Replies: 1 comment 1 reply
-
|
bumping this up. @cgarciae perhaps you have some insights? |
Beta Was this translation helpful? Give feedback.
1 reply
-
|
Would really appreciate any available insight into this! |
Beta Was this translation helpful? Give feedback.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
In torch you can repeat/broadcast an
nn.Paramand use it for further processing. This is useful for trainable initialization states of LSTMs and latent variable models. And even in the case of masked auto encoders (although no repetition occurs there).How can I repeat a
nnx.Paramtype using einops or jnp#tile or jnp#repeat before using it innnx.Scan?Beta Was this translation helpful? Give feedback.
All reactions