roberta-large-ner-ghtk-cs-6-label-new-data-3090-14Sep-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: 0.1772
• Tk: {'precision': 0.7345132743362832, 'recall': 0.7155172413793104, 'f1': 0.7248908296943231, 'number': 116}
• Gày: {'precision': 0.7380952380952381, 'recall': 0.9117647058823529, 'f1': 0.8157894736842106, 'number': 34}
• Gày trừu tượng: {'precision': 0.9118852459016393, 'recall': 0.9118852459016393, 'f1': 0.9118852459016393, 'number': 488}
• Ã đơn: {'precision': 0.8514851485148515, 'recall': 0.8472906403940886, 'f1': 0.8493827160493828, 'number': 203}
• Đt: {'precision': 0.9291084854994629, 'recall': 0.9851936218678815, 'f1': 0.9563294637921502, 'number': 878}
• Đt trừu tượng: {'precision': 0.8259109311740891, 'recall': 0.8755364806866953, 'f1': 0.85, 'number': 233}
• Overall Precision: 0.8898
• Overall Recall: 0.9221
• Overall F1: 0.9057
• Overall Accuracy: 0.9665

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	0.1542	{'precision': 0.0, 'recall': 0.0, 'f1': 0.0, 'number': 116}	{'precision': 0.5384615384615384, 'recall': 0.8235294117647058, 'f1': 0.6511627906976744, 'number': 34}	{'precision': 0.7964285714285714, 'recall': 0.9139344262295082, 'f1': 0.8511450381679388, 'number': 488}	{'precision': 0.8045977011494253, 'recall': 0.6896551724137931, 'f1': 0.7427055702917771, 'number': 203}	{'precision': 0.8341323106423778, 'recall': 0.9908883826879271, 'f1': 0.9057782404997397, 'number': 878}	{'precision': 0.7639484978540773, 'recall': 0.7639484978540773, 'f1': 0.7639484978540771, 'number': 233}	0.8021	0.8514	0.8260	0.9463
0.263	2.0	934	0.1126	{'precision': 0.6966292134831461, 'recall': 0.5344827586206896, 'f1': 0.6048780487804878, 'number': 116}	{'precision': 0.6041666666666666, 'recall': 0.8529411764705882, 'f1': 0.7073170731707317, 'number': 34}	{'precision': 0.8535645472061657, 'recall': 0.9077868852459017, 'f1': 0.8798411122144986, 'number': 488}	{'precision': 0.9135802469135802, 'recall': 0.729064039408867, 'f1': 0.810958904109589, 'number': 203}	{'precision': 0.9209694415173867, 'recall': 0.9954441913439636, 'f1': 0.9567597153804049, 'number': 878}	{'precision': 0.7540983606557377, 'recall': 0.7896995708154506, 'f1': 0.7714884696016772, 'number': 233}	0.8652	0.8914	0.8781	0.9574
0.1028	3.0	1401	0.1177	{'precision': 0.868421052631579, 'recall': 0.5689655172413793, 'f1': 0.6875000000000001, 'number': 116}	{'precision': 0.6122448979591837, 'recall': 0.8823529411764706, 'f1': 0.7228915662650602, 'number': 34}	{'precision': 0.8669275929549902, 'recall': 0.9077868852459017, 'f1': 0.8868868868868869, 'number': 488}	{'precision': 0.8507462686567164, 'recall': 0.8423645320197044, 'f1': 0.8465346534653465, 'number': 203}	{'precision': 0.9059561128526645, 'recall': 0.9874715261958997, 'f1': 0.9449591280653951, 'number': 878}	{'precision': 0.8034188034188035, 'recall': 0.8068669527896996, 'f1': 0.8051391862955032, 'number': 233}	0.8703	0.9042	0.8869	0.9600
0.077	4.0	1868	0.1431	{'precision': 0.8666666666666667, 'recall': 0.33620689655172414, 'f1': 0.484472049689441, 'number': 116}	{'precision': 0.7435897435897436, 'recall': 0.8529411764705882, 'f1': 0.7945205479452054, 'number': 34}	{'precision': 0.87109375, 'recall': 0.9139344262295082, 'f1': 0.892, 'number': 488}	{'precision': 0.8972972972972973, 'recall': 0.8177339901477833, 'f1': 0.8556701030927836, 'number': 203}	{'precision': 0.8969072164948454, 'recall': 0.9908883826879271, 'f1': 0.9415584415584417, 'number': 878}	{'precision': 0.6175637393767706, 'recall': 0.9356223175965666, 'f1': 0.7440273037542662, 'number': 233}	0.8403	0.9057	0.8718	0.9591
0.053	5.0	2335	0.1367	{'precision': 0.7065217391304348, 'recall': 0.5603448275862069, 'f1': 0.625, 'number': 116}	{'precision': 0.8181818181818182, 'recall': 0.7941176470588235, 'f1': 0.8059701492537314, 'number': 34}	{'precision': 0.8993963782696177, 'recall': 0.9159836065573771, 'f1': 0.9076142131979696, 'number': 488}	{'precision': 0.8440860215053764, 'recall': 0.7733990147783252, 'f1': 0.8071979434447302, 'number': 203}	{'precision': 0.9205508474576272, 'recall': 0.989749430523918, 'f1': 0.9538968166849615, 'number': 878}	{'precision': 0.7114093959731543, 'recall': 0.9098712446351931, 'f1': 0.7984934086629002, 'number': 233}	0.8668	0.9103	0.8881	0.9625
0.0404	6.0	2802	0.1269	{'precision': 0.7959183673469388, 'recall': 0.6724137931034483, 'f1': 0.7289719626168225, 'number': 116}	{'precision': 0.7209302325581395, 'recall': 0.9117647058823529, 'f1': 0.8051948051948051, 'number': 34}	{'precision': 0.9168399168399168, 'recall': 0.9036885245901639, 'f1': 0.9102167182662538, 'number': 488}	{'precision': 0.8967391304347826, 'recall': 0.812807881773399, 'f1': 0.8527131782945736, 'number': 203}	{'precision': 0.9555555555555556, 'recall': 0.979498861047836, 'f1': 0.9673790776152982, 'number': 878}	{'precision': 0.7619047619047619, 'recall': 0.8927038626609443, 'f1': 0.8221343873517787, 'number': 233}	0.9010	0.9134	0.9071	0.9664
0.0232	7.0	3269	0.1361	{'precision': 0.7818181818181819, 'recall': 0.7413793103448276, 'f1': 0.7610619469026548, 'number': 116}	{'precision': 0.7209302325581395, 'recall': 0.9117647058823529, 'f1': 0.8051948051948051, 'number': 34}	{'precision': 0.896, 'recall': 0.9180327868852459, 'f1': 0.9068825910931175, 'number': 488}	{'precision': 0.875, 'recall': 0.8620689655172413, 'f1': 0.8684863523573201, 'number': 203}	{'precision': 0.9505494505494505, 'recall': 0.9851936218678815, 'f1': 0.9675615212527963, 'number': 878}	{'precision': 0.8504273504273504, 'recall': 0.8540772532188842, 'f1': 0.8522483940042828, 'number': 233}	0.9034	0.9242	0.9136	0.9693
0.0192	8.0	3736	0.1610	{'precision': 0.7478991596638656, 'recall': 0.7672413793103449, 'f1': 0.7574468085106383, 'number': 116}	{'precision': 0.7435897435897436, 'recall': 0.8529411764705882, 'f1': 0.7945205479452054, 'number': 34}	{'precision': 0.9102040816326531, 'recall': 0.9139344262295082, 'f1': 0.9120654396728016, 'number': 488}	{'precision': 0.8442211055276382, 'recall': 0.8275862068965517, 'f1': 0.8358208955223881, 'number': 203}	{'precision': 0.9361471861471862, 'recall': 0.9851936218678815, 'f1': 0.9600443951165373, 'number': 878}	{'precision': 0.8326359832635983, 'recall': 0.8540772532188842, 'f1': 0.8432203389830507, 'number': 233}	0.8935	0.9201	0.9066	0.9661
0.01	9.0	4203	0.1725	{'precision': 0.7368421052631579, 'recall': 0.603448275862069, 'f1': 0.6635071090047393, 'number': 116}	{'precision': 0.7045454545454546, 'recall': 0.9117647058823529, 'f1': 0.794871794871795, 'number': 34}	{'precision': 0.9087221095334685, 'recall': 0.9180327868852459, 'f1': 0.9133537206931702, 'number': 488}	{'precision': 0.8613861386138614, 'recall': 0.8571428571428571, 'f1': 0.8592592592592593, 'number': 203}	{'precision': 0.9261241970021413, 'recall': 0.9851936218678815, 'f1': 0.9547461368653422, 'number': 878}	{'precision': 0.8326359832635983, 'recall': 0.8540772532188842, 'f1': 0.8432203389830507, 'number': 233}	0.8904	0.9155	0.9028	0.9658
0.0082	10.0	4670	0.1772	{'precision': 0.7345132743362832, 'recall': 0.7155172413793104, 'f1': 0.7248908296943231, 'number': 116}	{'precision': 0.7380952380952381, 'recall': 0.9117647058823529, 'f1': 0.8157894736842106, 'number': 34}	{'precision': 0.9118852459016393, 'recall': 0.9118852459016393, 'f1': 0.9118852459016393, 'number': 488}	{'precision': 0.8514851485148515, 'recall': 0.8472906403940886, 'f1': 0.8493827160493828, 'number': 203}	{'precision': 0.9291084854994629, 'recall': 0.9851936218678815, 'f1': 0.9563294637921502, 'number': 878}	{'precision': 0.8259109311740891, 'recall': 0.8755364806866953, 'f1': 0.85, 'number': 233}	0.8898	0.9221	0.9057	0.9665

Framework versions

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