Skip to content

docs: fix incorrect spectral radius computation in README example #50

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

Open
sunnyc0206 wants to merge 3 commits into
base: main
Choose a base branch
from
Open

docs: fix incorrect spectral radius computation in README example #50

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
3 commits
Select commit Hold shift + click to select a range
c697e81
Update spectral radius logging in README
sunnyc0206 Apr 22, 2026
732c460
Merge branch 'main' into main
sunnyc0206 Apr 23, 2026
7eb3a5b
Add assertion to check for NaN loss
sunnyc0206 Apr 23, 2026
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
2 changes: 1 addition & 1 deletion README.md
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -99,7 +99,7 @@ print(f"[{attn_type.upper()}] Generated shape: {out.shape}")
A = model.recurrent.injection.get_A()
rho = torch.linalg.eigvals(A).abs().max().item()
print(
f"[{attn_type.upper()}] Spectral radius ρ(A) = {rho:.4f} (must be < 1)"
f"[{attn_type.upper()}] Spectral radius ρ(A): {rho:.4f} (must be < 1)"
)
```

Expand Down
1 change: 1 addition & 0 deletions examples/moda_example.py
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -39,6 +39,7 @@
labels = torch.randint(0, cfg.vocab_size, (B, T), device=device)

logits, loss = model(input_ids, labels)
assert not torch.isnan(loss), "Loss is NaN!"
assert logits.shape == (B, T, cfg.vocab_size)
print(f"Logits shape : {logits.shape}")
print(f"Loss (LM + balance): {loss.item():.4f}")
Expand Down