debug_info" : "787",
+      "name" : "787",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "787",
+      "alpha" : 0.125,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "tensor.67",
+      "top" : "tensor.67",
+      "type" : "elementwise",
+      "name" : "tensor.67",
+      "beta" : 0
+    },
+    {
+      "name" : "791",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.67",
+      "debug_info" : "791",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "791"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor.69",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 373,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.67",
+      "blob_biases" : 371,
+      "has_tanh" : 0,
+      "debug_info" : "tensor.69",
+      "name" : "tensor.69",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "797",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor.69",
+      "debug_info" : "797",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "797"
+    },
+    {
+      "nB" : 768,
+      "top" : "tensor",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 377,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "hidden_states.67",
+      "blob_biases" : 375,
+      "has_tanh" : 0,
+      "debug_info" : "tensor",
+      "name" : "tensor",
+      "has_prelu" : 0
+    },
+    {
+      "name" : "804",
+      "weights" : {
+
+      },
+      "dst_w" : 64,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 12,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "tensor",
+      "debug_info" : "804",
+      "dst_seq" : 1,
+      "dst_k" : -1,
+      "top" : "804"
+    },
+    {
+      "axis_h" : 0,
+      "axis_w" : 2,
+      "bottom" : "797",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_61",
+      "top" : "transpose_61",
+      "type" : "transpose",
+      "name" : "transpose_61"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "791",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_62",
+      "top" : "transpose_62",
+      "type" : "transpose",
+      "name" : "transpose_62"
+    },
+    {
+      "bottom" : "transpose_62,transpose_61",
+      "weights" : {
+
+      },
+      "debug_info" : "input.249",
+      "top" : "input.249",
+      "type" : "batch_matmul",
+      "name" : "input.249",
+      "channel_mode" : false
+    },
+    {
+      "bottom" : "input.249",
+      "weights" : {
+
+      },
+      "nd_axis" : -1,
+      "debug_info" : "input.251",
+      "top" : "input.251",
+      "type" : "softmax_nd",
+      "name" : "input.251"
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "804",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_63",
+      "top" : "transpose_63",
+      "type" : "transpose",
+      "name" : "transpose_63"
+    },
+    {
+      "bottom" : "input.251,transpose_63",
+      "weights" : {
+
+      },
+      "debug_info" : "attn_output.45",
+      "top" : "attn_output.45",
+      "type" : "batch_matmul",
+      "name" : "attn_output.45",
+      "channel_mode" : false
+    },
+    {
+      "axis_h" : 2,
+      "axis_w" : 0,
+      "bottom" : "attn_output.45",
+      "axis_k" : 1,
+      "axis_n" : 3,
+      "axis_seq" : 4,
+      "weights" : {
+
+      },
+      "debug_info" : "transpose_60",
+      "top" : "transpose_60",
+      "type" : "transpose",
+      "name" : "transpose_60"
+    },
+    {
+      "name" : "input.253",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "transpose_60",
+      "debug_info" : "input.253",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.253"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.255",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 381,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.253",
+      "blob_biases" : 379,
+      "has_tanh" : 0,
+      "debug_info" : "input.255",
+      "name" : "input.255",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.247,input.255",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.257",
+      "top" : "input.257",
+      "type" : "elementwise",
+      "name" : "input.257",
+      "beta" : 0
+    },
+    {
+      "name" : "input.259_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.257",
+      "debug_info" : "input.259_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "input.259_reshape"
+    },
+    {
+      "bottom" : "input.259_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "input.259_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "input.259_mvn",
+      "type" : "l2_normalize",
+      "name" : "input.259_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.259_scale",
+      "constant_blob" : 383,
+      "top" : "input.259_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.259_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "input.259_scale",
+      "constant_blob" : 385,
+      "top" : "input.259_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "input.259_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "input.259_mvn,input.259_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.259_scale",
+      "top" : "input.259_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "input.259_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "input.259_scale_mul_out,input.259_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input.259_scale",
+      "top" : "input.259_scale",
+      "type" : "elementwise",
+      "name" : "input.259_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "input.259",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input.259_scale",
+      "debug_info" : "input.259",
+      "dst_seq" : 1,
+      "dst_k" : 1,
+      "top" : "input.259"
+    },
+    {
+      "nB" : 768,
+      "top" : "input.261",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 3072,
+      "blob_weights" : 389,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.259",
+      "blob_biases" : 387,
+      "has_tanh" : 0,
+      "debug_info" : "input.261",
+      "name" : "input.261",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.261",
+      "weights" : {
+
+      },
+      "mode" : 19,
+      "debug_info" : "input.263",
+      "top" : "input.263",
+      "type" : "activation",
+      "name" : "input.263"
+    },
+    {
+      "nB" : 3072,
+      "top" : "input.267",
+      "has_biases" : 1,
+      "weights" : {
+
+      },
+      "nC" : 768,
+      "blob_weights" : 393,
+      "type" : "inner_product",
+      "has_relu" : 0,
+      "bottom" : "input.263",
+      "blob_biases" : 391,
+      "has_tanh" : 0,
+      "debug_info" : "input.267",
+      "name" : "input.267",
+      "has_prelu" : 0
+    },
+    {
+      "bottom" : "input.257,input.267",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "input",
+      "top" : "input",
+      "type" : "elementwise",
+      "name" : "input",
+      "beta" : 0
+    },
+    {
+      "name" : "837_reshape",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 4,
+      "type" : "reshape",
+      "dst_h" : 1,
+      "mode" : 0,
+      "dynamic_shape" : false,
+      "bottom" : "input",
+      "debug_info" : "837_reshape",
+      "dst_seq" : 1,
+      "dst_k" : 1500,
+      "top" : "837_reshape"
+    },
+    {
+      "bottom" : "837_reshape",
+      "weights" : {
+
+      },
+      "eps" : 9.9999997473787516e-06,
+      "debug_info" : "837_mvn",
+      "axis" : 1,
+      "normalization_mode" : 1,
+      "top" : "837_mvn",
+      "type" : "l2_normalize",
+      "name" : "837_mvn"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "837_scale",
+      "constant_blob" : 395,
+      "top" : "837_scale_constant_in_scale",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "837_scale_constant_in_scale"
+    },
+    {
+      "bottom" : "",
+      "k" : 1,
+      "n" : 1,
+      "weights" : {
+
+      },
+      "h" : 1,
+      "debug_info" : "837_scale",
+      "constant_blob" : 397,
+      "top" : "837_scale_constant_in_bias",
+      "w" : 768,
+      "type" : "load_constant",
+      "name" : "837_scale_constant_in_bias"
+    },
+    {
+      "bottom" : "837_mvn,837_scale_constant_in_scale",
+      "alpha" : 1,
+      "operation" : 1,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "837_scale",
+      "top" : "837_scale_mul_out",
+      "type" : "elementwise",
+      "name" : "837_scale_mul_out",
+      "beta" : 0
+    },
+    {
+      "bottom" : "837_scale_mul_out,837_scale_constant_in_bias",
+      "alpha" : 1,
+      "operation" : 0,
+      "weights" : {
+
+      },
+      "fused_relu" : 0,
+      "debug_info" : "837_scale",
+      "top" : "837_scale",
+      "type" : "elementwise",
+      "name" : "837_scale",
+      "beta" : 0
+    },
+    {
+      "name" : "837",
+      "weights" : {
+
+      },
+      "dst_w" : 768,
+      "version" : 1,
+      "dst_n" : 1,
+      "dst_nd_rank" : 3,
+      "type" : "reshape",
+      "dst_h" : 1500,
+      "mode" : 0,
+      "attributes" : {
+        "is_output" : 1
+      },
+      "dynamic_shape" : false,
+      "bottom" : "837_scale",
+      "debug_info" : "837",
+      "dst_k" : 1,
+      "dst_seq" : 1,
+      "top" : "last_hidden_state"
+    }
+  ]
+}
\ No newline at end of file