I’m using resnet34 to train my fast.ai model to train it on the MNIST database, but I’m getting very poor results. Any idea how to use other architectures like Lenet-5 on fast.ai?
I’m using resnet34 to train my fast.ai model to train it on the MNIST database, but I’m getting very poor results. Any idea how to use other architectures like Lenet-5 on fast.ai?