xlm-roberta-base-ner-coin

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

• Loss: 0.0402
• Precision: 0.9865
• Recall: 0.9748
• F1: 0.9806
• Accuracy: 0.9957

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: 5e-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: 100

Training results

Training Loss	Epoch	Step	Validation Loss	Precision	Recall	F1	Accuracy
No log	1.0	26	0.0316	0.9075	0.9452	0.9260	0.9905
No log	2.0	52	0.0156	0.9762	0.9719	0.9740	0.9946
No log	3.0	78	0.0157	0.9880	0.9763	0.9821	0.9957
No log	4.0	104	0.0153	0.9894	0.9719	0.9806	0.9959
No log	5.0	130	0.0158	0.9850	0.9748	0.9799	0.9952
No log	6.0	156	0.0166	0.9909	0.9674	0.9790	0.9954
No log	7.0	182	0.0178	0.9865	0.9719	0.9791	0.9952
No log	8.0	208	0.0971	0.9260	0.8533	0.8882	0.9798
No log	9.0	234	0.2225	0.9314	0.5630	0.7018	0.9434
No log	10.0	260	0.0260	0.9740	0.9452	0.9594	0.9919
No log	11.0	286	0.0184	0.9777	0.9763	0.9770	0.9946
No log	12.0	312	0.0185	0.9910	0.9748	0.9828	0.9962
No log	13.0	338	0.0190	0.9807	0.9778	0.9792	0.9952
No log	14.0	364	0.0180	0.9791	0.9733	0.9762	0.9948
No log	15.0	390	0.0208	0.9792	0.9778	0.9785	0.9952
No log	16.0	416	0.0237	0.9851	0.9778	0.9814	0.9959
No log	17.0	442	0.0232	0.9880	0.9719	0.9798	0.9955
No log	18.0	468	0.0236	0.9895	0.9733	0.9813	0.9959
No log	19.0	494	0.0246	0.9909	0.9689	0.9798	0.9955
0.0374	20.0	520	0.0216	0.9778	0.9807	0.9793	0.9951
0.0374	21.0	546	0.0257	0.9880	0.9733	0.9806	0.9955
0.0374	22.0	572	0.0226	0.9880	0.9793	0.9836	0.9962
0.0374	23.0	598	0.0237	0.9792	0.9748	0.9770	0.9948
0.0374	24.0	624	0.0330	0.9734	0.9763	0.9749	0.9943
0.0374	25.0	650	0.0300	0.9821	0.9778	0.9800	0.9954
0.0374	26.0	676	0.0280	0.9821	0.9748	0.9784	0.9951
0.0374	27.0	702	0.0247	0.9939	0.9674	0.9805	0.9957
0.0374	28.0	728	0.0280	0.9924	0.9704	0.9813	0.9959
0.0374	29.0	754	0.0266	0.9806	0.9733	0.9770	0.9949
0.0374	30.0	780	0.0215	0.9866	0.9807	0.9837	0.9962
0.0374	31.0	806	0.0263	0.9820	0.9719	0.9769	0.9949
0.0374	32.0	832	0.0306	0.9880	0.9733	0.9806	0.9957
0.0374	33.0	858	0.0273	0.9850	0.9733	0.9791	0.9954
0.0374	34.0	884	0.0253	0.9787	0.9541	0.9662	0.9928
0.0374	35.0	910	0.0298	0.9863	0.9630	0.9745	0.9944
0.0374	36.0	936	0.0251	0.9849	0.9674	0.9761	0.9948
0.0374	37.0	962	0.0308	0.9806	0.9748	0.9777	0.9951
0.0374	38.0	988	0.0303	0.9706	0.9793	0.9749	0.9943
0.0027	39.0	1014	0.0288	0.9866	0.9793	0.9829	0.9962
0.0027	40.0	1040	0.0285	0.9792	0.9778	0.9785	0.9952
0.0027	41.0	1066	0.0316	0.9864	0.9704	0.9783	0.9952
0.0027	42.0	1092	0.0307	0.9836	0.9778	0.9807	0.9957
0.0027	43.0	1118	0.0312	0.9865	0.9763	0.9814	0.9959
0.0027	44.0	1144	0.0325	0.9880	0.9719	0.9798	0.9955
0.0027	45.0	1170	0.0330	0.9821	0.9778	0.9800	0.9955
0.0027	46.0	1196	0.0384	0.9864	0.9689	0.9776	0.9951
0.0027	47.0	1222	0.0349	0.9865	0.9748	0.9806	0.9957
0.0027	48.0	1248	0.0335	0.9836	0.9763	0.9799	0.9955
0.0027	49.0	1274	0.0319	0.9895	0.9763	0.9828	0.9962
0.0027	50.0	1300	0.0334	0.9865	0.9763	0.9814	0.9959
0.0027	51.0	1326	0.0346	0.9880	0.9763	0.9821	0.9960
0.0027	52.0	1352	0.0383	0.9821	0.9763	0.9792	0.9954
0.0027	53.0	1378	0.0354	0.9895	0.9733	0.9813	0.9959
0.0027	54.0	1404	0.0385	0.9806	0.9748	0.9777	0.9949
0.0027	55.0	1430	0.0360	0.9792	0.9763	0.9777	0.9949
0.0027	56.0	1456	0.0376	0.9821	0.9763	0.9792	0.9952
0.0027	57.0	1482	0.0367	0.9850	0.9748	0.9799	0.9955
0.0009	58.0	1508	0.0377	0.9792	0.9778	0.9785	0.9951
0.0009	59.0	1534	0.0395	0.9836	0.9748	0.9792	0.9952
0.0009	60.0	1560	0.0362	0.9851	0.9778	0.9814	0.9957
0.0009	61.0	1586	0.0303	0.9778	0.9793	0.9785	0.9951
0.0009	62.0	1612	0.0338	0.9822	0.9793	0.9807	0.9955
0.0009	63.0	1638	0.0354	0.9851	0.9793	0.9822	0.9960
0.0009	64.0	1664	0.0361	0.9836	0.9793	0.9814	0.9959
0.0009	65.0	1690	0.0366	0.9822	0.9793	0.9807	0.9957
0.0009	66.0	1716	0.0384	0.9851	0.9778	0.9814	0.9959
0.0009	67.0	1742	0.0391	0.9865	0.9763	0.9814	0.9959
0.0009	68.0	1768	0.0399	0.9851	0.9763	0.9807	0.9957
0.0009	69.0	1794	0.0396	0.9850	0.9748	0.9799	0.9955
0.0009	70.0	1820	0.0410	0.9865	0.9748	0.9806	0.9957
0.0009	71.0	1846	0.0361	0.9836	0.9748	0.9792	0.9954
0.0009	72.0	1872	0.0371	0.9880	0.9719	0.9798	0.9955
0.0009	73.0	1898	0.0365	0.9865	0.9748	0.9806	0.9957
0.0009	74.0	1924	0.0359	0.9836	0.9763	0.9799	0.9955
0.0009	75.0	1950	0.0369	0.9850	0.9748	0.9799	0.9955
0.0009	76.0	1976	0.0380	0.9850	0.9733	0.9791	0.9954
0.0005	77.0	2002	0.0382	0.9835	0.9733	0.9784	0.9952
0.0005	78.0	2028	0.0384	0.9850	0.9733	0.9791	0.9954
0.0005	79.0	2054	0.0385	0.9864	0.9704	0.9783	0.9952
0.0005	80.0	2080	0.0384	0.9865	0.9733	0.9799	0.9955
0.0005	81.0	2106	0.0375	0.9835	0.9733	0.9784	0.9952
0.0005	82.0	2132	0.0375	0.9850	0.9733	0.9791	0.9954
0.0005	83.0	2158	0.0380	0.9850	0.9733	0.9791	0.9954
0.0005	84.0	2184	0.0385	0.9850	0.9733	0.9791	0.9954
0.0005	85.0	2210	0.0387	0.9880	0.9733	0.9806	0.9957
0.0005	86.0	2236	0.0391	0.9895	0.9733	0.9813	0.9959
0.0005	87.0	2262	0.0388	0.9880	0.9748	0.9814	0.9959
0.0005	88.0	2288	0.0390	0.9880	0.9748	0.9814	0.9959
0.0005	89.0	2314	0.0391	0.9850	0.9748	0.9799	0.9955
0.0005	90.0	2340	0.0393	0.9850	0.9748	0.9799	0.9955
0.0005	91.0	2366	0.0392	0.9850	0.9748	0.9799	0.9955
0.0005	92.0	2392	0.0394	0.9850	0.9748	0.9799	0.9955
0.0005	93.0	2418	0.0394	0.9850	0.9748	0.9799	0.9955
0.0005	94.0	2444	0.0395	0.9850	0.9748	0.9799	0.9955
0.0005	95.0	2470	0.0395	0.9850	0.9748	0.9799	0.9955
0.0005	96.0	2496	0.0396	0.9850	0.9748	0.9799	0.9955
0.0003	97.0	2522	0.0398	0.9850	0.9748	0.9799	0.9955
0.0003	98.0	2548	0.0399	0.9865	0.9748	0.9806	0.9957
0.0003	99.0	2574	0.0401	0.9865	0.9748	0.9806	0.9957
0.0003	100.0	2600	0.0402	0.9865	0.9748	0.9806	0.9957

Framework versions

• Transformers 4.40.2
• Pytorch 2.1.0+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1