35 lines
784 B
JSON
35 lines
784 B
JSON
{
|
|
"schema_version": 1,
|
|
"objectives": [
|
|
"helpfulness",
|
|
"safety",
|
|
"conciseness"
|
|
],
|
|
"normalization": "per_prompt_minmax",
|
|
"adversary": {
|
|
"steps": 25,
|
|
"alpha": 1.0,
|
|
"kappa": 0.05,
|
|
"preference_scale": 8.0
|
|
},
|
|
"ronpo_estimator": "Rao-Blackwellized full sigma expectation with three common response pairs per prompt",
|
|
"counts": {
|
|
"train": {
|
|
"avg": 16746,
|
|
"ronpo_full_expect": 50340,
|
|
"ronpo_k_only": 50340,
|
|
"ht_mnpo_helpfulness": 16743,
|
|
"ht_mnpo_safety": 16764,
|
|
"ht_mnpo_conciseness": 16755
|
|
},
|
|
"validation": {
|
|
"avg": 1859,
|
|
"ronpo_full_expect": 5592,
|
|
"ronpo_k_only": 5592,
|
|
"ht_mnpo_helpfulness": 1864,
|
|
"ht_mnpo_safety": 1862,
|
|
"ht_mnpo_conciseness": 1858
|
|
}
|
|
}
|
|
}
|