← Back to Architecture
MHA 8/8
ArchitectureUsed in
1 PRs
Best BPB
0.1582
Avg BPB
0.1582
Submissions
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 859 | {"layers":8,"heads":8} |
| pr_number | parameters |
|---|---|
| 859 | {"layers":8,"heads":8} |