Datasetpublic

anirudhb11/rebase_gpt-oss-20b_rg_cognition_ns8_md4_bt0_1_seed65_rg_cognition__v0

Aggregated Metrics Weighted mean of per-shard wandb summary values (weights = shard rows). See _meta/aggregated_metrics.json for raw values and _meta/splits.json for the per-shard wandb run names. Aggregated from 10 shards. Metric Value avg_response_tokens 5759.1 generation_phase_time_s 271.844 judge_output_tokens_level_1 31023.8 judge_output_tokens_level_2 8833.8 judge_output_tokens_level_3 5082.6 judge_output_tokens_level_4 4306.8… See the full description on the dataset page: https://huggingface.co/datasets/anirudhb11/rebase_gpt-oss-20b_rg_cognition_ns8_md4_bt0_1_seed65_rg_cognition__v0.

sourceHugging Faceupdated 5mo agoView on Hugging Face

0likes4downloads

Dataset Card

Aggregated Metrics

Weighted mean of per-shard wandb summary values (weights = shard rows). See `meta/aggregatedmetrics.json` for raw values and `meta/splits.json` for the per-shard wandb run names._

Aggregated from 10 shards.

Metric	Value
`avg_response_tokens`	5759.1
`generation_phase_time_s`	271.844
`judge_output_tokens_level_1`	31023.8
`judge_output_tokens_level_2`	8833.8
`judge_output_tokens_level_3`	5082.6
`judge_output_tokens_level_4`	4306.8
`judge_output_tokens_search`	49247
`judge_skipped_completed_level_1`	47
`judge_skipped_completed_level_2`	16.3
`judge_skipped_completed_level_3`	5.3
`judge_skipped_completed_level_4`	2.6
`judge_token_cdf_level_1`	1
`judge_token_cdf_level_2`	1
`judge_token_cdf_level_3`	1
`judge_token_cdf_level_4`	1
`judge_token_count_level_1`	6835
`judge_token_count_level_2`	2060
`judge_token_count_level_3`	1015
`judge_token_count_level_4`	1145
`maj@1`	0.480211
`maj@1_std`	0.0678333
`maj@2`	0.48697
`maj@2_std`	0.0650374
`maj@4`	0.518102
`maj@4_std`	0.0509059
`maj@8`	0.52316
`maj@8_std`	0.0328563
`median_response_tokens`	2744.3
`num_unique_answers@1`	0.932
`num_unique_answers@2`	1.365
`num_unique_answers@4`	2.063
`num_unique_answers@8`	3.32
`pass@1`	0.4575
`pass@2`	0.534286
`pass@4`	0.571429
`pass@8`	0.6
`policy_output_tokens_final_completion`	75062.8
`policy_output_tokens_level_1`	187391
`policy_output_tokens_level_2`	100893
`policy_output_tokens_level_3`	55917.1
`policy_output_tokens_level_4`	41531.4
`policy_output_tokens_search_phases`	385733
`policy_output_tokens_total_so_far`	460795
`token_cdf`	1
`token_count`	26650
`total_generated_output_tokens`	550838
`total_judge_output_tokens`	90042.6
`total_judge_output_tokens_search`	49247
`total_judge_output_tokens_weighted_mv`	40795.6
`total_policy_output_tokens`	460795
`total_time_s`	311.382
`truncated_count_level_1`	0.3
`truncated_count_level_2`	0
`truncated_count_level_3`	0
`truncated_count_level_4`	0
`truncated_rate_level_1`	0.00858586
`truncated_rate_level_2`	0
`truncated_rate_level_3`	0
`truncated_rate_level_4`	0
`w_best@1`	0.492082
`w_best@1_std`	0.0698718
`w_best@2`	0.491226
`w_best@2_std`	0.0698433
`w_best@4`	0.480332
`w_best@4_std`	0.0689084
`w_best@8`	0.49261
`w_best@8_std`	0.0550959
`w_maj@1`	0.480211
`w_maj@1_std`	0.0678333
`w_maj@2`	0.491904
`w_maj@2_std`	0.0625095
`w_maj@4`	0.525595
`w_maj@4_std`	0.0493452
`w_maj@8`	0.52818
`w_maj@8_std`	0.0329451
`w_num_unique_answers@1`	0.932
`w_num_unique_answers@2`	1.365
`w_num_unique_answers@4`	2.063
`w_num_unique_answers@8`	3.32