Irina Saparina

@senthurRam33 I also think the problem is somewhere in gradients. I changed the loss to 'label_smooth' (see #10) and got more stable training.

Hi! Sorry about the delay with the answer. Did you try to change ```num_workers``` parameter in all DataLoaders (like [here](https://github.com/saparina/neural-ode/blob/e1107abd9e89d975a82d13d3f847faed3f30ea98/experiments/data_helper.py#L96))?

@CannonWWW Sorry, I don't remember the details but I think I just followed this advice: https://github.com/salesforce/TabularSemanticParsing/issues/14#issuecomment-790586474

Hi! My explanation is that tensor decomposition methods require more mathematical operations: instead of one (highly optimized in Pytorch) matrix multiplication, we have several. I think it is possible to...