2 ms·

The hosting looks cool but not efficient. Would you consider support batch inference for a more scalable solution?