Bauxi-23M / eval /baseline.json
classifyre's picture
Bauxi 1.0.0: weights, graphs, evaluation
0746964 verified
Raw History Blame Contribute Delete
2.11 kB
{
"model": "cross-encoder/nli-MiniLM2-L6-H768",
"revision": "b95119ce93d3e065de6214e38cd4a97b0f2f2c6d",
"parameters": 82120707,
"holdout": {
"count": 150,
"accuracy": 0.43333333333333335,
"macro_f1": 0.3870198371746359,
"invalid_recall": 0.44,
"false_safe": 1,
"false_safe_rate": 0.02,
"per_class": {
"valid": {
"precision": 0.7142857142857143,
"recall": 0.1,
"f1": 0.1754385964912281
},
"invalid": {
"precision": 0.55,
"recall": 0.44,
"f1": 0.48888888888888893
},
"unknown": {
"precision": 0.36893203883495146,
"recall": 0.76,
"f1": 0.4967320261437908
}
},
"confusion": {
"valid": {
"valid": 5,
"invalid": 7,
"unknown": 38
},
"invalid": {
"valid": 1,
"invalid": 22,
"unknown": 27
},
"unknown": {
"valid": 1,
"invalid": 11,
"unknown": 38
}
},
"nll": 1.731837511062622,
"brier_score": 0.8633853580312644
},
"regression": {
"count": 888,
"accuracy": 0.40878378378378377,
"macro_f1": 0.3026087574653168,
"invalid_recall": 0.17567567567567569,
"false_safe": 6,
"false_safe_rate": 0.02027027027027027,
"per_class": {
"valid": {
"precision": 0.7142857142857143,
"recall": 0.05859375,
"f1": 0.10830324909747294
},
"invalid": {
"precision": 0.5252525252525253,
"recall": 0.17567567567567569,
"f1": 0.26329113924050634
},
"unknown": {
"precision": 0.3854166666666667,
"recall": 0.8809523809523809,
"f1": 0.5362318840579711
}
},
"confusion": {
"valid": {
"valid": 15,
"invalid": 7,
"unknown": 234
},
"invalid": {
"valid": 6,
"invalid": 52,
"unknown": 238
},
"unknown": {
"valid": 0,
"invalid": 40,
"unknown": 296
}
},
"nll": 1.797816276550293,
"brier_score": 0.9120152167371303
}
}