-
## ❓Question
Layer 55: mvit_block_0_transformer_0_LN1
Input shape: (None, 4, 1024, 144)
Output shape: (None, 4, 1024, 144)
Input spec: None
Layer 56: mvit_block_0_transformer_0_attention
…
-
Token indices sequence length is longer than the specified maximum sequence length for this model (2409 > 2048). Running this sequence through the model will result in indexing errors
I got this wa…
-
when `initial_connection` value in the config set to fs_neat_hidden, partial_direct, full_direct, and and `num_hidden` is other than zero, creates population of size, different from what is set in con…
-
Thanks for the great paper!
I've created another open source implementation of the SHA RNN here: https://github.com/talonvoice/sha-rnn
I trained with similar parameters to the single head model …
-
Assuming an AR-HMM, I want to know if it's possible to constrain the AR model parameters. Specifically, is it possible to set the 'A' matrix to fixed parameters and only estimate parameters for the in…
-
Hey,
Do you think it is possible to add the initialization from https://openreview.net/forum?id=H1lma24tPB to this model?
Thanks,
Ohad
-
||link|
|----|---|
|paper| [Fully Convolutional Networks for Semantic Segmentation](https://openaccess.thecvf.com/content_cvpr_2015/html/Long_Fully_Convolutional_Networks_2015_CVPR_paper.html) |
|c…
-
I tried to run scripts like the one in multigpu examples but still it isn't as nice as prisma. I know that prisma is just derived from this network but still, this should work even better then. It som…
-
Is this a trivial thing or more tricky?
Hello, so I'm on linux (Zorin OS 16) and I seem to need CImg.h, I tried to upgrade:
> You can upgrade your CImg.h to version 2.8.4 or later by getting it …
-
For linear compositions and so on, for example when using `seqP`, `seqPLoop`, and `ur`, it makes sense to reset the clock, so that a loop can be started 'now'.
There is a quick hack for resetting t…