Skip to content

语义分割炼丹技巧：mean std #16

Open
Open
语义分割炼丹技巧：mean std#16
@gemfield

Description

@gemfield
gemfield
opened on May 29, 2021

不同mean std的pk

数据集

  • 训练集：clothes std 2.1
  • 验证集：LIP986

炼丹参数

  • config.input_w = 384
  • config.input_h = 384
  • config.core.cls_num = 4
  • config.aug.ImageWithMasksRandomRotateAug.max_angle = 45
  • config.core.batch_size = 16
  • config.core.model_path = "/opt/public/pretrain/ESPNetv2/imagenet/espnetv2_s_2.0.pth"
  • config.core.optimizer = torch.optim.Adam(config.core.net.parameters(), 3e-4, (0.9, 0.999), eps=1e-08, weight_decay=5e-4)
  • lambda_lr = lambda epoch: round ((1 - epoch/config.core.epoch_num) ** 0.9, 8)
  • config.core.scheduler = optim.lr_scheduler.LambdaLR(config.core.optimizer, lr_lambda=lambda_lr)
  • config.core.criterion = torch.nn.CrossEntropyLoss(weight)
  • AugFactory('SpeckleAug@0.1 => GaussianAug@0.1 => HorlineAug@0.1 => VerlineAug@0.1 => LRmotionAug@0.1 =>UDmotionAug@0.1 => NoisyAug@0.1 => DarkAug@0.1 => ColorJitterAug@0.15 => BrightnessJitterAug@0.15 =>ContrastJitterAug@0.15 => ImageWithMasksRandomRotateAug@0.6 => ImageWithMasksNormalizeAug =>ImageWithMasksCenterCropAug => ImageWithMasksScaleAug => ImageWithMasksHFlipAug@0.5 =>ImageWithMasksToTensorAug', deepvac_config)

TRAIN

config.core.mean = config.data['mean'] && config.core.std = config.data['std']

Epoch No.: 0    TRAIN Loss = 0.8305      TRAIN mIOU = 0.6219
Epoch No.: 1    TRAIN Loss = 0.6207      TRAIN mIOU = 0.6758
Epoch No.: 2    TRAIN Loss = 0.5139      TRAIN mIOU = 0.7286
Epoch No.: 3    TRAIN Loss = 0.4646      TRAIN mIOU = 0.7510
Epoch No.: 4    TRAIN Loss = 0.4634      TRAIN mIOU = 0.7521
Epoch No.: 5    TRAIN Loss = 0.4377      TRAIN mIOU = 0.7588
Epoch No.: 6    TRAIN Loss = 0.4065      TRAIN mIOU = 0.7717
Epoch No.: 7    TRAIN Loss = 0.4041      TRAIN mIOU = 0.7728
Epoch No.: 8    TRAIN Loss = 0.4023      TRAIN mIOU = 0.7758
Epoch No.: 9    TRAIN Loss = 0.3936      TRAIN mIOU = 0.7782
Epoch No.: 10   TRAIN Loss = 0.3864      TRAIN mIOU = 0.7768
Epoch No.: 11   TRAIN Loss = 0.3854      TRAIN mIOU = 0.7785
Epoch No.: 12   TRAIN Loss = 0.3883      TRAIN mIOU = 0.7790
Epoch No.: 13   TRAIN Loss = 0.3796      TRAIN mIOU = 0.7800
Epoch No.: 14   TRAIN Loss = 0.3667      TRAIN mIOU = 0.7873
Epoch No.: 15   TRAIN Loss = 0.3524      TRAIN mIOU = 0.7932
Epoch No.: 16   TRAIN Loss = 0.3539      TRAIN mIOU = 0.7935
Epoch No.: 17   TRAIN Loss = 0.3466      TRAIN mIOU = 0.7947
Epoch No.: 18   TRAIN Loss = 0.3523      TRAIN mIOU = 0.7939
Epoch No.: 19   TRAIN Loss = 0.3518      TRAIN mIOU = 0.7919
Epoch No.: 20   TRAIN Loss = 0.3384      TRAIN mIOU = 0.8006
Epoch No.: 21   TRAIN Loss = 0.3494      TRAIN mIOU = 0.7959
Epoch No.: 22   TRAIN Loss = 0.3488      TRAIN mIOU = 0.7953
Epoch No.: 23   TRAIN Loss = 0.3383      TRAIN mIOU = 0.8008
Epoch No.: 24   TRAIN Loss = 0.3317      TRAIN mIOU = 0.8019
Epoch No.: 25   TRAIN Loss = 0.3454      TRAIN mIOU = 0.7980
Epoch No.: 26   TRAIN Loss = 0.3399      TRAIN mIOU = 0.8008
Epoch No.: 27   TRAIN Loss = 0.3283      TRAIN mIOU = 0.8043
Epoch No.: 28   TRAIN Loss = 0.3313      TRAIN mIOU = 0.8052
Epoch No.: 29   TRAIN Loss = 0.3149      TRAIN mIOU = 0.8144
Epoch No.: 30   TRAIN Loss = 0.3252      TRAIN mIOU = 0.8099
Epoch No.: 31   TRAIN Loss = 0.3250      TRAIN mIOU = 0.8086
Epoch No.: 32   TRAIN Loss = 0.3217      TRAIN mIOU = 0.8123
Epoch No.: 33   TRAIN Loss = 0.3145      TRAIN mIOU = 0.8153
Epoch No.: 34   TRAIN Loss = 0.3095      TRAIN mIOU = 0.8145
Epoch No.: 35   TRAIN Loss = 0.3217      TRAIN mIOU = 0.8118
Epoch No.: 36   TRAIN Loss = 0.3023      TRAIN mIOU = 0.8188
Epoch No.: 37   TRAIN Loss = 0.3479      TRAIN mIOU = 0.8003
Epoch No.: 38   TRAIN Loss = 0.3139      TRAIN mIOU = 0.8162
Epoch No.: 39   TRAIN Loss = 0.3228      TRAIN mIOU = 0.8084
Epoch No.: 40   TRAIN Loss = 0.3105      TRAIN mIOU = 0.8169
Epoch No.: 41   TRAIN Loss = 0.3098      TRAIN mIOU = 0.8176
Epoch No.: 42   TRAIN Loss = 0.3020      TRAIN mIOU = 0.8221
Epoch No.: 43   TRAIN Loss = 0.3232      TRAIN mIOU = 0.8106
Epoch No.: 44   TRAIN Loss = 0.3113      TRAIN mIOU = 0.8178
Epoch No.: 45   TRAIN Loss = 0.3171      TRAIN mIOU = 0.8139
Epoch No.: 46   TRAIN Loss = 0.3011      TRAIN mIOU = 0.8238
Epoch No.: 47   TRAIN Loss = 0.3028      TRAIN mIOU = 0.8196
Epoch No.: 48   TRAIN Loss = 0.2992      TRAIN mIOU = 0.8217
Epoch No.: 49   TRAIN Loss = 0.3661      TRAIN mIOU = 0.7944
Epoch No.: 50   TRAIN Loss = 0.2963      TRAIN mIOU = 0.8231
Epoch No.: 51   TRAIN Loss = 0.2979      TRAIN mIOU = 0.8209
Epoch No.: 52   TRAIN Loss = 0.2880      TRAIN mIOU = 0.8268
Epoch No.: 53   TRAIN Loss = 0.3200      TRAIN mIOU = 0.8141
Epoch No.: 54   TRAIN Loss = 0.2867      TRAIN mIOU = 0.8280
Epoch No.: 55   TRAIN Loss = 0.2851      TRAIN mIOU = 0.8299
Epoch No.: 56   TRAIN Loss = 0.2963      TRAIN mIOU = 0.8249
Epoch No.: 57   TRAIN Loss = 0.2867      TRAIN mIOU = 0.8282
Epoch No.: 58   TRAIN Loss = 0.2872      TRAIN mIOU = 0.8284
Epoch No.: 59   TRAIN Loss = 0.2830      TRAIN mIOU = 0.8283
Epoch No.: 60   TRAIN Loss = 0.2824      TRAIN mIOU = 0.8320
Epoch No.: 61   TRAIN Loss = 0.2788      TRAIN mIOU = 0.8347
Epoch No.: 62   TRAIN Loss = 0.3024      TRAIN mIOU = 0.8256
Epoch No.: 63   TRAIN Loss = 0.2851      TRAIN mIOU = 0.8298
Epoch No.: 64   TRAIN Loss = 0.2906      TRAIN mIOU = 0.8269
Epoch No.: 65   TRAIN Loss = 0.2852      TRAIN mIOU = 0.8279
Epoch No.: 66   TRAIN Loss = 0.2748      TRAIN mIOU = 0.8362
Epoch No.: 67   TRAIN Loss = 0.2889      TRAIN mIOU = 0.8313
Epoch No.: 68   TRAIN Loss = 0.2904      TRAIN mIOU = 0.8271
Epoch No.: 69   TRAIN Loss = 0.2932      TRAIN mIOU = 0.8270
Epoch No.: 70   TRAIN Loss = 0.2758      TRAIN mIOU = 0.8387
Epoch No.: 71   TRAIN Loss = 0.2819      TRAIN mIOU = 0.8336

config.core.mean = np.array([0.406, 0.456, 0.485]) * 255 && config.core.std = np.array([0.224, 0.225, 0.229]) * 255

Epoch No.: 0    TRAIN Loss = 0.8366      TRAIN mIOU = 0.6301
Epoch No.: 1    TRAIN Loss = 0.5948      TRAIN mIOU = 0.6986
Epoch No.: 2    TRAIN Loss = 0.5243      TRAIN mIOU = 0.7221
Epoch No.: 3    TRAIN Loss = 0.4927      TRAIN mIOU = 0.7373
Epoch No.: 4    TRAIN Loss = 0.4506      TRAIN mIOU = 0.7494
Epoch No.: 5    TRAIN Loss = 0.4363      TRAIN mIOU = 0.7573
Epoch No.: 6    TRAIN Loss = 0.4201      TRAIN mIOU = 0.7674
Epoch No.: 7    TRAIN Loss = 0.4050      TRAIN mIOU = 0.7667
Epoch No.: 8    TRAIN Loss = 0.3794      TRAIN mIOU = 0.7790
Epoch No.: 9    TRAIN Loss = 0.4004      TRAIN mIOU = 0.7725
Epoch No.: 10   TRAIN Loss = 0.3861      TRAIN mIOU = 0.7744
Epoch No.: 11   TRAIN Loss = 0.3822      TRAIN mIOU = 0.7762
Epoch No.: 12   TRAIN Loss = 0.3700      TRAIN mIOU = 0.7805
Epoch No.: 13   TRAIN Loss = 0.3603      TRAIN mIOU = 0.7861
Epoch No.: 14   TRAIN Loss = 0.3559      TRAIN mIOU = 0.7890
Epoch No.: 15   TRAIN Loss = 0.3692      TRAIN mIOU = 0.7817
Epoch No.: 16   TRAIN Loss = 0.3653      TRAIN mIOU = 0.7813
Epoch No.: 17   TRAIN Loss = 0.3682      TRAIN mIOU = 0.7851
Epoch No.: 18   TRAIN Loss = 0.3405      TRAIN mIOU = 0.7950
Epoch No.: 19   TRAIN Loss = 0.3447      TRAIN mIOU = 0.7958
Epoch No.: 20   TRAIN Loss = 0.3428      TRAIN mIOU = 0.7926
Epoch No.: 21   TRAIN Loss = 0.3439      TRAIN mIOU = 0.7919
Epoch No.: 22   TRAIN Loss = 0.3546      TRAIN mIOU = 0.7839
Epoch No.: 23   TRAIN Loss = 0.3414      TRAIN mIOU = 0.7954
Epoch No.: 24   TRAIN Loss = 0.3426      TRAIN mIOU = 0.7926
Epoch No.: 25   TRAIN Loss = 0.3539      TRAIN mIOU = 0.7960
Epoch No.: 26   TRAIN Loss = 0.3400      TRAIN mIOU = 0.7940
Epoch No.: 27   TRAIN Loss = 0.3334      TRAIN mIOU = 0.7996
Epoch No.: 28   TRAIN Loss = 0.3527      TRAIN mIOU = 0.7933
Epoch No.: 29   TRAIN Loss = 0.3269      TRAIN mIOU = 0.8054
Epoch No.: 30   TRAIN Loss = 0.3221      TRAIN mIOU = 0.8070
Epoch No.: 31   TRAIN Loss = 0.3345      TRAIN mIOU = 0.7999
Epoch No.: 32   TRAIN Loss = 0.3271      TRAIN mIOU = 0.8024
Epoch No.: 33   TRAIN Loss = 0.3218      TRAIN mIOU = 0.8073
Epoch No.: 34   TRAIN Loss = 0.3178      TRAIN mIOU = 0.8121
Epoch No.: 35   TRAIN Loss = 0.3109      TRAIN mIOU = 0.8157
Epoch No.: 36   TRAIN Loss = 0.3431      TRAIN mIOU = 0.8027
Epoch No.: 37   TRAIN Loss = 0.3178      TRAIN mIOU = 0.8111
Epoch No.: 38   TRAIN Loss = 0.3020      TRAIN mIOU = 0.8208
Epoch No.: 39   TRAIN Loss = 0.3194      TRAIN mIOU = 0.8142
Epoch No.: 40   TRAIN Loss = 0.3122      TRAIN mIOU = 0.8157
Epoch No.: 41   TRAIN Loss = 0.3096      TRAIN mIOU = 0.8186
Epoch No.: 42   TRAIN Loss = 0.2946      TRAIN mIOU = 0.8237
Epoch No.: 43   TRAIN Loss = 0.3123      TRAIN mIOU = 0.8167
Epoch No.: 44   TRAIN Loss = 0.2982      TRAIN mIOU = 0.8211
Epoch No.: 45   TRAIN Loss = 0.3216      TRAIN mIOU = 0.8107
Epoch No.: 46   TRAIN Loss = 0.3048      TRAIN mIOU = 0.8212
Epoch No.: 47   TRAIN Loss = 0.2985      TRAIN mIOU = 0.8211
Epoch No.: 48   TRAIN Loss = 0.3184      TRAIN mIOU = 0.8148
Epoch No.: 49   TRAIN Loss = 0.2981      TRAIN mIOU = 0.8221
Epoch No.: 50   TRAIN Loss = 0.2956      TRAIN mIOU = 0.8196
Epoch No.: 51   TRAIN Loss = 0.3136      TRAIN mIOU = 0.8112
Epoch No.: 52   TRAIN Loss = 0.3168      TRAIN mIOU = 0.8119
Epoch No.: 53   TRAIN Loss = 0.2957      TRAIN mIOU = 0.8197
Epoch No.: 54   TRAIN Loss = 0.2979      TRAIN mIOU = 0.8232
Epoch No.: 55   TRAIN Loss = 0.3038      TRAIN mIOU = 0.8206
Epoch No.: 56   TRAIN Loss = 0.2890      TRAIN mIOU = 0.8256
Epoch No.: 57   TRAIN Loss = 0.2870      TRAIN mIOU = 0.8271
Epoch No.: 58   TRAIN Loss = 0.2857      TRAIN mIOU = 0.8259
Epoch No.: 59   TRAIN Loss = 0.2945      TRAIN mIOU = 0.8229
Epoch No.: 60   TRAIN Loss = 0.2899      TRAIN mIOU = 0.8265
Epoch No.: 61   TRAIN Loss = 0.2808      TRAIN mIOU = 0.8332
Epoch No.: 62   TRAIN Loss = 0.2891      TRAIN mIOU = 0.8296
Epoch No.: 63   TRAIN Loss = 0.2807      TRAIN mIOU = 0.8308
Epoch No.: 64   TRAIN Loss = 0.2798      TRAIN mIOU = 0.8298
Epoch No.: 65   TRAIN Loss = 0.2956      TRAIN mIOU = 0.8232
Epoch No.: 66   TRAIN Loss = 0.2755      TRAIN mIOU = 0.8371
Epoch No.: 67   TRAIN Loss = 0.2910      TRAIN mIOU = 0.8253
Epoch No.: 68   TRAIN Loss = 0.2880      TRAIN mIOU = 0.8264
Epoch No.: 69   TRAIN Loss = 0.2793      TRAIN mIOU = 0.8322
Epoch No.: 70   TRAIN Loss = 0.2931      TRAIN mIOU = 0.8288
Epoch No.: 71   TRAIN Loss = 0.3069      TRAIN mIOU = 0.8180

VAL

config.core.mean = config.data['mean'] && config.core.std = config.data['std']

Epoch No.: 0    VAL Loss = 0.5173        VAL mIOU = 0.6432
Epoch No.: 1    VAL Loss = 0.7352        VAL mIOU = 0.6698
Epoch No.: 2    VAL Loss = 0.9506        VAL mIOU = 0.6749
Epoch No.: 3    VAL Loss = 0.4759        VAL mIOU = 0.7032
Epoch No.: 4    VAL Loss = 0.4187        VAL mIOU = 0.6993
Epoch No.: 5    VAL Loss = 0.5501        VAL mIOU = 0.6915
Epoch No.: 6    VAL Loss = 0.5055        VAL mIOU = 0.6912
Epoch No.: 7    VAL Loss = 0.3540        VAL mIOU = 0.7052
Epoch No.: 8    VAL Loss = 0.3820        VAL mIOU = 0.7039
Epoch No.: 9    VAL Loss = 0.4253        VAL mIOU = 0.6960
Epoch No.: 10   VAL Loss = 0.3428        VAL mIOU = 0.7053
Epoch No.: 11   VAL Loss = 0.3238        VAL mIOU = 0.7105
Epoch No.: 12   VAL Loss = 0.4548        VAL mIOU = 0.7008
Epoch No.: 13   VAL Loss = 0.2605        VAL mIOU = 0.7088
Epoch No.: 14   VAL Loss = 0.2577        VAL mIOU = 0.7024
Epoch No.: 15   VAL Loss = 0.4431        VAL mIOU = 0.6901
Epoch No.: 16   VAL Loss = 0.6530        VAL mIOU = 0.7028
Epoch No.: 17   VAL Loss = 0.3773        VAL mIOU = 0.7138
Epoch No.: 18   VAL Loss = 0.3961        VAL mIOU = 0.7098
Epoch No.: 19   VAL Loss = 0.5007        VAL mIOU = 0.7003
Epoch No.: 20   VAL Loss = 0.4277        VAL mIOU = 0.7065
Epoch No.: 21   VAL Loss = 0.3111        VAL mIOU = 0.7018
Epoch No.: 22   VAL Loss = 0.4636        VAL mIOU = 0.7042
Epoch No.: 23   VAL Loss = 0.3106        VAL mIOU = 0.7164
Epoch No.: 24   VAL Loss = 0.4415        VAL mIOU = 0.7028
Epoch No.: 25   VAL Loss = 0.4288        VAL mIOU = 0.7096
Epoch No.: 26   VAL Loss = 0.4786        VAL mIOU = 0.7097
Epoch No.: 27   VAL Loss = 0.2658        VAL mIOU = 0.7024
Epoch No.: 28   VAL Loss = 0.3159        VAL mIOU = 0.7124
Epoch No.: 29   VAL Loss = 0.2794        VAL mIOU = 0.6917
Epoch No.: 30   VAL Loss = 0.4634        VAL mIOU = 0.7142
Epoch No.: 31   VAL Loss = 0.3339        VAL mIOU = 0.7104
Epoch No.: 32   VAL Loss = 0.3904        VAL mIOU = 0.7146
Epoch No.: 33   VAL Loss = 0.1813        VAL mIOU = 0.7036
Epoch No.: 34   VAL Loss = 0.2867        VAL mIOU = 0.7112
Epoch No.: 35   VAL Loss = 0.3358        VAL mIOU = 0.7012
Epoch No.: 36   VAL Loss = 0.3156        VAL mIOU = 0.7102
Epoch No.: 37   VAL Loss = 0.6730        VAL mIOU = 0.7118
Epoch No.: 38   VAL Loss = 0.3842        VAL mIOU = 0.7124
Epoch No.: 39   VAL Loss = 0.2802        VAL mIOU = 0.7041
Epoch No.: 40   VAL Loss = 0.3072        VAL mIOU = 0.7008
Epoch No.: 41   VAL Loss = 0.2515        VAL mIOU = 0.7058
Epoch No.: 42   VAL Loss = 0.3218        VAL mIOU = 0.7141
Epoch No.: 43   VAL Loss = 0.3874        VAL mIOU = 0.7141
Epoch No.: 44   VAL Loss = 0.3264        VAL mIOU = 0.7046
Epoch No.: 45   VAL Loss = 0.2222        VAL mIOU = 0.7133
Epoch No.: 46   VAL Loss = 0.2573        VAL mIOU = 0.7175
Epoch No.: 47   VAL Loss = 0.2777        VAL mIOU = 0.7070
Epoch No.: 48   VAL Loss = 0.2893        VAL mIOU = 0.7089
Epoch No.: 49   VAL Loss = 0.3707        VAL mIOU = 0.7077
Epoch No.: 50   VAL Loss = 0.2880        VAL mIOU = 0.7086
Epoch No.: 51   VAL Loss = 0.2251        VAL mIOU = 0.7092
Epoch No.: 52   VAL Loss = 0.2015        VAL mIOU = 0.7095
Epoch No.: 53   VAL Loss = 0.3612        VAL mIOU = 0.7125
Epoch No.: 54   VAL Loss = 0.4589        VAL mIOU = 0.7110
Epoch No.: 55   VAL Loss = 0.1923        VAL mIOU = 0.7142
Epoch No.: 56   VAL Loss = 0.3266        VAL mIOU = 0.7219
Epoch No.: 57   VAL Loss = 0.4174        VAL mIOU = 0.7158
Epoch No.: 58   VAL Loss = 0.2626        VAL mIOU = 0.7116
Epoch No.: 59   VAL Loss = 0.2723        VAL mIOU = 0.7077
Epoch No.: 60   VAL Loss = 0.2446        VAL mIOU = 0.7148
Epoch No.: 61   VAL Loss = 0.2165        VAL mIOU = 0.7193
Epoch No.: 62   VAL Loss = 0.2425        VAL mIOU = 0.7183
Epoch No.: 63   VAL Loss = 0.1794        VAL mIOU = 0.7070
Epoch No.: 64   VAL Loss = 0.2505        VAL mIOU = 0.7088
Epoch No.: 65   VAL Loss = 0.2996        VAL mIOU = 0.7113
Epoch No.: 66   VAL Loss = 0.2840        VAL mIOU = 0.7256
Epoch No.: 67   VAL Loss = 0.1732        VAL mIOU = 0.7209
Epoch No.: 68   VAL Loss = 0.3359        VAL mIOU = 0.7211
Epoch No.: 69   VAL Loss = 0.2815        VAL mIOU = 0.7205
Epoch No.: 70   VAL Loss = 0.3863        VAL mIOU = 0.7236
Epoch No.: 71   VAL Loss = 0.4573        VAL mIOU = 0.7125

config.core.mean = np.array([0.406, 0.456, 0.485]) * 255 && config.core.std = np.array([0.224, 0.225, 0.229]) * 255

Epoch No.: 0    VAL Loss = 1.0027        VAL mIOU = 0.6609
Epoch No.: 1    VAL Loss = 0.4098        VAL mIOU = 0.6891
Epoch No.: 2    VAL Loss = 0.4527        VAL mIOU = 0.6816
Epoch No.: 3    VAL Loss = 0.4802        VAL mIOU = 0.6926
Epoch No.: 4    VAL Loss = 0.4131        VAL mIOU = 0.7090
Epoch No.: 5    VAL Loss = 0.3669        VAL mIOU = 0.6910
Epoch No.: 6    VAL Loss = 0.3852        VAL mIOU = 0.7039
Epoch No.: 7    VAL Loss = 0.4639        VAL mIOU = 0.7010
Epoch No.: 8    VAL Loss = 0.2421        VAL mIOU = 0.7056
Epoch No.: 9    VAL Loss = 0.3121        VAL mIOU = 0.7064
Epoch No.: 10   VAL Loss = 0.4149        VAL mIOU = 0.6891
Epoch No.: 11   VAL Loss = 0.4101        VAL mIOU = 0.6920
Epoch No.: 12   VAL Loss = 0.5366        VAL mIOU = 0.7054
Epoch No.: 13   VAL Loss = 0.2483        VAL mIOU = 0.7016
Epoch No.: 14   VAL Loss = 0.3290        VAL mIOU = 0.7000
Epoch No.: 15   VAL Loss = 0.4951        VAL mIOU = 0.7046
Epoch No.: 16   VAL Loss = 0.2388        VAL mIOU = 0.7102
Epoch No.: 17   VAL Loss = 0.4184        VAL mIOU = 0.7130
Epoch No.: 18   VAL Loss = 0.2305        VAL mIOU = 0.6993
Epoch No.: 19   VAL Loss = 0.2024        VAL mIOU = 0.7130
Epoch No.: 20   VAL Loss = 0.2712        VAL mIOU = 0.7119
Epoch No.: 21   VAL Loss = 0.3688        VAL mIOU = 0.7067
Epoch No.: 22   VAL Loss = 0.2932        VAL mIOU = 0.7037
Epoch No.: 23   VAL Loss = 0.3351        VAL mIOU = 0.7116
Epoch No.: 24   VAL Loss = 0.3120        VAL mIOU = 0.6907
Epoch No.: 25   VAL Loss = 0.3701        VAL mIOU = 0.7075
Epoch No.: 26   VAL Loss = 0.2840        VAL mIOU = 0.7055
Epoch No.: 27   VAL Loss = 0.1913        VAL mIOU = 0.7009
Epoch No.: 28   VAL Loss = 0.3571        VAL mIOU = 0.7058
Epoch No.: 29   VAL Loss = 0.2842        VAL mIOU = 0.6992
Epoch No.: 30   VAL Loss = 0.2221        VAL mIOU = 0.6947
Epoch No.: 31   VAL Loss = 0.4760        VAL mIOU = 0.7040
Epoch No.: 32   VAL Loss = 0.3142        VAL mIOU = 0.7053
Epoch No.: 33   VAL Loss = 0.3420        VAL mIOU = 0.7011
Epoch No.: 34   VAL Loss = 0.2332        VAL mIOU = 0.7141
Epoch No.: 35   VAL Loss = 0.3380        VAL mIOU = 0.7150
Epoch No.: 36   VAL Loss = 0.3132        VAL mIOU = 0.6955
Epoch No.: 37   VAL Loss = 0.2566        VAL mIOU = 0.7078
Epoch No.: 38   VAL Loss = 0.3091        VAL mIOU = 0.7179
Epoch No.: 39   VAL Loss = 0.2164        VAL mIOU = 0.7153
Epoch No.: 40   VAL Loss = 0.2183        VAL mIOU = 0.7151
Epoch No.: 41   VAL Loss = 0.3004        VAL mIOU = 0.7105
Epoch No.: 42   VAL Loss = 0.2718        VAL mIOU = 0.7142
Epoch No.: 43   VAL Loss = 0.2290        VAL mIOU = 0.7099
Epoch No.: 44   VAL Loss = 0.3576        VAL mIOU = 0.7211
Epoch No.: 45   VAL Loss = 0.4168        VAL mIOU = 0.7007
Epoch No.: 46   VAL Loss = 0.3234        VAL mIOU = 0.7180
Epoch No.: 47   VAL Loss = 0.4667        VAL mIOU = 0.6993
Epoch No.: 48   VAL Loss = 0.4768        VAL mIOU = 0.6966
Epoch No.: 49   VAL Loss = 0.2444        VAL mIOU = 0.7161
Epoch No.: 50   VAL Loss = 0.2509        VAL mIOU = 0.7061
Epoch No.: 51   VAL Loss = 0.3836        VAL mIOU = 0.7025
Epoch No.: 52   VAL Loss = 0.2030        VAL mIOU = 0.7112
Epoch No.: 53   VAL Loss = 0.3260        VAL mIOU = 0.7081
Epoch No.: 54   VAL Loss = 0.5848        VAL mIOU = 0.7049
Epoch No.: 55   VAL Loss = 0.2594        VAL mIOU = 0.7151
Epoch No.: 56   VAL Loss = 0.2592        VAL mIOU = 0.7108
Epoch No.: 57   VAL Loss = 0.2903        VAL mIOU = 0.7142
Epoch No.: 58   VAL Loss = 0.2276        VAL mIOU = 0.7182
Epoch No.: 59   VAL Loss = 0.2374        VAL mIOU = 0.7116
Epoch No.: 60   VAL Loss = 0.2516        VAL mIOU = 0.7113
Epoch No.: 61   VAL Loss = 0.4056        VAL mIOU = 0.7084
Epoch No.: 62   VAL Loss = 0.2798        VAL mIOU = 0.7176
Epoch No.: 63   VAL Loss = 0.4840        VAL mIOU = 0.7171
Epoch No.: 64   VAL Loss = 0.2694        VAL mIOU = 0.7125
Epoch No.: 65   VAL Loss = 0.2726        VAL mIOU = 0.7120
Epoch No.: 66   VAL Loss = 0.2584        VAL mIOU = 0.7047
Epoch No.: 67   VAL Loss = 0.2813        VAL mIOU = 0.7164
Epoch No.: 68   VAL Loss = 0.2304        VAL mIOU = 0.7129
Epoch No.: 69   VAL Loss = 0.2463        VAL mIOU = 0.7255
Epoch No.: 70   VAL Loss = 0.2372        VAL mIOU = 0.7171
Epoch No.: 71   VAL Loss = 0.3434        VAL mIOU = 0.7097

Metadata

Metadata

Assignees

No one assigned

    Labels

    No labels
    No labels

    Type

    No type

    Fields

    No fields configured for issues without a type.

    Projects

    No projects

    Milestone

    No milestone

    Relationships

    None yet

    Development

    No branches or pull requests

    Issue actions