← Back to Architecture
Transformer size
ArchitectureUsed in
1 PRs
Best BPB
1.6231
Avg BPB
1.6231
Submissions
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 248 | {"layers":8,"model_dim":512} |
| pr_number | parameters |
|---|---|
| 248 | {"layers":8,"model_dim":512} |