k4rk0or/

Handwritten-Digits-Recognition

Datasets MNIST

Epoch 0 - Training loss: 0.6262280630039126 Epoch 1 - Training loss: 0.27758935431856463 Epoch 2 - Training loss: 0.21520252936064943 Epoch 3 - Training loss: 0.1722082135114652 Epoch 4 - Training loss: 0.14558795913021322 Epoch 5 - Training loss: 0.1254206331395137 Epoch 6 - Training loss: 0.11103267670432324 Epoch 7 - Training loss: 0.09751890437887993 Epoch 8 - Training loss: 0.0890058792992505 Epoch 9 - Training loss: 0.080117929476633 Epoch 10 - Training loss: 0.07290435441807389 Epoch 11 - Training loss: 0.06701811250777784 Epoch 12 - Training loss: 0.061741279120912464 Epoch 13 - Training loss: 0.05696900679157979 Epoch 14 - Training loss: 0.051936965980090455

10000 Images Tested Model is 97.46000000000001% accurate