CEFR_Bert_Fine-tuned / eval_per_level.json
SNALYF's picture
Init Repo
46d93f9 verified
{
"checkpoint": "checkpoints/reproduce/final.pt",
"state_used": "best",
"overall": {
"mse": 0.5665906071662903,
"mae": 0.5687209367752075,
"acc_exact": 0.5146067415730337,
"acc_pm1": 0.9393258426966292,
"n": 445
},
"per_level": {
"A1": {
"n": 39,
"acc_exact": 0.5128205128205128,
"acc_pm1": 1.0,
"mse": 0.3645050823688507,
"mae": 0.5215429067611694,
"mean_pred": 1.5343633890151978,
"mean_target": 1.0128204822540283
},
"A2": {
"n": 86,
"acc_exact": 0.47674418604651164,
"acc_pm1": 0.9534883720930233,
"mse": 0.4580419659614563,
"mae": 0.5539605617523193,
"mean_pred": 2.3083271980285645,
"mean_target": 2.1046512126922607
},
"B1": {
"n": 52,
"acc_exact": 0.4423076923076923,
"acc_pm1": 0.9807692307692307,
"mse": 0.5185056924819946,
"mae": 0.5999152064323425,
"mean_pred": 3.1202821731567383,
"mean_target": 3.0
},
"B2": {
"n": 128,
"acc_exact": 0.4609375,
"acc_pm1": 0.890625,
"mse": 0.6966836452484131,
"mae": 0.6493818163871765,
"mean_pred": 3.892303943634033,
"mean_target": 3.9873046875
},
"C1": {
"n": 62,
"acc_exact": 0.46774193548387094,
"acc_pm1": 0.9354838709677419,
"mse": 0.9029945731163025,
"mae": 0.6774065494537354,
"mean_pred": 4.686697006225586,
"mean_target": 4.967741966247559
},
"C2": {
"n": 78,
"acc_exact": 0.7307692307692307,
"acc_pm1": 0.9487179487179487,
"mse": 0.33848774433135986,
"mae": 0.36903026700019836,
"mean_pred": 5.595437049865723,
"mean_target": 5.895833492279053
}
},
"confusion_matrix": [
[
20,
19,
0,
0,
0,
0
],
[
12,
41,
29,
4,
0,
0
],
[
0,
13,
23,
15,
1,
0
],
[
1,
9,
27,
59,
28,
4
],
[
1,
1,
2,
19,
29,
10
],
[
0,
0,
1,
3,
17,
57
]
],
"cefr_to_idx": {
"A1": 0,
"A2": 1,
"B1": 2,
"B2": 3,
"C1": 4,
"C2": 5
}
}