abcalan

I wonder if this pruning approach can be adapted to net like resnet

Why training accuracy is 0.000000?

3

epoch 0, total_step 300, total loss is 51.74 , inference loss is 37.28, weight deacy loss is 14.46, training accuracy is 0.000000, time 33.227 samples/sec epoch 0, total_step 320, total...

TypeError: must be real number, not Tensor

2

Tensor("tower_0/resnet_v1_100/E_BN1/Identity:0", shape=(512,), dtype=float32, device=/device:GPU:0) Tensor("tower_0/resnet_v1_100/E_BN1/Identity_1:0", shape=(512,), dtype=float32, device=/device:GPU:0) Traceback (most recent call last): File "train_nets_mgpu_new.py", line 135, in net = get_resnet(images_s[i], args.net_depth, type='ir', w_init=w_init_method, trainable=True, keep_rate=dropout_rate) File "/home/zhangweiwei/InsightFace_TF-master/nets/L_Resnet_E_IR_MGPU.py", line 223,...

Why training accuracy is small after total_step 46680?

1

epoch 0, total_step 46560, total loss is 24.23 , inference loss is 12.90, weight deacy loss is 11.33, training accuracy is 0.062500, time 32.760 samples/sec epoch 0, total_step 46580, total...

Caused by op 'tower_3/resnet_v1_100/Placeholder', defined at: File "/home/zhangweiwei/InsightFace_TF-master/train_nets_mgpu_new.py", line 135, in net = get_resnet(images_s[i], args.net_depth, type='ir', w_init=w_init_method, trainable=True, keep_rate=dropout_rate) File "/home/zhangweiwei/InsightFace_TF-master/nets/L_Resnet_E_IR_MGPU.py", line 223, in get_resnet scope='resnet_v1_%d' % num_layers) File "/home/zhangweiwei/InsightFace_TF-master/nets/L_Resnet_E_IR_MGPU.py",...

n the training process, the center loss becomes larger and larger. What is the mistake

5


yolov3.weights是在什么数据集上预训练的权重

那相当于博主这个项目并没有实现shorcut部分的剪枝？这样会不会影响最终效果？

mAP上有区别吗？使用上有区别吗

我用自己的模型进行推理

3

推理结果为空，请问是什么原因呢