← Back to Test-Time Training
readout_only
Test-Time TrainingUsed in
1 PRs
Best BPB
1.0976
Avg BPB
1.0976
Submissions
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 1704 | {"learning_rate":0.005,"epochs":3,"prefix_chunk_ratio":0.2,"prefix_epochs":4,"prefix_lr_scale":1.15,"prefix_hard_window_fraction":0.25} |