Skip to content
New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

FIX: SAELens new format has "scaling_factor" key, which causes assert to fail #39

Merged
merged 3 commits into from
Apr 20, 2024
Merged

FIX: SAELens new format has "scaling_factor" key, which causes assert to fail #39

Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
6852170
Allow SAELens autoencoder keys to be superset of required keys, inste…
hijohnnylin Apr 19, 2024
21b0086
Merge branch 'main' into fix_loading_saelens_sae
hijohnnylin Apr 19, 2024
13b8106
Fix Formatting
hijohnnylin Apr 20, 2024
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
8 changes: 5 additions & 3 deletions sae_vis/data_storing_fns.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -1016,14 +1016,16 @@ def create(

# If encoder isn't an AutoEncoder, we wrap it in one
if not isinstance(encoder, AutoEncoder):
assert (
set(encoder.state_dict().keys()) == {"W_enc", "W_dec", "b_enc", "b_dec"}
assert set(
encoder.state_dict().keys()
).issuperset(
{"W_enc", "W_dec", "b_enc", "b_dec"}
), "If encoder isn't an AutoEncoder, it should have weights 'W_enc', 'W_dec', 'b_enc', 'b_dec'"
d_in, d_hidden = encoder.W_enc.shape
device = encoder.W_enc.device
encoder_cfg = AutoEncoderConfig(d_in=d_in, d_hidden=d_hidden)
encoder_wrapper = AutoEncoder(encoder_cfg).to(device)
encoder_wrapper.load_state_dict(encoder.state_dict())
encoder_wrapper.load_state_dict(encoder.state_dict(), strict=False)
else:
encoder_wrapper = encoder

Expand Down