← Back to Architecture
Turbo-Muon
ArchitectureUsed in
1 PRs
Best BPB
1.1091
Avg BPB
1.1091
Submissions
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 1126 | {"newton_schulz_iterations":4} |
| pr_number | parameters |
|---|---|
| 1126 | {"newton_schulz_iterations":4} |