<
   ir_version: 10,
   opset_import: ["" : 23],
   producer_name: "pytorch"
>
main_graph (uint8[batch,224,224,3] image) => (float[batch,512] image_embedding) 
   <
      float[batch,1024,14,14] add_1016
      float[batch,1024,14,14] add_1092
      float[batch,1024,14,14] add_1168
      float[batch,1024,14,14] add_1244
      float[batch,1024,14,14] add_1320
      float[batch,1024,14,14] add_1396
      float[batch,256,56,56] add_140
      float[batch,1024,14,14] add_1472
      float[batch,1024,14,14] add_1548
      float[batch,1024,14,14] add_1624
      float[batch,1024,14,14] add_1700
      float[batch,1024,14,14] add_1776
      float[batch,1024,14,14] add_1852
      float[batch,1024,14,14] add_1928
      float[batch,1024,14,14] add_2004
      float[batch,1024,14,14] add_2080
      float[batch,1024,14,14] add_2156
      float[batch,256,56,56] add_216
      float[batch,1024,14,14] add_2232
      float[batch,1024,14,14] add_2308
      float[batch,1024,14,14] add_2384
      float[batch,2048,7,7] add_2480
      float[batch,2048,7,7] add_2556
      float[batch,2048,7,7] add_2632
      float[50,batch,2048] add_2662
      float[batch,256,56,56] add_292
      float[batch,512,28,28] add_388
      float[batch,512,28,28] add_464
      float[batch,512,28,28] add_540
      float[batch,512,28,28] add_616
      float[batch,1024,14,14] add_712
      float[batch,1024,14,14] add_788
      float[batch,1024,14,14] add_864
      float[batch,1024,14,14] add_940
      float[batch,64,56,56] avg_pool2d
      float[batch,128,28,28] avg_pool2d_2
      float[batch,256,28,28] avg_pool2d_3
      float[batch,256,14,14] avg_pool2d_4
      float[batch,512,14,14] avg_pool2d_5
      float[batch,512,7,7] avg_pool2d_6
      float[batch,1024,7,7] avg_pool2d_7
      float[50,batch,2048] cat
      float[batch,1] clamp_min
      float[batch,32,112,112] getitem
      float[batch,256,14,14] getitem_102
      float[batch,1024,14,14] getitem_105
      float[batch,256,14,14] getitem_108
      float[batch,256,14,14] getitem_111
      float[batch,1024,14,14] getitem_114
      float[batch,256,14,14] getitem_117
      float[batch,64,56,56] getitem_12
      float[batch,256,14,14] getitem_120
      float[batch,1024,14,14] getitem_123
      float[batch,256,14,14] getitem_126
      float[batch,256,14,14] getitem_129
      float[batch,1024,14,14] getitem_132
      float[batch,256,14,14] getitem_135
      float[batch,256,14,14] getitem_138
      float[batch,1024,14,14] getitem_141
      float[batch,256,14,14] getitem_144
      float[batch,256,14,14] getitem_147
      float[batch,256,56,56] getitem_15
      float[batch,1024,14,14] getitem_150
      float[batch,256,14,14] getitem_153
      float[batch,256,14,14] getitem_156
      float[batch,1024,14,14] getitem_159
      float[batch,256,14,14] getitem_162
      float[batch,256,14,14] getitem_165
      float[batch,1024,14,14] getitem_168
      float[batch,256,14,14] getitem_171
      float[batch,256,14,14] getitem_174
      float[batch,1024,14,14] getitem_177
      float[batch,256,56,56] getitem_18
      float[batch,256,14,14] getitem_180
      float[batch,256,14,14] getitem_183
      float[batch,1024,14,14] getitem_186
      float[batch,256,14,14] getitem_189
      float[batch,256,14,14] getitem_192
      float[batch,1024,14,14] getitem_195
      float[batch,256,14,14] getitem_198
      float[batch,256,14,14] getitem_201
      float[batch,1024,14,14] getitem_204
      float[batch,256,14,14] getitem_207
      float[batch,64,56,56] getitem_21
      float[batch,256,14,14] getitem_210
      float[batch,1024,14,14] getitem_213
      float[batch,256,14,14] getitem_216
      float[batch,256,14,14] getitem_219
      float[batch,1024,14,14] getitem_222
      float[batch,256,14,14] getitem_225
      float[batch,256,14,14] getitem_228
      float[batch,1024,14,14] getitem_231
      float[batch,256,14,14] getitem_234
      float[batch,256,14,14] getitem_237
      float[batch,64,56,56] getitem_24
      float[batch,1024,14,14] getitem_240
      float[batch,256,14,14] getitem_243
      float[batch,256,14,14] getitem_246
      float[batch,1024,14,14] getitem_249
      float[batch,256,14,14] getitem_252
      float[batch,256,14,14] getitem_255
      float[batch,1024,14,14] getitem_258
      float[batch,256,14,14] getitem_261
      float[batch,256,14,14] getitem_264
      float[batch,1024,14,14] getitem_267
      float[batch,256,56,56] getitem_27
      float[batch,256,14,14] getitem_270
      float[batch,256,14,14] getitem_273
      float[batch,1024,14,14] getitem_276
      float[batch,256,14,14] getitem_279
      float[batch,256,14,14] getitem_282
      float[batch,1024,14,14] getitem_285
      float[batch,512,14,14] getitem_288
      float[batch,512,14,14] getitem_291
      float[batch,2048,7,7] getitem_294
      float[batch,2048,7,7] getitem_297
      float[batch,32,112,112] getitem_3
      float[batch,64,56,56] getitem_30
      float[batch,512,7,7] getitem_300
      float[batch,512,7,7] getitem_303
      float[batch,2048,7,7] getitem_306
      float[batch,512,7,7] getitem_309
      float[batch,512,7,7] getitem_312
      float[batch,2048,7,7] getitem_315
      float[batch,64,56,56] getitem_33
      float[batch,256,56,56] getitem_36
      float[batch,128,56,56] getitem_39
      float[batch,128,56,56] getitem_42
      float[batch,512,28,28] getitem_45
      float[batch,512,28,28] getitem_48
      float[batch,128,28,28] getitem_51
      float[batch,128,28,28] getitem_54
      float[batch,512,28,28] getitem_57
      float[batch,64,112,112] getitem_6
      float[batch,128,28,28] getitem_60
      float[batch,128,28,28] getitem_63
      float[batch,512,28,28] getitem_66
      float[batch,128,28,28] getitem_69
      float[batch,128,28,28] getitem_72
      float[batch,512,28,28] getitem_75
      float[batch,256,28,28] getitem_78
      float[batch,256,28,28] getitem_81
      float[batch,1024,14,14] getitem_84
      float[batch,1024,14,14] getitem_87
      float[batch,64,56,56] getitem_9
      float[batch,256,14,14] getitem_90
      float[batch,256,14,14] getitem_93
      float[batch,1024,14,14] getitem_96
      float[batch,256,14,14] getitem_99
      float[batch,3,224,224] image_chw
      float[batch,224,224,3] image_f32
      float[batch,224,224,3] image_shifted
      float[batch,1] linalg_vector_norm
      float[1,batch,2048] linear
      float[50,batch,2048] linear_1
      float[50,batch,2048] linear_2
      float[batch,512] linear_3
      float[1,batch,2048] mean
      float[1,batch,2048] node_scaled_dot_product_attention_q_row
      float[49,batch,2048] permute_1
      float[1,batch,32,64] permute_2
      float[batch,32,112,112] relu
      float[batch,32,112,112] relu_1
      float[batch,64,56,56] relu_10
      float[batch,512,7,7] relu_100
      float[batch,2048,7,7] relu_101
      float[batch,256,56,56] relu_11
      float[batch,128,56,56] relu_12
      float[batch,128,56,56] relu_13
      float[batch,512,28,28] relu_14
      float[batch,128,28,28] relu_15
      float[batch,128,28,28] relu_16
      float[batch,512,28,28] relu_17
      float[batch,128,28,28] relu_18
      float[batch,128,28,28] relu_19
      float[batch,64,112,112] relu_2
      float[batch,512,28,28] relu_20
      float[batch,128,28,28] relu_21
      float[batch,128,28,28] relu_22
      float[batch,512,28,28] relu_23
      float[batch,256,28,28] relu_24
      float[batch,256,28,28] relu_25
      float[batch,1024,14,14] relu_26
      float[batch,256,14,14] relu_27
      float[batch,256,14,14] relu_28
      float[batch,1024,14,14] relu_29
      float[batch,64,56,56] relu_3
      float[batch,256,14,14] relu_30
      float[batch,256,14,14] relu_31
      float[batch,1024,14,14] relu_32
      float[batch,256,14,14] relu_33
      float[batch,256,14,14] relu_34
      float[batch,1024,14,14] relu_35
      float[batch,256,14,14] relu_36
      float[batch,256,14,14] relu_37
      float[batch,1024,14,14] relu_38
      float[batch,256,14,14] relu_39
      float[batch,64,56,56] relu_4
      float[batch,256,14,14] relu_40
      float[batch,1024,14,14] relu_41
      float[batch,256,14,14] relu_42
      float[batch,256,14,14] relu_43
      float[batch,1024,14,14] relu_44
      float[batch,256,14,14] relu_45
      float[batch,256,14,14] relu_46
      float[batch,1024,14,14] relu_47
      float[batch,256,14,14] relu_48
      float[batch,256,14,14] relu_49
      float[batch,256,56,56] relu_5
      float[batch,1024,14,14] relu_50
      float[batch,256,14,14] relu_51
      float[batch,256,14,14] relu_52
      float[batch,1024,14,14] relu_53
      float[batch,256,14,14] relu_54
      float[batch,256,14,14] relu_55
      float[batch,1024,14,14] relu_56
      float[batch,256,14,14] relu_57
      float[batch,256,14,14] relu_58
      float[batch,1024,14,14] relu_59
      float[batch,64,56,56] relu_6
      float[batch,256,14,14] relu_60
      float[batch,256,14,14] relu_61
      float[batch,1024,14,14] relu_62
      float[batch,256,14,14] relu_63
      float[batch,256,14,14] relu_64
      float[batch,1024,14,14] relu_65
      float[batch,256,14,14] relu_66
      float[batch,256,14,14] relu_67
      float[batch,1024,14,14] relu_68
      float[batch,256,14,14] relu_69
      float[batch,64,56,56] relu_7
      float[batch,256,14,14] relu_70
      float[batch,1024,14,14] relu_71
      float[batch,256,14,14] relu_72
      float[batch,256,14,14] relu_73
      float[batch,1024,14,14] relu_74
      float[batch,256,14,14] relu_75
      float[batch,256,14,14] relu_76
      float[batch,1024,14,14] relu_77
      float[batch,256,14,14] relu_78
      float[batch,256,14,14] relu_79
      float[batch,256,56,56] relu_8
      float[batch,1024,14,14] relu_80
      float[batch,256,14,14] relu_81
      float[batch,256,14,14] relu_82
      float[batch,1024,14,14] relu_83
      float[batch,256,14,14] relu_84
      float[batch,256,14,14] relu_85
      float[batch,1024,14,14] relu_86
      float[batch,256,14,14] relu_87
      float[batch,256,14,14] relu_88
      float[batch,1024,14,14] relu_89
      float[batch,64,56,56] relu_9
      float[batch,256,14,14] relu_90
      float[batch,256,14,14] relu_91
      float[batch,1024,14,14] relu_92
      float[batch,512,14,14] relu_93
      float[batch,512,14,14] relu_94
      float[batch,2048,7,7] relu_95
      float[batch,512,7,7] relu_96
      float[batch,512,7,7] relu_97
      float[batch,2048,7,7] relu_98
      float[batch,512,7,7] relu_99
      float[batch,32,1,64] scaled_dot_product_attention
      float[batch,512] select
      float[2048] split_split_0
      float[2048] split_split_1
      float[2048] split_split_2
      float[unk__1,1,64] transpose
      float[unk__1,50,64] transpose_1
      float[unk__1,50,64] transpose_2
      float[50,1,2048] unsqueeze
      float[1,batch,2048] val_7
      float[50,batch,2048] val_8
      float[50,batch,2048] val_9
      float[1,batch,512] view_10
      float[batch,2048,49] view_2
      float[1,unk__1,64] view_3
      float[50,unk__1,64] view_4
      float[50,unk__1,64] view_5
      float[batch,32,1,64] view_6
      float[batch,32,50,64] view_7
      float[batch,32,50,64] view_8
      float[batch,2048] view_9
   >
{
   [pre_cast] image_f32 = Cast <to: int = 1> (image)
   [pre_shift] image_shifted = Sub (image_f32, image_shift)
   [pre_nhwc_to_nchw] image_chw = Transpose <perm: ints = [0, 3, 1, 2]> (image_shifted)
   getitem = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [2, 2]> (image_chw, "visual.conv1.weight", "visual.conv1.weight_bias")
   [node_relu] relu = Relu (getitem)
   getitem_3 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu, "visual.conv2.weight", "visual.conv2.weight_bias")
   relu_1 = Relu (getitem_3)
   getitem_6 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_1, "visual.conv3.weight", "visual.conv3.weight_bias")
   relu_2 = Relu (getitem_6)
   [node_avg_pool2d] avg_pool2d = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_2)
   getitem_9 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d, "visual.layer1.0.conv1.weight", "visual.layer1.0.conv1.weight_bias")
   relu_3 = Relu (getitem_9)
   getitem_12 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_3, "visual.layer1.0.conv2.weight", "visual.layer1.0.conv2.weight_bias")
   relu_4 = Relu (getitem_12)
   getitem_15 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_4, "visual.layer1.0.conv3.weight", "visual.layer1.0.conv3.weight_bias")
   getitem_18 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d, "visual.layer1.0.downsample.0.weight", "visual.layer1.0.downsample.0.weight_bias")
   add_140 = Add (getitem_15, getitem_18)
   relu_5 = Relu (add_140)
   getitem_21 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_5, "visual.layer1.1.conv1.weight", "visual.layer1.1.conv1.weight_bias")
   relu_6 = Relu (getitem_21)
   getitem_24 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_6, "visual.layer1.1.conv2.weight", "visual.layer1.1.conv2.weight_bias")
   relu_7 = Relu (getitem_24)
   getitem_27 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_7, "visual.layer1.1.conv3.weight", "visual.layer1.1.conv3.weight_bias")
   add_216 = Add (getitem_27, relu_5)
   relu_8 = Relu (add_216)
   getitem_30 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_8, "visual.layer1.2.conv1.weight", "visual.layer1.2.conv1.weight_bias")
   relu_9 = Relu (getitem_30)
   getitem_33 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_9, "visual.layer1.2.conv2.weight", "visual.layer1.2.conv2.weight_bias")
   relu_10 = Relu (getitem_33)
   getitem_36 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_10, "visual.layer1.2.conv3.weight", "visual.layer1.2.conv3.weight_bias")
   add_292 = Add (getitem_36, relu_8)
   relu_11 = Relu (add_292)
   getitem_39 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_11, "visual.layer2.0.conv1.weight", "visual.layer2.0.conv1.weight_bias")
   relu_12 = Relu (getitem_39)
   getitem_42 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_12, "visual.layer2.0.conv2.weight", "visual.layer2.0.conv2.weight_bias")
   relu_13 = Relu (getitem_42)
   avg_pool2d_2 = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_13)
   getitem_45 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d_2, "visual.layer2.0.conv3.weight", "visual.layer2.0.conv3.weight_bias")
   avg_pool2d_3 = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_11)
   getitem_48 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d_3, "visual.layer2.0.downsample.0.weight", "visual.layer2.0.downsample.0.weight_bias")
   add_388 = Add (getitem_45, getitem_48)
   relu_14 = Relu (add_388)
   getitem_51 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_14, "visual.layer2.1.conv1.weight", "visual.layer2.1.conv1.weight_bias")
   relu_15 = Relu (getitem_51)
   getitem_54 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_15, "visual.layer2.1.conv2.weight", "visual.layer2.1.conv2.weight_bias")
   relu_16 = Relu (getitem_54)
   getitem_57 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_16, "visual.layer2.1.conv3.weight", "visual.layer2.1.conv3.weight_bias")
   add_464 = Add (getitem_57, relu_14)
   relu_17 = Relu (add_464)
   getitem_60 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_17, "visual.layer2.2.conv1.weight", "visual.layer2.2.conv1.weight_bias")
   relu_18 = Relu (getitem_60)
   getitem_63 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_18, "visual.layer2.2.conv2.weight", "visual.layer2.2.conv2.weight_bias")
   relu_19 = Relu (getitem_63)
   getitem_66 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_19, "visual.layer2.2.conv3.weight", "visual.layer2.2.conv3.weight_bias")
   add_540 = Add (getitem_66, relu_17)
   relu_20 = Relu (add_540)
   getitem_69 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_20, "visual.layer2.3.conv1.weight", "visual.layer2.3.conv1.weight_bias")
   relu_21 = Relu (getitem_69)
   getitem_72 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_21, "visual.layer2.3.conv2.weight", "visual.layer2.3.conv2.weight_bias")
   relu_22 = Relu (getitem_72)
   getitem_75 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_22, "visual.layer2.3.conv3.weight", "visual.layer2.3.conv3.weight_bias")
   add_616 = Add (getitem_75, relu_20)
   relu_23 = Relu (add_616)
   getitem_78 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_23, "visual.layer3.0.conv1.weight", "visual.layer3.0.conv1.weight_bias")
   relu_24 = Relu (getitem_78)
   getitem_81 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_24, "visual.layer3.0.conv2.weight", "visual.layer3.0.conv2.weight_bias")
   relu_25 = Relu (getitem_81)
   avg_pool2d_4 = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_25)
   getitem_84 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d_4, "visual.layer3.0.conv3.weight", "visual.layer3.0.conv3.weight_bias")
   avg_pool2d_5 = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_23)
   getitem_87 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d_5, "visual.layer3.0.downsample.0.weight", "visual.layer3.0.downsample.0.weight_bias")
   add_712 = Add (getitem_84, getitem_87)
   relu_26 = Relu (add_712)
   getitem_90 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_26, "visual.layer3.1.conv1.weight", "visual.layer3.1.conv1.weight_bias")
   relu_27 = Relu (getitem_90)
   getitem_93 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_27, "visual.layer3.1.conv2.weight", "visual.layer3.1.conv2.weight_bias")
   relu_28 = Relu (getitem_93)
   getitem_96 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_28, "visual.layer3.1.conv3.weight", "visual.layer3.1.conv3.weight_bias")
   add_788 = Add (getitem_96, relu_26)
   relu_29 = Relu (add_788)
   getitem_99 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_29, "visual.layer3.2.conv1.weight", "visual.layer3.2.conv1.weight_bias")
   relu_30 = Relu (getitem_99)
   getitem_102 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_30, "visual.layer3.2.conv2.weight", "visual.layer3.2.conv2.weight_bias")
   relu_31 = Relu (getitem_102)
   getitem_105 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_31, "visual.layer3.2.conv3.weight", "visual.layer3.2.conv3.weight_bias")
   add_864 = Add (getitem_105, relu_29)
   relu_32 = Relu (add_864)
   getitem_108 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_32, "visual.layer3.3.conv1.weight", "visual.layer3.3.conv1.weight_bias")
   relu_33 = Relu (getitem_108)
   getitem_111 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_33, "visual.layer3.3.conv2.weight", "visual.layer3.3.conv2.weight_bias")
   relu_34 = Relu (getitem_111)
   getitem_114 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_34, "visual.layer3.3.conv3.weight", "visual.layer3.3.conv3.weight_bias")
   add_940 = Add (getitem_114, relu_32)
   relu_35 = Relu (add_940)
   getitem_117 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_35, "visual.layer3.4.conv1.weight", "visual.layer3.4.conv1.weight_bias")
   relu_36 = Relu (getitem_117)
   getitem_120 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_36, "visual.layer3.4.conv2.weight", "visual.layer3.4.conv2.weight_bias")
   relu_37 = Relu (getitem_120)
   getitem_123 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_37, "visual.layer3.4.conv3.weight", "visual.layer3.4.conv3.weight_bias")
   add_1016 = Add (getitem_123, relu_35)
   relu_38 = Relu (add_1016)
   getitem_126 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_38, "visual.layer3.5.conv1.weight", "visual.layer3.5.conv1.weight_bias")
   relu_39 = Relu (getitem_126)
   getitem_129 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_39, "visual.layer3.5.conv2.weight", "visual.layer3.5.conv2.weight_bias")
   relu_40 = Relu (getitem_129)
   getitem_132 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_40, "visual.layer3.5.conv3.weight", "visual.layer3.5.conv3.weight_bias")
   add_1092 = Add (getitem_132, relu_38)
   relu_41 = Relu (add_1092)
   getitem_135 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_41, "visual.layer3.6.conv1.weight", "visual.layer3.6.conv1.weight_bias")
   relu_42 = Relu (getitem_135)
   getitem_138 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_42, "visual.layer3.6.conv2.weight", "visual.layer3.6.conv2.weight_bias")
   relu_43 = Relu (getitem_138)
   getitem_141 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_43, "visual.layer3.6.conv3.weight", "visual.layer3.6.conv3.weight_bias")
   add_1168 = Add (getitem_141, relu_41)
   relu_44 = Relu (add_1168)
   getitem_144 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_44, "visual.layer3.7.conv1.weight", "visual.layer3.7.conv1.weight_bias")
   relu_45 = Relu (getitem_144)
   getitem_147 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_45, "visual.layer3.7.conv2.weight", "visual.layer3.7.conv2.weight_bias")
   relu_46 = Relu (getitem_147)
   getitem_150 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_46, "visual.layer3.7.conv3.weight", "visual.layer3.7.conv3.weight_bias")
   add_1244 = Add (getitem_150, relu_44)
   relu_47 = Relu (add_1244)
   getitem_153 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_47, "visual.layer3.8.conv1.weight", "visual.layer3.8.conv1.weight_bias")
   relu_48 = Relu (getitem_153)
   getitem_156 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_48, "visual.layer3.8.conv2.weight", "visual.layer3.8.conv2.weight_bias")
   relu_49 = Relu (getitem_156)
   getitem_159 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_49, "visual.layer3.8.conv3.weight", "visual.layer3.8.conv3.weight_bias")
   add_1320 = Add (getitem_159, relu_47)
   relu_50 = Relu (add_1320)
   getitem_162 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_50, "visual.layer3.9.conv1.weight", "visual.layer3.9.conv1.weight_bias")
   relu_51 = Relu (getitem_162)
   getitem_165 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_51, "visual.layer3.9.conv2.weight", "visual.layer3.9.conv2.weight_bias")
   relu_52 = Relu (getitem_165)
   getitem_168 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_52, "visual.layer3.9.conv3.weight", "visual.layer3.9.conv3.weight_bias")
   add_1396 = Add (getitem_168, relu_50)
   relu_53 = Relu (add_1396)
   getitem_171 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_53, "visual.layer3.10.conv1.weight", "visual.layer3.10.conv1.weight_bias")
   relu_54 = Relu (getitem_171)
   getitem_174 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_54, "visual.layer3.10.conv2.weight", "visual.layer3.10.conv2.weight_bias")
   relu_55 = Relu (getitem_174)
   getitem_177 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_55, "visual.layer3.10.conv3.weight", "visual.layer3.10.conv3.weight_bias")
   add_1472 = Add (getitem_177, relu_53)
   relu_56 = Relu (add_1472)
   getitem_180 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_56, "visual.layer3.11.conv1.weight", "visual.layer3.11.conv1.weight_bias")
   relu_57 = Relu (getitem_180)
   getitem_183 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_57, "visual.layer3.11.conv2.weight", "visual.layer3.11.conv2.weight_bias")
   relu_58 = Relu (getitem_183)
   getitem_186 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_58, "visual.layer3.11.conv3.weight", "visual.layer3.11.conv3.weight_bias")
   add_1548 = Add (getitem_186, relu_56)
   relu_59 = Relu (add_1548)
   getitem_189 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_59, "visual.layer3.12.conv1.weight", "visual.layer3.12.conv1.weight_bias")
   relu_60 = Relu (getitem_189)
   getitem_192 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_60, "visual.layer3.12.conv2.weight", "visual.layer3.12.conv2.weight_bias")
   relu_61 = Relu (getitem_192)
   getitem_195 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_61, "visual.layer3.12.conv3.weight", "visual.layer3.12.conv3.weight_bias")
   add_1624 = Add (getitem_195, relu_59)
   relu_62 = Relu (add_1624)
   getitem_198 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_62, "visual.layer3.13.conv1.weight", "visual.layer3.13.conv1.weight_bias")
   relu_63 = Relu (getitem_198)
   getitem_201 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_63, "visual.layer3.13.conv2.weight", "visual.layer3.13.conv2.weight_bias")
   relu_64 = Relu (getitem_201)
   getitem_204 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_64, "visual.layer3.13.conv3.weight", "visual.layer3.13.conv3.weight_bias")
   add_1700 = Add (getitem_204, relu_62)
   relu_65 = Relu (add_1700)
   getitem_207 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_65, "visual.layer3.14.conv1.weight", "visual.layer3.14.conv1.weight_bias")
   relu_66 = Relu (getitem_207)
   getitem_210 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_66, "visual.layer3.14.conv2.weight", "visual.layer3.14.conv2.weight_bias")
   relu_67 = Relu (getitem_210)
   getitem_213 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_67, "visual.layer3.14.conv3.weight", "visual.layer3.14.conv3.weight_bias")
   add_1776 = Add (getitem_213, relu_65)
   relu_68 = Relu (add_1776)
   getitem_216 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_68, "visual.layer3.15.conv1.weight", "visual.layer3.15.conv1.weight_bias")
   relu_69 = Relu (getitem_216)
   getitem_219 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_69, "visual.layer3.15.conv2.weight", "visual.layer3.15.conv2.weight_bias")
   relu_70 = Relu (getitem_219)
   getitem_222 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_70, "visual.layer3.15.conv3.weight", "visual.layer3.15.conv3.weight_bias")
   add_1852 = Add (getitem_222, relu_68)
   relu_71 = Relu (add_1852)
   getitem_225 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_71, "visual.layer3.16.conv1.weight", "visual.layer3.16.conv1.weight_bias")
   relu_72 = Relu (getitem_225)
   getitem_228 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_72, "visual.layer3.16.conv2.weight", "visual.layer3.16.conv2.weight_bias")
   relu_73 = Relu (getitem_228)
   getitem_231 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_73, "visual.layer3.16.conv3.weight", "visual.layer3.16.conv3.weight_bias")
   add_1928 = Add (getitem_231, relu_71)
   relu_74 = Relu (add_1928)
   getitem_234 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_74, "visual.layer3.17.conv1.weight", "visual.layer3.17.conv1.weight_bias")
   relu_75 = Relu (getitem_234)
   getitem_237 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_75, "visual.layer3.17.conv2.weight", "visual.layer3.17.conv2.weight_bias")
   relu_76 = Relu (getitem_237)
   getitem_240 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_76, "visual.layer3.17.conv3.weight", "visual.layer3.17.conv3.weight_bias")
   add_2004 = Add (getitem_240, relu_74)
   relu_77 = Relu (add_2004)
   getitem_243 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_77, "visual.layer3.18.conv1.weight", "visual.layer3.18.conv1.weight_bias")
   relu_78 = Relu (getitem_243)
   getitem_246 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_78, "visual.layer3.18.conv2.weight", "visual.layer3.18.conv2.weight_bias")
   relu_79 = Relu (getitem_246)
   getitem_249 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_79, "visual.layer3.18.conv3.weight", "visual.layer3.18.conv3.weight_bias")
   add_2080 = Add (getitem_249, relu_77)
   relu_80 = Relu (add_2080)
   getitem_252 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_80, "visual.layer3.19.conv1.weight", "visual.layer3.19.conv1.weight_bias")
   relu_81 = Relu (getitem_252)
   getitem_255 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_81, "visual.layer3.19.conv2.weight", "visual.layer3.19.conv2.weight_bias")
   relu_82 = Relu (getitem_255)
   getitem_258 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_82, "visual.layer3.19.conv3.weight", "visual.layer3.19.conv3.weight_bias")
   add_2156 = Add (getitem_258, relu_80)
   relu_83 = Relu (add_2156)
   getitem_261 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_83, "visual.layer3.20.conv1.weight", "visual.layer3.20.conv1.weight_bias")
   relu_84 = Relu (getitem_261)
   getitem_264 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_84, "visual.layer3.20.conv2.weight", "visual.layer3.20.conv2.weight_bias")
   relu_85 = Relu (getitem_264)
   getitem_267 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_85, "visual.layer3.20.conv3.weight", "visual.layer3.20.conv3.weight_bias")
   add_2232 = Add (getitem_267, relu_83)
   relu_86 = Relu (add_2232)
   getitem_270 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_86, "visual.layer3.21.conv1.weight", "visual.layer3.21.conv1.weight_bias")
   relu_87 = Relu (getitem_270)
   getitem_273 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_87, "visual.layer3.21.conv2.weight", "visual.layer3.21.conv2.weight_bias")
   relu_88 = Relu (getitem_273)
   getitem_276 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_88, "visual.layer3.21.conv3.weight", "visual.layer3.21.conv3.weight_bias")
   add_2308 = Add (getitem_276, relu_86)
   relu_89 = Relu (add_2308)
   getitem_279 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_89, "visual.layer3.22.conv1.weight", "visual.layer3.22.conv1.weight_bias")
   relu_90 = Relu (getitem_279)
   getitem_282 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_90, "visual.layer3.22.conv2.weight", "visual.layer3.22.conv2.weight_bias")
   relu_91 = Relu (getitem_282)
   getitem_285 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_91, "visual.layer3.22.conv3.weight", "visual.layer3.22.conv3.weight_bias")
   add_2384 = Add (getitem_285, relu_89)
   relu_92 = Relu (add_2384)
   getitem_288 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_92, "visual.layer4.0.conv1.weight", "visual.layer4.0.conv1.weight_bias")
   relu_93 = Relu (getitem_288)
   getitem_291 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_93, "visual.layer4.0.conv2.weight", "visual.layer4.0.conv2.weight_bias")
   relu_94 = Relu (getitem_291)
   avg_pool2d_6 = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_94)
   getitem_294 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d_6, "visual.layer4.0.conv3.weight", "visual.layer4.0.conv3.weight_bias")
   avg_pool2d_7 = AveragePool <auto_pad: string = "NOTSET", ceil_mode: int = 0, count_include_pad: int = 1, kernel_shape: ints = [2, 2], pads: ints = [0, 0, 0, 0], strides: ints = [2, 2]> (relu_92)
   getitem_297 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (avg_pool2d_7, "visual.layer4.0.downsample.0.weight", "visual.layer4.0.downsample.0.weight_bias")
   add_2480 = Add (getitem_294, getitem_297)
   relu_95 = Relu (add_2480)
   getitem_300 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_95, "visual.layer4.1.conv1.weight", "visual.layer4.1.conv1.weight_bias")
   relu_96 = Relu (getitem_300)
   getitem_303 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_96, "visual.layer4.1.conv2.weight", "visual.layer4.1.conv2.weight_bias")
   relu_97 = Relu (getitem_303)
   getitem_306 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_97, "visual.layer4.1.conv3.weight", "visual.layer4.1.conv3.weight_bias")
   add_2556 = Add (getitem_306, relu_95)
   relu_98 = Relu (add_2556)
   getitem_309 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_98, "visual.layer4.2.conv1.weight", "visual.layer4.2.conv1.weight_bias")
   relu_99 = Relu (getitem_309)
   getitem_312 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [1, 1, 1, 1], strides: ints = [1, 1]> (relu_99, "visual.layer4.2.conv2.weight", "visual.layer4.2.conv2.weight_bias")
   relu_100 = Relu (getitem_312)
   getitem_315 = Conv <auto_pad: string = "NOTSET", dilations: ints = [1, 1], group: int = 1, pads: ints = [0, 0, 0, 0], strides: ints = [1, 1]> (relu_100, "visual.layer4.2.conv3.weight", "visual.layer4.2.conv3.weight_bias")
   add_2632 = Add (getitem_315, relu_98)
   relu_101 = Relu (add_2632)
   view_2 = Reshape <allowzero: int = 1> (relu_101, view_2_target)
   permute_1 = Transpose <perm: ints = [2, 0, 1]> (view_2)
   [node_mean] mean = ReduceMean <keepdims: int = 1, noop_with_empty_axes: int = 0> (permute_1, val_0)
   [node_cat] cat = Concat <axis: int = 0> (mean, permute_1)
   [node_unsqueeze] unsqueeze = Unsqueeze ("visual.attnpool.positional_embedding", val_3)
   add_2662 = Add (cat, unsqueeze)
   split_split_0, split_split_1, split_split_2 = Split <axis: int = 0, num_outputs: int = 3> (cat_1)
   node_scaled_dot_product_attention_q_row = Slice (add_2662, val_0, val_3, val_0)
   val_7 = MatMul (node_scaled_dot_product_attention_q_row, val_4)
   [node_linear] linear = Add (val_7, split_split_0)
   val_8 = MatMul (add_2662, val_5)
   linear_1 = Add (val_8, split_split_1)
   val_9 = MatMul (add_2662, val_6)
   linear_2 = Add (val_9, split_split_2)
   view_3 = Reshape <allowzero: int = 1> (linear, node_scaled_dot_product_attention_q_unpack_1)
   [node_transpose] transpose = Transpose <perm: ints = [1, 0, 2]> (view_3)
   view_4 = Reshape <allowzero: int = 1> (linear_1, view_4_target)
   transpose_1 = Transpose <perm: ints = [1, 0, 2]> (view_4)
   view_5 = Reshape <allowzero: int = 1> (linear_2, view_4_target)
   transpose_2 = Transpose <perm: ints = [1, 0, 2]> (view_5)
   view_6 = Reshape <allowzero: int = 1> (transpose, node_scaled_dot_product_attention_q_pack_1)
   view_7 = Reshape <allowzero: int = 1> (transpose_1, view_7_target)
   view_8 = Reshape <allowzero: int = 1> (transpose_2, view_7_target)
   [node_scaled_dot_product_attention] scaled_dot_product_attention = Attention <is_causal: int = 0, qk_matmul_output_mode: int = 0, softcap: float = 0> (view_6, view_7, view_8)
   permute_2 = Transpose <perm: ints = [2, 0, 1, 3]> (scaled_dot_product_attention)
   view_9 = Reshape <allowzero: int = 1> (permute_2, view_9_target)
   linear_3 = Gemm <alpha: float = 1, beta: float = 1, transA: int = 0, transB: int = 1> (view_9, "visual.attnpool.c_proj.weight", "visual.attnpool.c_proj.bias")
   view_10 = Reshape <allowzero: int = 1> (linear_3, node_scaled_dot_product_attention_out_1)
   select = Squeeze (view_10, val_0)
   [node_linalg_vector_norm] linalg_vector_norm = ReduceL2 <keepdims: int = 1, noop_with_empty_axes: int = 0> (select, val_1)
   [node_clamp_min] clamp_min = Clip (linalg_vector_norm, val_2)
   [node_div] image_embedding = Div (select, clamp_min)
}

weights:
cat_1 FLOAT[6144] 43b166ed7897
image_shift FLOAT[3] 2f7a50e604ad
node_scaled_dot_product_attention_out_1 INT64[3] b6639c8d94ec
node_scaled_dot_product_attention_q_pack_1 INT64[4] f7b9b5620534
node_scaled_dot_product_attention_q_unpack_1 INT64[3] cfe34a386daf
val_0 INT64[1] af5570f5a181
val_1 INT64[1] 12a3ae445661
val_2 FLOAT[] 6708d9be4956
val_3 INT64[1] 7c9fa136d441
val_4 FLOAT[2048,2048] 1830071595a0
val_5 FLOAT[2048,2048] 0d0c88316b4a
val_6 FLOAT[2048,2048] 2715735c2c1a
view_2_target INT64[3] 68ffb9ecb5ec
view_4_target INT64[3] c7a3d94c4eb2
view_7_target INT64[4] e0ea1841387b
view_9_target INT64[2] 76aecb4697fd
visual.attnpool.c_proj.bias FLOAT[512] e318e3e8d1b8
visual.attnpool.c_proj.weight FLOAT[512,2048] d9262bb69358
visual.attnpool.positional_embedding FLOAT[50,2048] 34fb1db035ae
visual.conv1.weight FLOAT[32,3,3,3] 64abe16c2280
visual.conv1.weight_bias FLOAT[32] 124e690ec767
visual.conv2.weight FLOAT[32,32,3,3] a105b19f5ae3
visual.conv2.weight_bias FLOAT[32] dacb7232414b
visual.conv3.weight FLOAT[64,32,3,3] 60b15252b92e
visual.conv3.weight_bias FLOAT[64] f226ddd0b29f
visual.layer1.0.conv1.weight FLOAT[64,64,1,1] e195c32f5591
visual.layer1.0.conv1.weight_bias FLOAT[64] d2ca9824261c
visual.layer1.0.conv2.weight FLOAT[64,64,3,3] 7b7759b1f908
visual.layer1.0.conv2.weight_bias FLOAT[64] 24675ace9379
visual.layer1.0.conv3.weight FLOAT[256,64,1,1] 164ee686902a
visual.layer1.0.conv3.weight_bias FLOAT[256] 1fcd5d5e8c80
visual.layer1.0.downsample.0.weight FLOAT[256,64,1,1] bef1164339fd
visual.layer1.0.downsample.0.weight_bias FLOAT[256] 8e2c9e8ee01f
visual.layer1.1.conv1.weight FLOAT[64,256,1,1] a15a64b1292f
visual.layer1.1.conv1.weight_bias FLOAT[64] 98e61e621efc
visual.layer1.1.conv2.weight FLOAT[64,64,3,3] 144c8d32188c
visual.layer1.1.conv2.weight_bias FLOAT[64] e00a89dbf4f0
visual.layer1.1.conv3.weight FLOAT[256,64,1,1] 542aae81da85
visual.layer1.1.conv3.weight_bias FLOAT[256] 2827aa5546c4
visual.layer1.2.conv1.weight FLOAT[64,256,1,1] 538f399c5716
visual.layer1.2.conv1.weight_bias FLOAT[64] 96d99ca66fa7
visual.layer1.2.conv2.weight FLOAT[64,64,3,3] e58a9c4314b7
visual.layer1.2.conv2.weight_bias FLOAT[64] 365d8479aaaa
visual.layer1.2.conv3.weight FLOAT[256,64,1,1] 20fae160bef4
visual.layer1.2.conv3.weight_bias FLOAT[256] 180418939541
visual.layer2.0.conv1.weight FLOAT[128,256,1,1] bf7f3546e69f
visual.layer2.0.conv1.weight_bias FLOAT[128] 19ae59f827a0
visual.layer2.0.conv2.weight FLOAT[128,128,3,3] 9abd5129aeba
visual.layer2.0.conv2.weight_bias FLOAT[128] 943d4fcd74bc
visual.layer2.0.conv3.weight FLOAT[512,128,1,1] 9707078edbd2
visual.layer2.0.conv3.weight_bias FLOAT[512] c4d81efc28ca
visual.layer2.0.downsample.0.weight FLOAT[512,256,1,1] c6caa02d8537
visual.layer2.0.downsample.0.weight_bias FLOAT[512] 3966ec107a2b
visual.layer2.1.conv1.weight FLOAT[128,512,1,1] 0aa4b8b16d09
visual.layer2.1.conv1.weight_bias FLOAT[128] 0fc98e7d30d0
visual.layer2.1.conv2.weight FLOAT[128,128,3,3] e74d0ce70e23
visual.layer2.1.conv2.weight_bias FLOAT[128] 3fb8c3c32826
visual.layer2.1.conv3.weight FLOAT[512,128,1,1] 07522ea89d64
visual.layer2.1.conv3.weight_bias FLOAT[512] fe4778ae2f35
visual.layer2.2.conv1.weight FLOAT[128,512,1,1] 27f8d8b63847
visual.layer2.2.conv1.weight_bias FLOAT[128] 8c848b272f84
visual.layer2.2.conv2.weight FLOAT[128,128,3,3] 6e3d210305b6
visual.layer2.2.conv2.weight_bias FLOAT[128] d786b710ee57
visual.layer2.2.conv3.weight FLOAT[512,128,1,1] 61088d0c1408
visual.layer2.2.conv3.weight_bias FLOAT[512] 8f3ac9629094
visual.layer2.3.conv1.weight FLOAT[128,512,1,1] 83c0d544ba19
visual.layer2.3.conv1.weight_bias FLOAT[128] 9139c65714bb
visual.layer2.3.conv2.weight FLOAT[128,128,3,3] 5a92a46b0f7f
visual.layer2.3.conv2.weight_bias FLOAT[128] 56179726dc82
visual.layer2.3.conv3.weight FLOAT[512,128,1,1] 3935cdc24cc8
visual.layer2.3.conv3.weight_bias FLOAT[512] ec827bd45606
visual.layer3.0.conv1.weight FLOAT[256,512,1,1] e37983058e82
visual.layer3.0.conv1.weight_bias FLOAT[256] d91a009e585b
visual.layer3.0.conv2.weight FLOAT[256,256,3,3] d42d6104446d
visual.layer3.0.conv2.weight_bias FLOAT[256] 8b9969e0c8c0
visual.layer3.0.conv3.weight FLOAT[1024,256,1,1] d8c78ce9ef74
visual.layer3.0.conv3.weight_bias FLOAT[1024] 950b201ba19f
visual.layer3.0.downsample.0.weight FLOAT[1024,512,1,1] 3747fbb0d7dd
visual.layer3.0.downsample.0.weight_bias FLOAT[1024] 81cf7a0d5bcf
visual.layer3.1.conv1.weight FLOAT[256,1024,1,1] 0f7cbb3ccbc5
visual.layer3.1.conv1.weight_bias FLOAT[256] 468421dc2fe0
visual.layer3.1.conv2.weight FLOAT[256,256,3,3] 8799b1c2a315
visual.layer3.1.conv2.weight_bias FLOAT[256] f848659b7ccb
visual.layer3.1.conv3.weight FLOAT[1024,256,1,1] cc3a04bd880a
visual.layer3.1.conv3.weight_bias FLOAT[1024] 0d490e7cd4de
visual.layer3.10.conv1.weight FLOAT[256,1024,1,1] eff05ebef0a6
visual.layer3.10.conv1.weight_bias FLOAT[256] 557f1b79b311
visual.layer3.10.conv2.weight FLOAT[256,256,3,3] dd8271cac6e0
visual.layer3.10.conv2.weight_bias FLOAT[256] 7ab27b1d6e2b
visual.layer3.10.conv3.weight FLOAT[1024,256,1,1] 19617be1cb83
visual.layer3.10.conv3.weight_bias FLOAT[1024] e8475a5e9427
visual.layer3.11.conv1.weight FLOAT[256,1024,1,1] d7258f7f9ecf
visual.layer3.11.conv1.weight_bias FLOAT[256] 022250d83dc7
visual.layer3.11.conv2.weight FLOAT[256,256,3,3] b76ac1b5f4db
visual.layer3.11.conv2.weight_bias FLOAT[256] 066f410cc1e9
visual.layer3.11.conv3.weight FLOAT[1024,256,1,1] 9ab6c7fcf594
visual.layer3.11.conv3.weight_bias FLOAT[1024] d28e5671632e
visual.layer3.12.conv1.weight FLOAT[256,1024,1,1] f101d8db595a
visual.layer3.12.conv1.weight_bias FLOAT[256] 00683f611205
visual.layer3.12.conv2.weight FLOAT[256,256,3,3] a249af066395
visual.layer3.12.conv2.weight_bias FLOAT[256] 5341715541bf
visual.layer3.12.conv3.weight FLOAT[1024,256,1,1] 715a68d9f254
visual.layer3.12.conv3.weight_bias FLOAT[1024] 79464a731dbd
visual.layer3.13.conv1.weight FLOAT[256,1024,1,1] 238fee3e9b7e
visual.layer3.13.conv1.weight_bias FLOAT[256] 509670797ab2
visual.layer3.13.conv2.weight FLOAT[256,256,3,3] 8b429266d488
visual.layer3.13.conv2.weight_bias FLOAT[256] a966c4c47166
visual.layer3.13.conv3.weight FLOAT[1024,256,1,1] 60d4a61f3178
visual.layer3.13.conv3.weight_bias FLOAT[1024] 65cdf2f53817
visual.layer3.14.conv1.weight FLOAT[256,1024,1,1] 93e604164188
visual.layer3.14.conv1.weight_bias FLOAT[256] 90c272f28259
visual.layer3.14.conv2.weight FLOAT[256,256,3,3] bbbe57156768
visual.layer3.14.conv2.weight_bias FLOAT[256] da20369c55fc
visual.layer3.14.conv3.weight FLOAT[1024,256,1,1] 9965e4ec1ac6
visual.layer3.14.conv3.weight_bias FLOAT[1024] cf494f7d6c3f
visual.layer3.15.conv1.weight FLOAT[256,1024,1,1] b65355538dff
visual.layer3.15.conv1.weight_bias FLOAT[256] 47dfeaeddae8
visual.layer3.15.conv2.weight FLOAT[256,256,3,3] d746daee990d
visual.layer3.15.conv2.weight_bias FLOAT[256] 7775c1101726
visual.layer3.15.conv3.weight FLOAT[1024,256,1,1] 7bd988155a96
visual.layer3.15.conv3.weight_bias FLOAT[1024] 635de314fbe5
visual.layer3.16.conv1.weight FLOAT[256,1024,1,1] 8c72a7f5d24b
visual.layer3.16.conv1.weight_bias FLOAT[256] 46c5adcf8919
visual.layer3.16.conv2.weight FLOAT[256,256,3,3] 628024376db8
visual.layer3.16.conv2.weight_bias FLOAT[256] 53625854e635
visual.layer3.16.conv3.weight FLOAT[1024,256,1,1] c40c4360ee14
visual.layer3.16.conv3.weight_bias FLOAT[1024] e2578a7bc679
visual.layer3.17.conv1.weight FLOAT[256,1024,1,1] 35a2ad37cfe6
visual.layer3.17.conv1.weight_bias FLOAT[256] 01ad0ebee036
visual.layer3.17.conv2.weight FLOAT[256,256,3,3] 20754130d77a
visual.layer3.17.conv2.weight_bias FLOAT[256] ec196fda11dc
visual.layer3.17.conv3.weight FLOAT[1024,256,1,1] ceb74147d09d
visual.layer3.17.conv3.weight_bias FLOAT[1024] ffd9373f84f1
visual.layer3.18.conv1.weight FLOAT[256,1024,1,1] e581d085e34f
visual.layer3.18.conv1.weight_bias FLOAT[256] 796dc15ab13e
visual.layer3.18.conv2.weight FLOAT[256,256,3,3] 217efcfbc758
visual.layer3.18.conv2.weight_bias FLOAT[256] 24b7c2693aff
visual.layer3.18.conv3.weight FLOAT[1024,256,1,1] 03aecbb7da0a
visual.layer3.18.conv3.weight_bias FLOAT[1024] 5243e4ecba01
visual.layer3.19.conv1.weight FLOAT[256,1024,1,1] fb40a9ec919a
visual.layer3.19.conv1.weight_bias FLOAT[256] 9f51e9ac974a
visual.layer3.19.conv2.weight FLOAT[256,256,3,3] b536cb636767
visual.layer3.19.conv2.weight_bias FLOAT[256] 0cd34046767f
visual.layer3.19.conv3.weight FLOAT[1024,256,1,1] f6704ecf40c0
visual.layer3.19.conv3.weight_bias FLOAT[1024] 46e4d6f9abd7
visual.layer3.2.conv1.weight FLOAT[256,1024,1,1] 1f90bd499f45
visual.layer3.2.conv1.weight_bias FLOAT[256] 09a5286ec62e
visual.layer3.2.conv2.weight FLOAT[256,256,3,3] a09f97660a8f
visual.layer3.2.conv2.weight_bias FLOAT[256] 99c53ca962f6
visual.layer3.2.conv3.weight FLOAT[1024,256,1,1] a03fa7438e70
visual.layer3.2.conv3.weight_bias FLOAT[1024] 96f65dbe8421
visual.layer3.20.conv1.weight FLOAT[256,1024,1,1] fad135805bd9
visual.layer3.20.conv1.weight_bias FLOAT[256] 9f1098557e68
visual.layer3.20.conv2.weight FLOAT[256,256,3,3] 4af4e544ef06
visual.layer3.20.conv2.weight_bias FLOAT[256] 9408c9056f53
visual.layer3.20.conv3.weight FLOAT[1024,256,1,1] c09d8abd8d79
visual.layer3.20.conv3.weight_bias FLOAT[1024] 403a2f85a6e7
visual.layer3.21.conv1.weight FLOAT[256,1024,1,1] 70ed3d7dd6c3
visual.layer3.21.conv1.weight_bias FLOAT[256] 67f9a835b1a1
visual.layer3.21.conv2.weight FLOAT[256,256,3,3] 22a4046e56c8
visual.layer3.21.conv2.weight_bias FLOAT[256] f4a8ac66782e
visual.layer3.21.conv3.weight FLOAT[1024,256,1,1] 921266e3a333
visual.layer3.21.conv3.weight_bias FLOAT[1024] 0f7a77a240b7
visual.layer3.22.conv1.weight FLOAT[256,1024,1,1] 51db9c477eb4
visual.layer3.22.conv1.weight_bias FLOAT[256] e72f2f703304
visual.layer3.22.conv2.weight FLOAT[256,256,3,3] cf532b522d69
visual.layer3.22.conv2.weight_bias FLOAT[256] 9d7c283fd588
visual.layer3.22.conv3.weight FLOAT[1024,256,1,1] 60bb6dc64dc7
visual.layer3.22.conv3.weight_bias FLOAT[1024] 2e10a4198e5a
visual.layer3.3.conv1.weight FLOAT[256,1024,1,1] 2f57a087c91f
visual.layer3.3.conv1.weight_bias FLOAT[256] 1f5077976f1b
visual.layer3.3.conv2.weight FLOAT[256,256,3,3] 1df6c8bf38f2
visual.layer3.3.conv2.weight_bias FLOAT[256] 7d9ea29abb45
visual.layer3.3.conv3.weight FLOAT[1024,256,1,1] ffaf486bb7c8
visual.layer3.3.conv3.weight_bias FLOAT[1024] 93c38aedf8a0
visual.layer3.4.conv1.weight FLOAT[256,1024,1,1] 74e3d9b7ef6e
visual.layer3.4.conv1.weight_bias FLOAT[256] 7c605256b20c
visual.layer3.4.conv2.weight FLOAT[256,256,3,3] 0b54cb2cc3fa
visual.layer3.4.conv2.weight_bias FLOAT[256] 73afb3b55285
visual.layer3.4.conv3.weight FLOAT[1024,256,1,1] 21febd1efa28
visual.layer3.4.conv3.weight_bias FLOAT[1024] 0e1d6645ebeb
visual.layer3.5.conv1.weight FLOAT[256,1024,1,1] 700b495bc5d9
visual.layer3.5.conv1.weight_bias FLOAT[256] 155d6dcdabc7
visual.layer3.5.conv2.weight FLOAT[256,256,3,3] 8700f45735bd
visual.layer3.5.conv2.weight_bias FLOAT[256] f9efbf55a305
visual.layer3.5.conv3.weight FLOAT[1024,256,1,1] 6de7a56c5ac3
visual.layer3.5.conv3.weight_bias FLOAT[1024] 8d3ce2b31a6b
visual.layer3.6.conv1.weight FLOAT[256,1024,1,1] 28b798e5d978
visual.layer3.6.conv1.weight_bias FLOAT[256] fef76cc29c73
visual.layer3.6.conv2.weight FLOAT[256,256,3,3] abeebafcf6e1
visual.layer3.6.conv2.weight_bias FLOAT[256] 9ff63eb59206
visual.layer3.6.conv3.weight FLOAT[1024,256,1,1] 9d43ccaac3af
visual.layer3.6.conv3.weight_bias FLOAT[1024] b4037171de0e
visual.layer3.7.conv1.weight FLOAT[256,1024,1,1] 1ab932ba181a
visual.layer3.7.conv1.weight_bias FLOAT[256] ca2b5f101456
visual.layer3.7.conv2.weight FLOAT[256,256,3,3] ec7c7a4462b9
visual.layer3.7.conv2.weight_bias FLOAT[256] 6602c75c4b54
visual.layer3.7.conv3.weight FLOAT[1024,256,1,1] a2a816ebebd4
visual.layer3.7.conv3.weight_bias FLOAT[1024] 42ead0570f3f
visual.layer3.8.conv1.weight FLOAT[256,1024,1,1] 65a1e6e2a9ac
visual.layer3.8.conv1.weight_bias FLOAT[256] f3ce23615022
visual.layer3.8.conv2.weight FLOAT[256,256,3,3] daa5a1a4aec1
visual.layer3.8.conv2.weight_bias FLOAT[256] 6dbb5cf51f67
visual.layer3.8.conv3.weight FLOAT[1024,256,1,1] d68a418f835c
visual.layer3.8.conv3.weight_bias FLOAT[1024] 4eca3e4a9949
visual.layer3.9.conv1.weight FLOAT[256,1024,1,1] ab40ed895315
visual.layer3.9.conv1.weight_bias FLOAT[256] 15eaf0e90b7a
visual.layer3.9.conv2.weight FLOAT[256,256,3,3] f2766f847a2d
visual.layer3.9.conv2.weight_bias FLOAT[256] a501c342e17d
visual.layer3.9.conv3.weight FLOAT[1024,256,1,1] 682f558e1a40
visual.layer3.9.conv3.weight_bias FLOAT[1024] e12217d407a6
visual.layer4.0.conv1.weight FLOAT[512,1024,1,1] 9390c2f115e6
visual.layer4.0.conv1.weight_bias FLOAT[512] a96a15f56ea2
visual.layer4.0.conv2.weight FLOAT[512,512,3,3] 6f46f3db8786
visual.layer4.0.conv2.weight_bias FLOAT[512] bbbb373cd8fc
visual.layer4.0.conv3.weight FLOAT[2048,512,1,1] 480da506450a
visual.layer4.0.conv3.weight_bias FLOAT[2048] 45911f9d4ede
visual.layer4.0.downsample.0.weight FLOAT[2048,1024,1,1] 1dfa4209cd69
visual.layer4.0.downsample.0.weight_bias FLOAT[2048] ec264c60e9a9
visual.layer4.1.conv1.weight FLOAT[512,2048,1,1] eb2f69345705
visual.layer4.1.conv1.weight_bias FLOAT[512] 5de1b4440e4d
visual.layer4.1.conv2.weight FLOAT[512,512,3,3] 3566c6b83f7e
visual.layer4.1.conv2.weight_bias FLOAT[512] 1e250c828acc
visual.layer4.1.conv3.weight FLOAT[2048,512,1,1] 6352b196366c
visual.layer4.1.conv3.weight_bias FLOAT[2048] 15c292abf864
visual.layer4.2.conv1.weight FLOAT[512,2048,1,1] 3a5c04bc8b5d
visual.layer4.2.conv1.weight_bias FLOAT[512] 0991b62339dc
visual.layer4.2.conv2.weight FLOAT[512,512,3,3] 3c9ac30e7976
visual.layer4.2.conv2.weight_bias FLOAT[512] a4fd49beab30
visual.layer4.2.conv3.weight FLOAT[2048,512,1,1] 80801ef70ead
visual.layer4.2.conv3.weight_bias FLOAT[2048] 27f0f2f0a15a
