Train a two-layer MLP on XOR using sequential { } and AdamW.
dotnet run --project examples/SimpleTrainingTraining XOR with AdamW...
epoch 100 loss = ...
epoch 200 loss = ...
epoch 300 loss = ...
epoch 400 loss = ...
epoch 500 loss = ...
Predictions (expected: 0, 1, 1, 0):
0 XOR 0 = 0.0xx
0 XOR 1 = 0.9xx
1 XOR 0 = 0.9xx
1 XOR 1 = 0.0xx
sequential { }CE for composingIModulelayersLinear.init,ReluactivationAdamWoptimizer,Loss.mse