Datasets:

eatang
/

Llama-3.2-1B-Instruct-best_of_n-prm-completions

Modalities:

Formats:

Size:

Libraries:

Dataset card Data Studio Files Files and versions Community

Dataset Viewer

Auto-converted to Parquet

Subset (2)

HuggingFaceH4_MATH-500--T-0.8--top_p-1.0--n-256--seed-0--agg_strategy-last--evals · 9 rows

Split (1)

train · 9 rows

Search is not available for this dataset

n int64 1 256	acc_naive float64 30.9 46.5	acc_weighted float64 30.7 51.6	acc_maj float64 30.5 46.1
64	46.484375	50	46.09375
16	42.96875	45.703125	42.1875
4	42.578125	43.359375	36.71875
128	45.507813	49.023438	45.3125
32	44.335938	47.65625	44.140625
8	43.164063	44.921875	38.476563
1	30.859375	30.664063	30.46875
2	37.695313	37.890625	30.46875
256	46.289063	51.5625	46.09375

README.md exists but content is empty.

Downloads last month: 31

Size of downloaded dataset files:

Size of the auto-converted Parquet files:

Number of rows: