Matias Häggman

icon indicating a website link https://www.linkedin.com/in/matias-h%C3%A4ggman/ icon indicating an email [email protected]

icon company University of Helsinki icon location Helsinki

Results 3 comments of Matias Häggman

I am getting nan values in the mlstm output

I am also having this issue, with both the slstm and the mlstm. After quick debugging I found the 'gates' tensor to be filled with nans in the forward pass...

I am getting nan values in the mlstm output

> we should be able to construct a combination architecture using both sLSTM and mLSTM blocks, infact that is exactly how the original research paper suggest we use xLSTM. >...

Loading extension module slstm_HS128BS8NH4NS4DBfDRbDWbDGbDSbDAfNG4SA1GRCV0GRC0d0FCV0FC0d0...

I have this problem aswell. This only occurs if you include the sLSTM module in the xLSTM stack, using only mLSTM works. I tested this on the lightning platform with...