search

KaihuaTang / ResNet50-Pytorch-Face-Recognition

Using Pytorch to implement a ResNet50 for Cross-Age Face Recognition
MIT License
144 stars 49 forks source link

res2net在imagenet预训练模型 #4

Open yile824 opened 5 years ago

yile824 commented 5 years ago

因为计算资源有限，想要您预训练好的模型。

zhangbaijin commented 4 years ago

Epoch : 0, Batch : 700, Loss : 7.609572, Batch Accuracy 0.000000 Input Label : tensor([ 754, 197, 1778, 1205], device='cuda:0') Output Label : tensor([1753, 1753, 1753, 1833], device='cuda:0') Epoch : 0, Batch : 750, Loss : 7.835464, Batch Accuracy 0.000000 Input Label : tensor([1194, 888, 1766, 1407], device='cuda:0') Output Label : tensor([1833, 1833, 1833, 1833], device='cuda:0') Epoch : 0, Batch : 800, Loss : 7.709994, Batch Accuracy 0.000000 Input Label : tensor([1380, 12, 598, 1513], device='cuda:0') Output Label : tensor([1833, 1833, 1833, 1833], device='cuda:0') Epoch : 0, Batch : 850, Loss : 7.423506, Batch Accuracy 0.000000 Input Label : tensor([ 969, 1653, 1255, 387], device='cuda:0') Output Label : tensor([1833, 1833, 1833, 1833], device='cuda:0') Epoch : 0, Batch : 900, Loss : 7.549356, Batch Accuracy 0.000000 Input Label : tensor([1299, 869, 1788, 1213], device='cuda:0') Output Label : tensor([1833, 1753, 1753, 1753], device='cuda:0') Epoch : 0, Batch : 950, Loss : 7.509100, Batch Accuracy 0.000000 Input Label : tensor([1340, 1549, 353, 59], device='cuda:0') Output Label : tensor([1833, 1753, 1753, 1753], device='cuda:0') Epoch : 0, Batch : 1000, Loss : 7.717154, Batch Accuracy 0.000000 请问你遇到过这样的情况吗？