add ability to vary more hyperparamaters in parallel training #6
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
g-w1
force-pushed
the
parallel-more-hyps
branch
2 times, most recently
from
d5d9ddf to
59785e2
Compare
This allows one to train multiple autoencoders off of a single layer since we aren't indexing a dictionary off the same thing. This could be useful for something like hyperparamater tuning where you only want to change one thing at a time. Here's an example: ``` submodules = [model.gpt_neox.layers[3].mlp, model.gpt_neox.layers[3].mlp, model.gpt_neox.layers[3].mlp] activation_dim = 512 # output dimension of the MLP dictionary_size = 16 * activation_dim learning_rates = [3e-4, 1e-3, 3e-3] ```
This allows one to re-train a sparse autoencoder on the same layer without re-generating all of the activations to train on.
g-w1
force-pushed
the
parallel-more-hyps
branch
from
a3ff78e to
72e23be
Compare
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
This can be useful for hyperparam search.