XLMRoBERTa_Lexical_Dataset45K

This model is a fine-tuned version of FacebookAI/xlm-roberta-base on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4964
• Accuracy: 0.8870
• F1: 0.8870

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: 2e-05
• train_batch_size: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 15

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.2841	200	0.4614	0.7988	0.7972
No log	0.5682	400	0.3716	0.8339	0.8335
No log	0.8523	600	0.3603	0.8456	0.8474
0.455	1.1364	800	0.3298	0.8618	0.8612
0.455	1.4205	1000	0.3466	0.8621	0.8598
0.455	1.7045	1200	0.3051	0.8697	0.8712
0.455	1.9886	1400	0.2924	0.8773	0.8774
0.3111	2.2727	1600	0.2764	0.8847	0.8853
0.3111	2.5568	1800	0.2648	0.8865	0.8866
0.3111	2.8409	2000	0.2571	0.8890	0.8894
0.2536	3.125	2200	0.2745	0.8901	0.8898
0.2536	3.4091	2400	0.2801	0.8729	0.8747
0.2536	3.6932	2600	0.2863	0.8892	0.8885
0.2536	3.9773	2800	0.2587	0.8909	0.8902
0.2249	4.2614	3000	0.2728	0.8905	0.8903
0.2249	4.5455	3200	0.2782	0.8873	0.8882
0.2249	4.8295	3400	0.2696	0.8900	0.8908
0.1984	5.1136	3600	0.3226	0.8896	0.8890
0.1984	5.3977	3800	0.2821	0.8921	0.8923
0.1984	5.6818	4000	0.3206	0.8818	0.8830
0.1984	5.9659	4200	0.2800	0.8938	0.8936
0.178	6.25	4400	0.2892	0.8938	0.8938
0.178	6.5341	4600	0.3262	0.8875	0.8880
0.178	6.8182	4800	0.3199	0.8876	0.8874
0.1556	7.1023	5000	0.3237	0.8889	0.8892
0.1556	7.3864	5200	0.3540	0.8883	0.8880
0.1556	7.6705	5400	0.3361	0.8925	0.8925
0.1556	7.9545	5600	0.3369	0.8868	0.8877
0.1344	8.2386	5800	0.3449	0.8889	0.8887
0.1344	8.5227	6000	0.3813	0.8888	0.8884
0.1344	8.8068	6200	0.3463	0.8908	0.8910
0.1203	9.0909	6400	0.3781	0.8886	0.8893
0.1203	9.375	6600	0.3825	0.8897	0.8897
0.1203	9.6591	6800	0.3842	0.8890	0.8895
0.1203	9.9432	7000	0.3894	0.8885	0.8893
0.1069	10.2273	7200	0.4063	0.8903	0.8900
0.1069	10.5114	7400	0.3976	0.8908	0.8912
0.1069	10.7955	7600	0.4013	0.8892	0.8893
0.0942	11.0795	7800	0.4663	0.8888	0.8886
0.0942	11.3636	8000	0.4394	0.8867	0.8871
0.0942	11.6477	8200	0.4770	0.8852	0.8859
0.0942	11.9318	8400	0.4278	0.8881	0.8881
0.0827	12.2159	8600	0.4431	0.8874	0.8877
0.0827	12.5	8800	0.4553	0.8851	0.8856
0.0827	12.7841	9000	0.4546	0.8859	0.8861
0.0772	13.0682	9200	0.4571	0.8883	0.8881
0.0772	13.3523	9400	0.4655	0.8872	0.8870
0.0772	13.6364	9600	0.4700	0.8872	0.8876
0.0772	13.9205	9800	0.4817	0.8863	0.8867
0.0692	14.2045	10000	0.4838	0.8864	0.8868
0.0692	14.4886	10200	0.4825	0.8868	0.8871
0.0692	14.7727	10400	0.4964	0.8870	0.8870

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.2
• Datasets 2.20.0
• Tokenizers 0.19.1