on-demand

for tailwindcss:

<div className="relative w-4 h-4 bg-red-400 after:absolute after:top-4 after:w-full after:h-full after:bg-blue-400" />

for windicss:

<div className="relative w-4 h-4 bg-red-400 after:(absolute top-4 w-full h-full bg-blue-400)" />
`

Hi Jiahui,

Thanks for the great work. I'm trying to reproduce AutoSlim for CIFAR-10 (Table 2).
Could you please provide a detailed hyperparameter you used for it?

I'm able to train the baseline MobileNetV2 1.0x to 7.9 Top-1 error using the following hyperparameters:

• 0.1 initial learning rate
• linear learning rate decay
• 128 batch size
• 300 epochs of training
• 5e-4 weight decay

As clamd already has support to run custom scripts.