roberta-large-ner-ghtk-cs-6-label-new-data-3090-11Sep-1

This model is a fine-tuned version of FacebookAI/xlm-roberta-large on the None dataset. It achieves the following results on the evaluation set:

• Loss: 0.1990
• Tk: {'precision': 0.8227848101265823, 'recall': 0.5603448275862069, 'f1': 0.6666666666666667, 'number': 116}
• Gày: {'precision': 0.7692307692307693, 'recall': 0.8823529411764706, 'f1': 0.8219178082191781, 'number': 34}
• Gày trừu tượng: {'precision': 0.9166666666666666, 'recall': 0.9241803278688525, 'f1': 0.9204081632653061, 'number': 488}
• Ã đơn: {'precision': 0.8413461538461539, 'recall': 0.8620689655172413, 'f1': 0.8515815085158149, 'number': 203}
• Đt: {'precision': 0.9330453563714903, 'recall': 0.9840546697038725, 'f1': 0.9578713968957872, 'number': 878}
• Đt trừu tượng: {'precision': 0.8502024291497976, 'recall': 0.9012875536480687, 'f1': 0.875, 'number': 233}
• Overall Precision: 0.9016
• Overall Recall: 0.9196
• Overall F1: 0.9105
• Overall Accuracy: 0.9689

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2.5e-05
• train_batch_size: 8
• eval_batch_size: 8
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Tk	Gày	Gày trừu tượng	à đơn	Đt	Đt trừu tượng	Overall Precision	Overall Recall	Overall F1	Overall Accuracy
No log	1.0	467	0.1481	{'precision': 1.0, 'recall': 0.04310344827586207, 'f1': 0.08264462809917356, 'number': 116}	{'precision': 0.5263157894736842, 'recall': 0.8823529411764706, 'f1': 0.6593406593406594, 'number': 34}	{'precision': 0.864476386036961, 'recall': 0.8627049180327869, 'f1': 0.8635897435897436, 'number': 488}	{'precision': 0.7544642857142857, 'recall': 0.8325123152709359, 'f1': 0.791569086651054, 'number': 203}	{'precision': 0.8652694610778443, 'recall': 0.9874715261958997, 'f1': 0.9223404255319149, 'number': 878}	{'precision': 0.8465346534653465, 'recall': 0.7339055793991416, 'f1': 0.7862068965517242, 'number': 233}	0.8412	0.8519	0.8465	0.9507
0.2706	2.0	934	0.1772	{'precision': 0.4626865671641791, 'recall': 0.2672413793103448, 'f1': 0.33879781420765026, 'number': 116}	{'precision': 0.5625, 'recall': 0.7941176470588235, 'f1': 0.6585365853658537, 'number': 34}	{'precision': 0.8574144486692015, 'recall': 0.9241803278688525, 'f1': 0.8895463510848126, 'number': 488}	{'precision': 0.8043478260869565, 'recall': 0.3645320197044335, 'f1': 0.5016949152542373, 'number': 203}	{'precision': 0.8913934426229508, 'recall': 0.9908883826879271, 'f1': 0.9385113268608414, 'number': 878}	{'precision': 0.6782006920415224, 'recall': 0.8412017167381974, 'f1': 0.7509578544061303, 'number': 233}	0.8253	0.8448	0.8349	0.9411
0.1019	3.0	1401	0.1537	{'precision': 0.84, 'recall': 0.1810344827586207, 'f1': 0.2978723404255319, 'number': 116}	{'precision': 0.6304347826086957, 'recall': 0.8529411764705882, 'f1': 0.725, 'number': 34}	{'precision': 0.8383458646616542, 'recall': 0.9139344262295082, 'f1': 0.8745098039215686, 'number': 488}	{'precision': 0.8407960199004975, 'recall': 0.8325123152709359, 'f1': 0.8366336633663367, 'number': 203}	{'precision': 0.88259526261586, 'recall': 0.9760820045558086, 'f1': 0.9269875608436994, 'number': 878}	{'precision': 0.8245614035087719, 'recall': 0.8068669527896996, 'f1': 0.8156182212581344, 'number': 233}	0.8537	0.8760	0.8647	0.9577
0.0747	4.0	1868	0.1361	{'precision': 0.7058823529411765, 'recall': 0.3103448275862069, 'f1': 0.4311377245508982, 'number': 116}	{'precision': 0.7560975609756098, 'recall': 0.9117647058823529, 'f1': 0.8266666666666665, 'number': 34}	{'precision': 0.8975903614457831, 'recall': 0.9159836065573771, 'f1': 0.9066937119675456, 'number': 488}	{'precision': 0.8907103825136612, 'recall': 0.8029556650246306, 'f1': 0.844559585492228, 'number': 203}	{'precision': 0.903125, 'recall': 0.9874715261958997, 'f1': 0.94341675734494, 'number': 878}	{'precision': 0.6449704142011834, 'recall': 0.9356223175965666, 'f1': 0.7635726795096323, 'number': 233}	0.8508	0.9027	0.8760	0.9610
0.0481	5.0	2335	0.1297	{'precision': 0.7428571428571429, 'recall': 0.4482758620689655, 'f1': 0.5591397849462366, 'number': 116}	{'precision': 0.7692307692307693, 'recall': 0.8823529411764706, 'f1': 0.8219178082191781, 'number': 34}	{'precision': 0.8933601609657947, 'recall': 0.9098360655737705, 'f1': 0.9015228426395939, 'number': 488}	{'precision': 0.9010416666666666, 'recall': 0.8522167487684729, 'f1': 0.8759493670886076, 'number': 203}	{'precision': 0.8815261044176707, 'recall': 1.0, 'f1': 0.9370330843116329, 'number': 878}	{'precision': 0.8477366255144033, 'recall': 0.8841201716738197, 'f1': 0.865546218487395, 'number': 233}	0.8753	0.9134	0.8940	0.9670
0.0367	6.0	2802	0.1397	{'precision': 0.8709677419354839, 'recall': 0.46551724137931033, 'f1': 0.6067415730337079, 'number': 116}	{'precision': 0.6829268292682927, 'recall': 0.8235294117647058, 'f1': 0.7466666666666667, 'number': 34}	{'precision': 0.898989898989899, 'recall': 0.9118852459016393, 'f1': 0.9053916581892167, 'number': 488}	{'precision': 0.8871794871794871, 'recall': 0.8522167487684729, 'f1': 0.8693467336683417, 'number': 203}	{'precision': 0.9268817204301075, 'recall': 0.9817767653758542, 'f1': 0.9535398230088497, 'number': 878}	{'precision': 0.8014981273408239, 'recall': 0.9184549356223176, 'f1': 0.856, 'number': 233}	0.8925	0.9098	0.9011	0.9681
0.0228	7.0	3269	0.1445	{'precision': 0.8372093023255814, 'recall': 0.6206896551724138, 'f1': 0.7128712871287128, 'number': 116}	{'precision': 0.7804878048780488, 'recall': 0.9411764705882353, 'f1': 0.8533333333333334, 'number': 34}	{'precision': 0.907070707070707, 'recall': 0.9200819672131147, 'f1': 0.9135300101729399, 'number': 488}	{'precision': 0.8271028037383178, 'recall': 0.8719211822660099, 'f1': 0.8489208633093526, 'number': 203}	{'precision': 0.9314775160599572, 'recall': 0.9908883826879271, 'f1': 0.9602649006622518, 'number': 878}	{'precision': 0.8708333333333333, 'recall': 0.8969957081545065, 'f1': 0.8837209302325582, 'number': 233}	0.9	0.9267	0.9132	0.9692
0.0162	8.0	3736	0.1854	{'precision': 0.9014084507042254, 'recall': 0.5517241379310345, 'f1': 0.6844919786096257, 'number': 116}	{'precision': 0.7222222222222222, 'recall': 0.7647058823529411, 'f1': 0.7428571428571428, 'number': 34}	{'precision': 0.9238683127572016, 'recall': 0.9200819672131147, 'f1': 0.9219712525667352, 'number': 488}	{'precision': 0.8293838862559242, 'recall': 0.8620689655172413, 'f1': 0.8454106280193237, 'number': 203}	{'precision': 0.9128151260504201, 'recall': 0.989749430523918, 'f1': 0.9497267759562841, 'number': 878}	{'precision': 0.8546255506607929, 'recall': 0.8326180257510729, 'f1': 0.8434782608695652, 'number': 233}	0.8961	0.9103	0.9032	0.9672
0.0084	9.0	4203	0.1853	{'precision': 0.8144329896907216, 'recall': 0.6810344827586207, 'f1': 0.7417840375586854, 'number': 116}	{'precision': 0.7435897435897436, 'recall': 0.8529411764705882, 'f1': 0.7945205479452054, 'number': 34}	{'precision': 0.9129554655870445, 'recall': 0.9241803278688525, 'f1': 0.9185336048879837, 'number': 488}	{'precision': 0.8333333333333334, 'recall': 0.8620689655172413, 'f1': 0.847457627118644, 'number': 203}	{'precision': 0.9492273730684326, 'recall': 0.979498861047836, 'f1': 0.9641255605381165, 'number': 878}	{'precision': 0.82421875, 'recall': 0.9055793991416309, 'f1': 0.8629856850715747, 'number': 233}	0.9016	0.9247	0.9130	0.9691
0.0057	10.0	4670	0.1990	{'precision': 0.8227848101265823, 'recall': 0.5603448275862069, 'f1': 0.6666666666666667, 'number': 116}	{'precision': 0.7692307692307693, 'recall': 0.8823529411764706, 'f1': 0.8219178082191781, 'number': 34}	{'precision': 0.9166666666666666, 'recall': 0.9241803278688525, 'f1': 0.9204081632653061, 'number': 488}	{'precision': 0.8413461538461539, 'recall': 0.8620689655172413, 'f1': 0.8515815085158149, 'number': 203}	{'precision': 0.9330453563714903, 'recall': 0.9840546697038725, 'f1': 0.9578713968957872, 'number': 878}	{'precision': 0.8502024291497976, 'recall': 0.9012875536480687, 'f1': 0.875, 'number': 233}	0.9016	0.9196	0.9105	0.9689

Framework versions

• Transformers 4.44.0
• Pytorch 2.3.1+cu121
• Datasets 2.19.1
• Tokenizers 0.19.1