← Back to Architecture

RMSNorm

Architecture
Used in
3 PRs
Best BPB
1.2156
Avg BPB
1.4006

Submissions

PR #85by hydeh3r3
1.2156
PR #1542by negrurv
1.5363
PR #1558by Subramanyam6
1.4500

Hyperparameters Across PRs

pr_numberparameters
85—
1542—
1558{"slot_parameters":512}