← Back to Architecture
PLE
ArchitectureUsed in
1 PRs
Best BPB
1.2195
Avg BPB
1.2195
Submissions
Hyperparameters Across PRs
| pr_number | parameters |
|---|---|
| 2062 | {"per_layer_embed_dim":64,"per_layer_embed_init_std":0.02} |