MMMMoK9922's picture
Upload candidate_value_qslow_active model and summary
626888c verified
Raw History Blame Contribute Delete
3.41 kB
{
"collection": {
"input_files": 61,
"complete_groups": 1116,
"class_counts": {
"all_fail": 13,
"all_success": 867,
"normal_needed": 103,
"other": 92,
"slow_needed": 41
},
"selected_groups": 449,
"selected_class_counts": {
"all_fail": 13,
"all_success": 200,
"normal_needed": 103,
"other": 92,
"slow_needed": 41
},
"targets": {
"retained_groups": 400,
"max_all_success": 200
},
"targets_met": true,
"records_output": "/kairos_vepfs_volc/data/manlichen/MetaRoboBench/candidate_value/p05ff500c30_qslow_active_g8_20260813/selected_records.jsonl"
},
"heads": 8,
"test": [
{
"groups": 72,
"selected_successes": 49,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5555555555555556,
"success_accuracy": 0.6574074074074074,
"selected_speed_counts": {
"fast": 56,
"normal": 12,
"slow": 4
}
},
{
"groups": 72,
"selected_successes": 50,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5694444444444444,
"success_accuracy": 0.6481481481481481,
"selected_speed_counts": {
"fast": 51,
"normal": 14,
"slow": 7
}
},
{
"groups": 72,
"selected_successes": 47,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5555555555555556,
"success_accuracy": 0.6435185185185185,
"selected_speed_counts": {
"fast": 55,
"normal": 15,
"slow": 2
}
},
{
"groups": 72,
"selected_successes": 47,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5555555555555556,
"success_accuracy": 0.6712962962962963,
"selected_speed_counts": {
"fast": 60,
"normal": 10,
"slow": 2
}
},
{
"groups": 72,
"selected_successes": 44,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5,
"success_accuracy": 0.6712962962962963,
"selected_speed_counts": {
"fast": 56,
"normal": 15,
"slow": 1
}
},
{
"groups": 72,
"selected_successes": 49,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5,
"success_accuracy": 0.6527777777777778,
"selected_speed_counts": {
"fast": 50,
"normal": 11,
"slow": 11
}
},
{
"groups": 72,
"selected_successes": 49,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5416666666666666,
"success_accuracy": 0.6851851851851852,
"selected_speed_counts": {
"fast": 55,
"normal": 13,
"slow": 4
}
},
{
"groups": 72,
"selected_successes": 50,
"slow_successes": 65,
"oracle_successes": 72,
"preferred_accuracy": 0.5416666666666666,
"success_accuracy": 0.6435185185185185,
"selected_speed_counts": {
"fast": 49,
"normal": 18,
"slow": 5
}
}
],
"selection": "argmax_value_no_threshold",
"label_semantics": "Q^Slow: success first; among successful candidate chunks choose the lowest final episode frame count, followed by Slow continuation"
}