SignBart-KArSL03-100

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

• Loss: 0.0066
• Accuracy: 0.9988
• Precision: 0.9989
• Recall: 0.9988

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
• lr_scheduler_warmup_ratio: 0.5
• num_epochs: 250

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall
5.5682	1.0	17	4.9643	0.01	0.0009	0.01
5.5976	2.0	34	4.9425	0.0138	0.0010	0.0138
5.5604	3.0	51	4.9106	0.0138	0.0010	0.0138
5.5398	4.0	68	4.8662	0.0138	0.0010	0.0138
5.4863	5.0	85	4.8148	0.0187	0.0125	0.0187
5.4532	6.0	102	4.7630	0.0213	0.0144	0.0213
5.3646	7.0	119	4.7042	0.0187	0.0141	0.0187
5.3801	8.0	136	4.6483	0.025	0.0131	0.025
5.3212	9.0	153	4.5928	0.0288	0.0087	0.0288
5.2989	10.0	170	4.5351	0.0288	0.0122	0.0288
5.2334	11.0	187	4.4720	0.0413	0.0365	0.0413
5.1654	12.0	204	4.3960	0.0462	0.0368	0.0462
5.1226	13.0	221	4.3142	0.065	0.0480	0.065
5.1777	14.0	238	4.2301	0.065	0.0320	0.065
5.0047	15.0	255	4.1303	0.1025	0.0576	0.1025
5.0417	16.0	272	4.0285	0.11	0.0667	0.11
4.9857	17.0	289	3.9299	0.1375	0.0803	0.1375
4.8391	18.0	306	3.8083	0.1737	0.1112	0.1737
4.7713	19.0	323	3.6839	0.1913	0.1487	0.1913
4.7983	20.0	340	3.5572	0.2313	0.1761	0.2313
4.5996	21.0	357	3.4163	0.2525	0.1731	0.2525
4.5774	22.0	374	3.2754	0.3025	0.2303	0.3025
4.4857	23.0	391	3.1421	0.3	0.2480	0.3
4.3894	24.0	408	3.0070	0.3312	0.2500	0.3312
4.4367	25.0	425	2.8953	0.355	0.3118	0.355
4.2151	26.0	442	2.7584	0.4363	0.3853	0.4363
4.1855	27.0	459	2.6366	0.5062	0.4244	0.5062
3.9657	28.0	476	2.5168	0.5437	0.4651	0.5437
3.9845	29.0	493	2.3876	0.5625	0.5258	0.5625
3.7952	30.0	510	2.2810	0.5837	0.5393	0.5837
3.6446	31.0	527	2.1693	0.6212	0.6093	0.6212
3.9823	32.0	544	2.0817	0.6288	0.6316	0.6288
3.557	33.0	561	1.9832	0.7025	0.6766	0.7025
3.5253	34.0	578	1.8723	0.7238	0.7192	0.7238
3.4051	35.0	595	1.7682	0.7412	0.7456	0.7412
3.3304	36.0	612	1.6787	0.7775	0.7866	0.7775
3.2686	37.0	629	1.5950	0.8037	0.8235	0.8037
3.2883	38.0	646	1.5212	0.8125	0.8055	0.8125
3.1299	39.0	663	1.4254	0.8237	0.8378	0.8237
3.0164	40.0	680	1.3583	0.855	0.8749	0.855
2.868	41.0	697	1.2796	0.8612	0.8838	0.8612
2.9247	42.0	714	1.1999	0.8862	0.8946	0.8862
3.0729	43.0	731	1.1417	0.8838	0.8927	0.8838
2.9102	44.0	748	1.0816	0.885	0.8908	0.885
2.6118	45.0	765	1.0099	0.9062	0.9154	0.9062
2.472	46.0	782	0.9481	0.9062	0.9120	0.9062
2.6815	47.0	799	0.8991	0.9175	0.9112	0.9175
2.4457	48.0	816	0.8231	0.9263	0.9311	0.9263
2.6512	49.0	833	0.7924	0.9287	0.9280	0.9287
2.4139	50.0	850	0.7454	0.93	0.9363	0.93
2.5069	51.0	867	0.7068	0.94	0.9387	0.94
2.2432	52.0	884	0.6504	0.9425	0.9426	0.9425
2.0071	53.0	901	0.6148	0.9525	0.9508	0.9525
2.2854	54.0	918	0.5782	0.9525	0.9516	0.9525
2.198	55.0	935	0.5534	0.9563	0.9534	0.9563
1.9973	56.0	952	0.5183	0.9613	0.9589	0.9613
2.1929	57.0	969	0.4900	0.9525	0.9596	0.9525
1.8044	58.0	986	0.4470	0.9563	0.9567	0.9563
1.8836	59.0	1003	0.4243	0.9613	0.9699	0.9613
1.7381	60.0	1020	0.3983	0.9637	0.9613	0.9637
1.8321	61.0	1037	0.3704	0.9688	0.9651	0.9688
1.7215	62.0	1054	0.3463	0.9712	0.9784	0.9712
1.7982	63.0	1071	0.3288	0.97	0.9776	0.97
1.7174	64.0	1088	0.3069	0.9712	0.9751	0.9712
1.3888	65.0	1105	0.2851	0.9788	0.9837	0.9788
1.6684	66.0	1122	0.2693	0.9825	0.9868	0.9825
1.5034	67.0	1139	0.2505	0.9775	0.9816	0.9775
1.5362	68.0	1156	0.2395	0.9812	0.9847	0.9812
1.4029	69.0	1173	0.2207	0.9875	0.9899	0.9875
1.3331	70.0	1190	0.2050	0.9875	0.9885	0.9875
1.3716	71.0	1207	0.1959	0.985	0.9887	0.985
1.1504	72.0	1224	0.1834	0.9838	0.9865	0.9838
1.1814	73.0	1241	0.1710	0.9862	0.9893	0.9862
1.4892	74.0	1258	0.1619	0.9912	0.9923	0.9912
1.3333	75.0	1275	0.1526	0.9925	0.9934	0.9925
1.1588	76.0	1292	0.1466	0.9888	0.9917	0.9888
0.9779	77.0	1309	0.1299	0.995	0.9958	0.995
1.2613	78.0	1326	0.1292	0.9912	0.9931	0.9912
1.2501	79.0	1343	0.1209	0.9938	0.9943	0.9938
1.1902	80.0	1360	0.1151	0.9938	0.9947	0.9938
0.9188	81.0	1377	0.1036	0.9962	0.9969	0.9962
0.8126	82.0	1394	0.0949	0.9975	0.9978	0.9975
1.0877	83.0	1411	0.0907	0.9962	0.9967	0.9962
1.0862	84.0	1428	0.0863	0.9962	0.9967	0.9962
0.7751	85.0	1445	0.0810	0.9962	0.9967	0.9962
0.7491	86.0	1462	0.0748	0.9962	0.9967	0.9962
0.9086	87.0	1479	0.0746	0.995	0.9958	0.995
0.8638	88.0	1496	0.0654	0.9988	0.9989	0.9988
0.9267	89.0	1513	0.0637	0.9962	0.9967	0.9962
0.7683	90.0	1530	0.0621	0.9975	0.9978	0.9975
0.6818	91.0	1547	0.0600	0.9925	0.9939	0.9925
0.6814	92.0	1564	0.0523	0.9962	0.9967	0.9962
0.9308	93.0	1581	0.0514	0.9962	0.9967	0.9962
0.8256	94.0	1598	0.0494	0.9962	0.9967	0.9962
0.7316	95.0	1615	0.0449	0.9975	0.9978	0.9975
0.8574	96.0	1632	0.0453	0.9975	0.9978	0.9975
0.6799	97.0	1649	0.0410	0.9988	0.9989	0.9988
0.4571	98.0	1666	0.0390	0.9975	0.9978	0.9975
0.5149	99.0	1683	0.0347	0.9975	0.9978	0.9975
0.5335	100.0	1700	0.0327	0.9975	0.9978	0.9975
0.581	101.0	1717	0.0336	0.9975	0.9978	0.9975
0.4442	102.0	1734	0.0304	0.9988	0.9989	0.9988
0.5522	103.0	1751	0.0287	0.9975	0.9978	0.9975
0.4082	104.0	1768	0.0292	0.9975	0.9978	0.9975
0.4623	105.0	1785	0.0271	0.9975	0.9978	0.9975
0.5586	106.0	1802	0.0261	0.9975	0.9978	0.9975
0.7277	107.0	1819	0.0252	0.9975	0.9978	0.9975
0.7515	108.0	1836	0.0227	0.9988	0.9989	0.9988
0.5521	109.0	1853	0.0214	0.9988	0.9989	0.9988
0.33	110.0	1870	0.0204	0.9988	0.9989	0.9988
0.4944	111.0	1887	0.0190	0.9988	0.9989	0.9988
0.6165	112.0	1904	0.0187	0.9975	0.9978	0.9975
0.3288	113.0	1921	0.0176	0.9975	0.9978	0.9975
0.3835	114.0	1938	0.0155	1.0	1.0	1.0
0.4359	115.0	1955	0.0157	1.0	1.0	1.0
0.3827	116.0	1972	0.0141	1.0	1.0	1.0
0.3317	117.0	1989	0.0146	1.0	1.0	1.0
0.295	118.0	2006	0.0155	0.9988	0.9989	0.9988
0.2978	119.0	2023	0.0147	1.0	1.0	1.0
0.4484	120.0	2040	0.0119	0.9988	0.9989	0.9988
0.2444	121.0	2057	0.0120	0.9988	0.9989	0.9988
0.3408	122.0	2074	0.0109	1.0	1.0	1.0
0.365	123.0	2091	0.0107	0.9988	0.9989	0.9988
0.4152	124.0	2108	0.0099	0.9988	0.9989	0.9988
0.2228	125.0	2125	0.0098	1.0	1.0	1.0
0.2466	126.0	2142	0.0089	1.0	1.0	1.0
0.4605	127.0	2159	0.0094	1.0	1.0	1.0
0.2572	128.0	2176	0.0083	1.0	1.0	1.0
0.2759	129.0	2193	0.0087	0.9988	0.9989	0.9988
0.4362	130.0	2210	0.0080	1.0	1.0	1.0
0.3868	131.0	2227	0.0077	1.0	1.0	1.0
0.3904	132.0	2244	0.0075	1.0	1.0	1.0
0.3206	133.0	2261	0.0080	0.9988	0.9989	0.9988
0.1986	134.0	2278	0.0064	1.0	1.0	1.0
0.2567	135.0	2295	0.0071	1.0	1.0	1.0
0.2234	136.0	2312	0.0060	1.0	1.0	1.0
0.334	137.0	2329	0.0070	0.9988	0.9989	0.9988
0.1759	138.0	2346	0.0060	1.0	1.0	1.0
0.3446	139.0	2363	0.0066	0.9988	0.9989	0.9988

Framework versions

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