← Back to Architecture
MLP6
ArchitectureUsed in
2 PRs
Best BPB
1.1636
Avg BPB
1.1636
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 1897 | {"layers":5,"mlp_mult":6,"model_dim":512} |
| 1917 | {"layers":5,"mlp_mult":6,"model_dim":512} |