littletrainingloop

Fix device placement bugs in loglikelihood evaluation (#2)

Add spelling benchmark for character-level evaluation (#1)

update growing mlp experiment (#39)

update growing mlp experiment

stackv2 training (#38)