SignBart-KArSL02-502

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

• Loss: 0.0288
• Accuracy: 0.9928
• Precision: 0.9937
• Recall: 0.9928

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: 0.0002
• train_batch_size: 256
• eval_batch_size: 256
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: cosine
• num_epochs: 1000

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall
6.5751	1.0	84	5.6281	0.0229	0.0143	0.0229
5.9133	2.0	168	4.7280	0.1113	0.0728	0.1113
5.2344	3.0	252	3.9383	0.3008	0.2830	0.3008
4.6682	4.0	336	3.2614	0.4771	0.4991	0.4771
4.1218	5.0	420	2.6372	0.6190	0.6422	0.6190
3.7013	6.0	504	2.1624	0.6887	0.7085	0.6887
3.2224	7.0	588	1.6954	0.7734	0.8037	0.7734
2.8422	8.0	672	1.3277	0.8277	0.8571	0.8277
2.5199	9.0	756	1.0722	0.8653	0.8867	0.8653
2.1217	10.0	840	0.8216	0.8989	0.9189	0.8989
2.0435	11.0	924	0.6716	0.9153	0.9296	0.9153
1.8002	12.0	1008	0.5558	0.9260	0.9397	0.9260
1.5256	13.0	1092	0.4512	0.9353	0.9495	0.9353
1.4743	14.0	1176	0.3772	0.9490	0.9592	0.9490
1.3304	15.0	1260	0.3114	0.9557	0.9666	0.9557
1.2249	16.0	1344	0.2686	0.9634	0.9722	0.9634
1.0887	17.0	1428	0.2315	0.9626	0.9691	0.9626
0.9725	18.0	1512	0.2024	0.9674	0.9746	0.9674
0.9389	19.0	1596	0.1752	0.9741	0.9800	0.9741
0.9236	20.0	1680	0.1593	0.9736	0.9775	0.9736
0.8081	21.0	1764	0.1443	0.9771	0.9813	0.9771
0.7524	22.0	1848	0.1310	0.9796	0.9835	0.9796
0.7976	23.0	1932	0.1217	0.9781	0.9825	0.9781
0.675	24.0	2016	0.1111	0.9808	0.9855	0.9808
0.6981	25.0	2100	0.1043	0.9783	0.9818	0.9783
0.654	26.0	2184	0.0927	0.9846	0.9872	0.9846
0.5625	27.0	2268	0.0896	0.9838	0.9872	0.9838
0.6833	28.0	2352	0.0803	0.9868	0.9891	0.9868
0.5077	29.0	2436	0.0716	0.9846	0.9880	0.9846
0.553	30.0	2520	0.0733	0.9863	0.9887	0.9863
0.4878	31.0	2604	0.0660	0.9880	0.9902	0.9880
0.4306	32.0	2688	0.0613	0.9893	0.9907	0.9893
0.4244	33.0	2772	0.0606	0.9900	0.9917	0.9900
0.4809	34.0	2856	0.0575	0.9875	0.9899	0.9875
0.4619	35.0	2940	0.0525	0.9898	0.9918	0.9898
0.4083	36.0	3024	0.0512	0.9903	0.9920	0.9903
0.36	37.0	3108	0.0523	0.9885	0.9900	0.9885
0.3592	38.0	3192	0.0500	0.9905	0.9922	0.9905
0.3933	39.0	3276	0.0505	0.9898	0.9918	0.9898
0.3261	40.0	3360	0.0445	0.9898	0.9915	0.9898
0.333	41.0	3444	0.0467	0.9875	0.9902	0.9875
0.3892	42.0	3528	0.0408	0.9918	0.9928	0.9918
0.3201	43.0	3612	0.0400	0.9923	0.9936	0.9923
0.415	44.0	3696	0.0444	0.9900	0.9920	0.9900
0.3712	45.0	3780	0.0371	0.9928	0.9938	0.9928
0.3379	46.0	3864	0.0371	0.9915	0.9930	0.9915
0.3076	47.0	3948	0.0395	0.9925	0.9938	0.9925
0.3687	48.0	4032	0.0372	0.9920	0.9934	0.9920
0.3415	49.0	4116	0.0344	0.9935	0.9945	0.9935
0.2651	50.0	4200	0.0364	0.9925	0.9935	0.9925
0.2858	51.0	4284	0.0349	0.9908	0.9921	0.9908
0.2511	52.0	4368	0.0322	0.9930	0.9942	0.9930
0.2788	53.0	4452	0.0319	0.9933	0.9941	0.9933
0.3445	54.0	4536	0.0325	0.9933	0.9943	0.9933
0.2614	55.0	4620	0.0308	0.9943	0.9950	0.9943
0.2448	56.0	4704	0.0322	0.9940	0.9949	0.9940
0.2548	57.0	4788	0.0307	0.9918	0.9932	0.9918
0.2499	58.0	4872	0.0297	0.9928	0.9941	0.9928
0.3037	59.0	4956	0.0284	0.9933	0.9944	0.9933
0.2937	60.0	5040	0.0303	0.9933	0.9942	0.9933
0.2379	61.0	5124	0.0265	0.9948	0.9954	0.9948
0.279	62.0	5208	0.0304	0.9925	0.9936	0.9925
0.3219	63.0	5292	0.0271	0.9940	0.9949	0.9940
0.2348	64.0	5376	0.0284	0.9943	0.9950	0.9943
0.3779	65.0	5460	0.0285	0.9930	0.9938	0.9930
0.185	66.0	5544	0.0288	0.9928	0.9937	0.9928

Framework versions

• Transformers 4.44.2
• Pytorch 2.4.1+cu121
• Datasets 3.2.0
• Tokenizers 0.19.1