decision-model / RESULT.json
gparitosh007's picture
v2: ModernBERT-base-nli, v2 data + open data, seed 42
2400eca verified
Raw History Blame Contribute Delete
3.2 kB
{
"data": "/home/eai-ajzxf/dm/decision-model/data/merged_v2_bal.jsonl,/home/eai-ajzxf/dm/decision-model/data/open_open_tasksource.jsonl,/home/eai-ajzxf/dm/decision-model/data/open_ratings.jsonl,/home/eai-ajzxf/dm/decision-model/data/open_support_intents.jsonl",
"model": "/home/eai-ajzxf/dm/models/ModernBERT-base-nli",
"init_from": null,
"domain_weights": {
"generic": 3.0,
"open_tasksource": 2.0
},
"seed": 42,
"epochs": 3,
"lr": 3e-05,
"max_len": 1024,
"batch_size": 16,
"device": "cuda",
"split_mode": "template",
"exclude_domains": [],
"split": {
"train": 256920,
"validation": 37978,
"heldout": 0,
"validation_template_ids": 1892
},
"class_weight": true,
"majority_baseline": {
"source": "train split",
"validation": {
"n": 35870,
"accuracy": 0.40398661834402005,
"n_unmapped": 2108
},
"heldout": null
},
"steps": 48174,
"final_loss": 0.06356,
"temperature": 3.986253,
"metrics_validation_raw": {
"n": 37978,
"accuracy": 0.46753383537837695,
"balanced_accuracy": 0.45502415494688436,
"ece": 0.2291721958145001,
"nll": 1.392849499783441
},
"metrics_validation_calibrated": {
"n": 37978,
"accuracy": 0.46753383537837695,
"balanced_accuracy": 0.45502415494688436,
"ece": 0.04692186695970413,
"nll": 1.0603329397388084
},
"validation_by_domain": {
"generic": {
"n": 4888,
"accuracy": 0.717266775777414,
"balanced_accuracy": 0.6540912238268553,
"ece": 0.19083873238932775,
"nll": 0.8106594818093334
},
"open_tasksource": {
"n": 8936,
"accuracy": 0.40689346463742165,
"balanced_accuracy": 0.46969241031024717,
"ece": 0.06341728585252776,
"nll": 1.071087839243584
},
"routing": {
"n": 6577,
"accuracy": 0.36658050783031776,
"balanced_accuracy": 0.4599181298323277,
"ece": 0.1523182318985272,
"nll": 1.2406663806953009
},
"shopping": {
"n": 5558,
"accuracy": 0.6765023389708529,
"balanced_accuracy": 0.6502757329467332,
"ece": 0.20365982418246473,
"nll": 0.9283356470128373
},
"support": {
"n": 6131,
"accuracy": 0.21497308758766923,
"balanced_accuracy": 0.2544055306860031,
"ece": 0.27600667550928965,
"nll": 1.2674843568678904
},
"tool_risk": {
"n": 5457,
"accuracy": 0.5041231445849368,
"balanced_accuracy": 0.3976298867419834,
"ece": 0.05821240913284683,
"nll": 0.991137123980093
},
"travel": {
"n": 431,
"accuracy": 0.8677494199535963,
"balanced_accuracy": 0.9055462327315874,
"ece": 0.2595401152518952,
"nll": 0.5485979475182198
}
},
"heldout_domain": "none",
"validation_by_question_type": {
"choice": {
"n": 17528,
"accuracy": 0.4361592879963487,
"balanced_accuracy": 0.5408672503546958
},
"noul": {
"n": 10052,
"accuracy": 0.6830481496219658,
"balanced_accuracy": 0.7107085773844286
},
"score": {
"n": 10398,
"accuracy": 0.31207924600884784,
"balanced_accuracy": 0.33391884754359086
}
}
}