← Back to Architecture

MLP activation

Architecture
Used in
6 PRs
Best BPB
1.2302
Avg BPB
1.4614

Submissions

PR #73by NishantDahal
1.3281
PR #554by chrisnkuno
1.4612
PR #675by ChideraIbe123
1.2302
PR #830by zlxi02
1.4096
PR #1299by Ribin545
1.8184
PR #1434by ranausmanai
1.5207

Hyperparameters Across PRs

pr_numberparameters
73—
554—
675{"negative_slope":0.5,"power":2}
830{"negative_slope":0.5}
1299—
1434{"activation":"silu2"}