Phil Wang

@Johnly1986 could you try again on the latest version?

> in autoregressive_q_learn() method of QLearner the time steps dimension is moved into batch dimension. E.g. if we have a batch of 16 samples with 3 timesteps each - this...

@yuriy10139 also, you should reach out to @2M-kotb, as i think he was playing around with the repo for his research some time back

@yuriy10139 thank you! i'll add a few things soon

@Michelangelo-Y thank you for reporting this could you try 0.4.3?

@forwiat is this the latest version? thought i fixed this 🤔 try actually turning off code expiration, and just use `rotation_trick = True`. this new research finding may obsolete all...

@JunityZhan are you sure? researchers are telling me it doesn't perform well at all with low precision.

@JunityZhan ok, if what you are saying is true, then i can think about it but your fix will break for your own condition `In training, I think it is...

@JunityZhan right right. the difference is whether i always enforce f32 during training so that people (non researchers) have a greater chance of success using the library, or allow for...

@JunityZhan do you want to see if setting [this](https://github.com/lucidrains/vector-quantize-pytorch/commit/be92f7908b938cf00c60d643c5ecffb01c7fa9fc#diff-3488501716ef0c4a1b84127592be7e1015e0faf0d7d3b630b69d88e55a134701R117) `False` works for you?