XMLRoberta_Lexical_Dataset59KBoDuoi

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: 0.6232
• Accuracy: 0.8988
• F1: 0.8992

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	0.2558	200	0.4593	0.7916	0.7880
No log	0.5115	400	0.3779	0.8222	0.8249
No log	0.7673	600	0.3462	0.8514	0.8497
0.4345	1.0230	800	0.3543	0.8554	0.8513
0.4345	1.2788	1000	0.3504	0.8573	0.8537
0.4345	1.5345	1200	0.3033	0.8767	0.8772
0.4345	1.7903	1400	0.2834	0.8778	0.8788
0.3071	2.0460	1600	0.3207	0.8671	0.8695
0.3071	2.3018	1800	0.2959	0.8822	0.8814
0.3071	2.5575	2000	0.2821	0.8778	0.8781
0.3071	2.8133	2200	0.3024	0.8872	0.8883
0.2523	3.0691	2400	0.2972	0.8888	0.8894
0.2523	3.3248	2600	0.2746	0.8883	0.8891
0.2523	3.5806	2800	0.2828	0.8909	0.8911
0.2523	3.8363	3000	0.2822	0.8941	0.8941
0.2177	4.0921	3200	0.2995	0.8898	0.8910
0.2177	4.3478	3400	0.2953	0.8887	0.8898
0.2177	4.6036	3600	0.2944	0.8925	0.8931
0.2177	4.8593	3800	0.3006	0.8957	0.8958
0.189	5.1151	4000	0.2816	0.8950	0.8955
0.189	5.3708	4200	0.2865	0.8956	0.8960
0.189	5.6266	4400	0.2794	0.8961	0.8966
0.189	5.8824	4600	0.2836	0.8980	0.8986
0.1637	6.1381	4800	0.3399	0.8949	0.8951
0.1637	6.3939	5000	0.3248	0.8952	0.8957
0.1637	6.6496	5200	0.3341	0.8976	0.8979
0.1637	6.9054	5400	0.2993	0.8962	0.8970
0.1388	7.1611	5600	0.3662	0.8967	0.8978
0.1388	7.4169	5800	0.3761	0.8962	0.8968
0.1388	7.6726	6000	0.3305	0.8953	0.8961
0.1388	7.9284	6200	0.3328	0.8966	0.8970
0.1193	8.1841	6400	0.3753	0.8980	0.8985
0.1193	8.4399	6600	0.3646	0.8974	0.8976
0.1193	8.6957	6800	0.3800	0.8963	0.8966
0.1193	8.9514	7000	0.3472	0.8980	0.8987
0.1059	9.2072	7200	0.3991	0.9002	0.9004
0.1059	9.4629	7400	0.4026	0.8967	0.8978
0.1059	9.7187	7600	0.3915	0.8983	0.8983
0.1059	9.9744	7800	0.3932	0.8997	0.8999
0.0923	10.2302	8000	0.4887	0.8939	0.8947
0.0923	10.4859	8200	0.4074	0.8977	0.8981
0.0923	10.7417	8400	0.3931	0.8998	0.9003
0.0806	10.9974	8600	0.4131	0.8955	0.8964
0.0806	11.2532	8800	0.4499	0.8963	0.8970
0.0806	11.5090	9000	0.4436	0.8999	0.9002
0.0806	11.7647	9200	0.4842	0.8965	0.8968
0.0697	12.0205	9400	0.4851	0.8961	0.8963
0.0697	12.2762	9600	0.5138	0.8999	0.9002
0.0697	12.5320	9800	0.5020	0.8963	0.8964
0.0697	12.7877	10000	0.5108	0.8929	0.8940
0.064	13.0435	10200	0.4893	0.8966	0.8968
0.064	13.2992	10400	0.5052	0.8973	0.8980
0.064	13.5550	10600	0.4917	0.8970	0.8971
0.064	13.8107	10800	0.5087	0.8965	0.8968
0.0571	14.0665	11000	0.5195	0.8970	0.8977
0.0571	14.3223	11200	0.5279	0.8932	0.8943
0.0571	14.5780	11400	0.5015	0.8974	0.8978
0.0571	14.8338	11600	0.5301	0.8961	0.8965
0.0538	15.0895	11800	0.5297	0.8951	0.8952
0.0538	15.3453	12000	0.5573	0.8976	0.8980
0.0538	15.6010	12200	0.5579	0.8955	0.8962
0.0538	15.8568	12400	0.5814	0.8969	0.8968
0.0481	16.1125	12600	0.5861	0.8972	0.8974
0.0481	16.3683	12800	0.5871	0.8968	0.8972
0.0481	16.6240	13000	0.5913	0.8978	0.8986
0.0481	16.8798	13200	0.6100	0.8957	0.8967
0.043	17.1355	13400	0.5895	0.8976	0.8982
0.043	17.3913	13600	0.5653	0.8978	0.8982
0.043	17.6471	13800	0.5914	0.8996	0.8999
0.043	17.9028	14000	0.5850	0.9005	0.9007
0.042	18.1586	14200	0.5927	0.8983	0.8988
0.042	18.4143	14400	0.6164	0.8997	0.8999
0.042	18.6701	14600	0.6324	0.8986	0.8992
0.042	18.9258	14800	0.6097	0.8996	0.9001
0.0383	19.1816	15000	0.6029	0.8985	0.8989
0.0383	19.4373	15200	0.6067	0.8988	0.8992
0.0383	19.6931	15400	0.6177	0.8987	0.8991
0.0383	19.9488	15600	0.6232	0.8988	0.8992

Framework versions

• Transformers 4.41.2
• Pytorch 2.1.2
• Datasets 2.19.2
• Tokenizers 0.19.1