SignBart-KArSL02-ALL-190

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

• Loss: 0.0257
• Accuracy: 0.9928
• Precision: 0.9932
• 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: 200

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall
4.7933	1.0	95	3.8980	0.3063	0.3230	0.3063
3.6437	2.0	190	2.6826	0.6176	0.6360	0.6176
2.7596	3.0	285	1.7858	0.7706	0.7850	0.7706
2.1591	4.0	380	1.2133	0.8498	0.8756	0.8498
1.7394	5.0	475	0.8731	0.8842	0.9031	0.8842
1.4682	6.0	570	0.6466	0.9102	0.9214	0.9102
1.2723	7.0	665	0.5106	0.9335	0.9443	0.9335
1.1765	8.0	760	0.4095	0.9444	0.9505	0.9444
0.9184	9.0	855	0.3387	0.9521	0.9571	0.9521
0.9107	10.0	950	0.2809	0.9630	0.9657	0.9630
0.856	11.0	1045	0.2431	0.9630	0.9673	0.9630
0.8468	12.0	1140	0.2215	0.9623	0.9666	0.9623
0.7021	13.0	1235	0.1932	0.9702	0.9729	0.9702
0.7215	14.0	1330	0.1638	0.9757	0.9778	0.9757
0.6853	15.0	1425	0.1527	0.9761	0.9780	0.9761
0.6369	16.0	1520	0.1352	0.9781	0.9799	0.9781
0.5833	17.0	1615	0.1181	0.9825	0.9837	0.9825
0.5716	18.0	1710	0.1076	0.9825	0.9839	0.9825
0.5362	19.0	1805	0.1031	0.9792	0.9812	0.9792
0.5133	20.0	1900	0.0923	0.9820	0.9836	0.9820
0.5345	21.0	1995	0.0845	0.9834	0.9846	0.9834
0.5236	22.0	2090	0.0823	0.9845	0.9862	0.9845
0.469	23.0	2185	0.0853	0.9831	0.9847	0.9831
0.4582	24.0	2280	0.0676	0.9871	0.9879	0.9871
0.4615	25.0	2375	0.0763	0.9816	0.9835	0.9816
0.4546	26.0	2470	0.0742	0.9834	0.9849	0.9834
0.4077	27.0	2565	0.0663	0.9873	0.9885	0.9873
0.4091	28.0	2660	0.0546	0.9891	0.9898	0.9891
0.3884	29.0	2755	0.0607	0.9860	0.9871	0.9860
0.401	30.0	2850	0.0535	0.9869	0.9878	0.9869
0.4272	31.0	2945	0.0505	0.9884	0.9892	0.9884
0.3691	32.0	3040	0.0480	0.9904	0.9910	0.9904
0.311	33.0	3135	0.0519	0.9880	0.9890	0.9880
0.3506	34.0	3230	0.0433	0.9915	0.9919	0.9915
0.368	35.0	3325	0.0552	0.9856	0.9870	0.9856
0.3914	36.0	3420	0.0464	0.9888	0.9902	0.9888
0.3822	37.0	3515	0.0432	0.9901	0.9913	0.9901
0.4152	38.0	3610	0.0501	0.9882	0.9893	0.9882
0.3778	39.0	3705	0.0406	0.9906	0.9913	0.9906
0.3661	40.0	3800	0.0404	0.9901	0.9911	0.9901
0.3812	41.0	3895	0.0481	0.9871	0.9886	0.9871
0.3118	42.0	3990	0.0400	0.9912	0.9919	0.9912
0.3657	43.0	4085	0.0391	0.9880	0.9889	0.9880
0.3387	44.0	4180	0.0344	0.9921	0.9927	0.9921
0.4042	45.0	4275	0.0365	0.9917	0.9926	0.9917
0.3001	46.0	4370	0.0364	0.9910	0.9920	0.9910
0.3939	47.0	4465	0.0345	0.9904	0.9911	0.9904
0.2917	48.0	4560	0.0322	0.9923	0.9929	0.9923
0.3421	49.0	4655	0.0326	0.9910	0.9917	0.9910
0.2996	50.0	4750	0.0332	0.9915	0.9923	0.9915
0.3169	51.0	4845	0.0364	0.9908	0.9914	0.9908
0.3373	52.0	4940	0.0302	0.9917	0.9923	0.9917
0.32	53.0	5035	0.0379	0.9891	0.9898	0.9891
0.2865	54.0	5130	0.0325	0.9912	0.9921	0.9912
0.3055	55.0	5225	0.0313	0.9934	0.9942	0.9934
0.3325	56.0	5320	0.0278	0.9928	0.9933	0.9928
0.327	57.0	5415	0.0327	0.9921	0.9927	0.9921
0.3881	58.0	5510	0.0289	0.9917	0.9923	0.9917
0.2695	59.0	5605	0.0297	0.9919	0.9923	0.9919
0.293	60.0	5700	0.0269	0.9930	0.9934	0.9930
0.3138	61.0	5795	0.0309	0.9915	0.9919	0.9915
0.3099	62.0	5890	0.0276	0.9923	0.9928	0.9923
0.2932	63.0	5985	0.0262	0.9930	0.9934	0.9930
0.3432	64.0	6080	0.0283	0.9923	0.9929	0.9923
0.2889	65.0	6175	0.0255	0.9928	0.9933	0.9928
0.3217	66.0	6270	0.0282	0.9919	0.9924	0.9919
0.2416	67.0	6365	0.0271	0.9926	0.9930	0.9926
0.3496	68.0	6460	0.0296	0.9930	0.9940	0.9930
0.3211	69.0	6555	0.0263	0.9930	0.9935	0.9930
0.288	70.0	6650	0.0242	0.9939	0.9942	0.9939
0.2781	71.0	6745	0.0295	0.9910	0.9918	0.9910
0.299	72.0	6840	0.0292	0.9921	0.9926	0.9921
0.3098	73.0	6935	0.0253	0.9917	0.9922	0.9917
0.2621	74.0	7030	0.0233	0.9934	0.9938	0.9934
0.2665	75.0	7125	0.0229	0.9934	0.9937	0.9934
0.3422	76.0	7220	0.0202	0.9943	0.9946	0.9943
0.2678	77.0	7315	0.0206	0.9941	0.9944	0.9941
0.2701	78.0	7410	0.0236	0.9930	0.9934	0.9930
0.2591	79.0	7505	0.0225	0.9943	0.9946	0.9943
0.2178	80.0	7600	0.0261	0.9937	0.9940	0.9937
0.3047	81.0	7695	0.0257	0.9928	0.9932	0.9928

Framework versions

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