After training the self-defined dataset(using coco data format) using yolov4 model
the training log specify that (mAP@0.50) is above 90%, however when using pretrained model to do inference,
it gets the following info:
conv 256 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.089 BFLOPs
78 Couldn't find activation function mish, going with ReLU
conv 256 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.797 BFLOPs
79 res 76 26 x 26 x 256 -> 26 x 26 x 256
80 Couldn't find activation function mish, going with ReLU
conv 256 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.089 BFLOPs
81 Couldn't find activation function mish, going with ReLU
conv 256 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.797 BFLOPs
82 res 79 26 x 26 x 256 -> 26 x 26 x 256
83 Couldn't find activation function mish, going with ReLU
conv 256 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.089 BFLOPs
84 route 83 56
85 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 512 0.354 BFLOPs
86 Couldn't find activation function mish, going with ReLU
conv 1024 3 x 3 / 2 26 x 26 x 512 -> 13 x 13 x1024 1.595 BFLOPs
87 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
88 route 86
89 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
90 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs
91 Couldn't find activation function mish, going with ReLU
conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs
92 res 89 13 x 13 x 512 -> 13 x 13 x 512
93 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs
94 Couldn't find activation function mish, going with ReLU
conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs
95 res 92 13 x 13 x 512 -> 13 x 13 x 512
96 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs
97 Couldn't find activation function mish, going with ReLU
conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs
98 res 95 13 x 13 x 512 -> 13 x 13 x 512
99 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs
100 Couldn't find activation function mish, going with ReLU
conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs
101 res 98 13 x 13 x 512 -> 13 x 13 x 512
102 Couldn't find activation function mish, going with ReLU
conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs
103 route 102 87
104 Couldn't find activation function mish, going with ReLU
conv 1024 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x1024 0.354 BFLOPs
105 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
106 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs
107 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
108 max 5 x 5 / 1 13 x 13 x 512 -> 13 x 13 x 512
109 route 107
110 max 9 x 9 / 1 13 x 13 x 512 -> 13 x 13 x 512
111 route 107
112 max 13 x 13 / 1 13 x 13 x 512 -> 13 x 13 x 512
113 route 112 110 108 107
114 conv 512 1 x 1 / 1 13 x 13 x2048 -> 13 x 13 x 512 0.354 BFLOPs
115 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs
116 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
117 conv 256 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 256 0.044 BFLOPs
118 upsample 2x 13 x 13 x 256 -> 26 x 26 x 256
119 route 85
120 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
121 route 120 118
122 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
123 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs
124 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
125 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs
126 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
127 conv 128 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 128 0.044 BFLOPs
128 upsample 2x 26 x 26 x 128 -> 52 x 52 x 128
129 route 54
130 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs
131 route 130 128
132 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs
133 conv 256 3 x 3 / 1 52 x 52 x 128 -> 52 x 52 x 256 1.595 BFLOPs
134 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs
135 conv 256 3 x 3 / 1 52 x 52 x 128 -> 52 x 52 x 256 1.595 BFLOPs
136 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs
137 conv 256 3 x 3 / 1 52 x 52 x 128 -> 52 x 52 x 256 1.595 BFLOPs
138 conv 75 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 75 0.104 BFLOPs
139 yolo
Unused field: 'scale_x_y = 1.2'
Unused field: 'iou_thresh = 0.213'
Unused field: 'cls_normalizer = 1.0'
Unused field: 'iou_normalizer = 0.07'
Unused field: 'iou_loss = ciou'
Unused field: 'nms_kind = greedynms'
Unused field: 'beta_nms = 0.6'
Unused field: 'max_delta = 5'
140 route 136
141 conv 256 3 x 3 / 2 52 x 52 x 128 -> 26 x 26 x 256 0.399 BFLOPs
142 route 141 126
143 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
144 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs
145 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
146 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs
147 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs
148 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs
149 conv 75 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 75 0.052 BFLOPs
150 yolo
Unused field: 'scale_x_y = 1.1'
Unused field: 'iou_thresh = 0.213'
Unused field: 'cls_normalizer = 1.0'
Unused field: 'iou_normalizer = 0.07'
Unused field: 'iou_loss = ciou'
Unused field: 'nms_kind = greedynms'
Unused field: 'beta_nms = 0.6'
Unused field: 'max_delta = 5'
151 route 147
152 conv 512 3 x 3 / 2 26 x 26 x 256 -> 13 x 13 x 512 0.399 BFLOPs
153 route 152 116
154 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
155 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs
156 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
157 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs
158 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs
159 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs
160 conv 75 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 75 0.026 BFLOPs
161 yolo
Unused field: 'scale_x_y = 1.05'
Unused field: 'iou_thresh = 0.213'
Unused field: 'cls_normalizer = 1.0'
Unused field: 'iou_normalizer = 0.07'
Unused field: 'iou_loss = ciou'
Unused field: 'nms_kind = greedynms'
Unused field: 'beta_nms = 0.6'
Unused field: 'max_delta = 5'
After training the self-defined dataset(using coco data format) using yolov4 model the training log specify that (mAP@0.50) is above 90%, however when using pretrained model to do inference, it gets the following info:
conv 256 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.089 BFLOPs 78 Couldn't find activation function mish, going with ReLU conv 256 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.797 BFLOPs 79 res 76 26 x 26 x 256 -> 26 x 26 x 256 80 Couldn't find activation function mish, going with ReLU conv 256 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.089 BFLOPs 81 Couldn't find activation function mish, going with ReLU conv 256 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.797 BFLOPs 82 res 79 26 x 26 x 256 -> 26 x 26 x 256 83 Couldn't find activation function mish, going with ReLU conv 256 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 256 0.089 BFLOPs 84 route 83 56 85 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 512 0.354 BFLOPs 86 Couldn't find activation function mish, going with ReLU conv 1024 3 x 3 / 2 26 x 26 x 512 -> 13 x 13 x1024 1.595 BFLOPs 87 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 88 route 86 89 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 90 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs 91 Couldn't find activation function mish, going with ReLU conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs 92 res 89 13 x 13 x 512 -> 13 x 13 x 512 93 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs 94 Couldn't find activation function mish, going with ReLU conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs 95 res 92 13 x 13 x 512 -> 13 x 13 x 512 96 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs 97 Couldn't find activation function mish, going with ReLU conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs 98 res 95 13 x 13 x 512 -> 13 x 13 x 512 99 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs 100 Couldn't find activation function mish, going with ReLU conv 512 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.797 BFLOPs 101 res 98 13 x 13 x 512 -> 13 x 13 x 512 102 Couldn't find activation function mish, going with ReLU conv 512 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 512 0.089 BFLOPs 103 route 102 87 104 Couldn't find activation function mish, going with ReLU conv 1024 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x1024 0.354 BFLOPs 105 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 106 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs 107 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 108 max 5 x 5 / 1 13 x 13 x 512 -> 13 x 13 x 512 109 route 107 110 max 9 x 9 / 1 13 x 13 x 512 -> 13 x 13 x 512 111 route 107 112 max 13 x 13 / 1 13 x 13 x 512 -> 13 x 13 x 512 113 route 112 110 108 107 114 conv 512 1 x 1 / 1 13 x 13 x2048 -> 13 x 13 x 512 0.354 BFLOPs 115 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs 116 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 117 conv 256 1 x 1 / 1 13 x 13 x 512 -> 13 x 13 x 256 0.044 BFLOPs 118 upsample 2x 13 x 13 x 256 -> 26 x 26 x 256 119 route 85 120 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 121 route 120 118 122 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 123 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs 124 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 125 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs 126 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 127 conv 128 1 x 1 / 1 26 x 26 x 256 -> 26 x 26 x 128 0.044 BFLOPs 128 upsample 2x 26 x 26 x 128 -> 52 x 52 x 128 129 route 54 130 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs 131 route 130 128 132 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs 133 conv 256 3 x 3 / 1 52 x 52 x 128 -> 52 x 52 x 256 1.595 BFLOPs 134 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs 135 conv 256 3 x 3 / 1 52 x 52 x 128 -> 52 x 52 x 256 1.595 BFLOPs 136 conv 128 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 128 0.177 BFLOPs 137 conv 256 3 x 3 / 1 52 x 52 x 128 -> 52 x 52 x 256 1.595 BFLOPs 138 conv 75 1 x 1 / 1 52 x 52 x 256 -> 52 x 52 x 75 0.104 BFLOPs 139 yolo Unused field: 'scale_x_y = 1.2' Unused field: 'iou_thresh = 0.213' Unused field: 'cls_normalizer = 1.0' Unused field: 'iou_normalizer = 0.07' Unused field: 'iou_loss = ciou' Unused field: 'nms_kind = greedynms' Unused field: 'beta_nms = 0.6' Unused field: 'max_delta = 5' 140 route 136 141 conv 256 3 x 3 / 2 52 x 52 x 128 -> 26 x 26 x 256 0.399 BFLOPs 142 route 141 126 143 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 144 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs 145 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 146 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs 147 conv 256 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 256 0.177 BFLOPs 148 conv 512 3 x 3 / 1 26 x 26 x 256 -> 26 x 26 x 512 1.595 BFLOPs 149 conv 75 1 x 1 / 1 26 x 26 x 512 -> 26 x 26 x 75 0.052 BFLOPs 150 yolo Unused field: 'scale_x_y = 1.1' Unused field: 'iou_thresh = 0.213' Unused field: 'cls_normalizer = 1.0' Unused field: 'iou_normalizer = 0.07' Unused field: 'iou_loss = ciou' Unused field: 'nms_kind = greedynms' Unused field: 'beta_nms = 0.6' Unused field: 'max_delta = 5' 151 route 147 152 conv 512 3 x 3 / 2 26 x 26 x 256 -> 13 x 13 x 512 0.399 BFLOPs 153 route 152 116 154 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 155 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs 156 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 157 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs 158 conv 512 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 512 0.177 BFLOPs 159 conv 1024 3 x 3 / 1 13 x 13 x 512 -> 13 x 13 x1024 1.595 BFLOPs 160 conv 75 1 x 1 / 1 13 x 13 x1024 -> 13 x 13 x 75 0.026 BFLOPs 161 yolo Unused field: 'scale_x_y = 1.05' Unused field: 'iou_thresh = 0.213' Unused field: 'cls_normalizer = 1.0' Unused field: 'iou_normalizer = 0.07' Unused field: 'iou_loss = ciou' Unused field: 'nms_kind = greedynms' Unused field: 'beta_nms = 0.6' Unused field: 'max_delta = 5'
the following is the cfg file for yolov4