sldbench

v1.0

SLDBench: A benchmark for scaling law discovery with symbolic regression tasks.

harbor run -d sldbench@1.0

Tasks (8)

data_constrained_scaling_law
harbor run -d sldbench@1.0 -t data_constrained_scaling_law
af3281a
domain_mixture_scaling_law
harbor run -d sldbench@1.0 -t domain_mixture_scaling_law
af3281a
easy_question_scaling_law
harbor run -d sldbench@1.0 -t easy_question_scaling_law
af3281a
lr_bsz_scaling_law
harbor run -d sldbench@1.0 -t lr_bsz_scaling_law
af3281a
moe_scaling_law
harbor run -d sldbench@1.0 -t moe_scaling_law
af3281a
parallel_scaling_law
harbor run -d sldbench@1.0 -t parallel_scaling_law
af3281a
sft_scaling_law
harbor run -d sldbench@1.0 -t sft_scaling_law
af3281a
vocab_scaling_law
harbor run -d sldbench@1.0 -t vocab_scaling_law
af3281a