SignBart-WLASL-100

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

• Loss: 0.8875
• Accuracy: 0.78
• Precision: 0.7083
• Recall: 0.78

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: Use adamw_torch with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: cosine
• num_epochs: 1000

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss	Precision	Recall
4.6294	1.0	6	0.01	4.6423	0.0029	0.01
4.6221	2.0	12	0.02	4.6363	0.0111	0.02
4.6023	3.0	18	0.01	4.6306	0.004	0.01
4.5958	4.0	24	0.01	4.6252	0.0033	0.01
4.5609	5.0	30	0.01	4.6195	0.005	0.01
4.548	6.0	36	0.0	4.6120	0.0	0.0
4.5145	7.0	42	0.0	4.6014	0.0	0.0
4.4954	8.0	48	0.005	4.5863	0.0009	0.005
4.4841	9.0	54	0.005	4.5616	0.0011	0.005
4.4363	10.0	60	0.035	4.5295	0.0103	0.035
4.4035	11.0	66	0.045	4.4939	0.0063	0.045
4.3784	12.0	72	0.045	4.4568	0.0074	0.045
4.3311	13.0	78	0.045	4.4236	0.0053	0.045
4.3205	14.0	84	0.045	4.3911	0.0053	0.045
4.2704	15.0	90	0.045	4.3577	0.0049	0.045
4.2605	16.0	96	0.045	4.3248	0.0042	0.045
4.1903	17.0	102	0.05	4.2867	0.0151	0.05
4.1656	18.0	108	0.05	4.2495	0.0091	0.05
4.1486	19.0	114	0.055	4.2056	0.0190	0.055
4.1166	20.0	120	0.07	4.1630	0.0262	0.07
4.0616	21.0	126	0.065	4.1183	0.0254	0.065
4.072	22.0	132	0.095	4.0716	0.0387	0.095
3.9844	23.0	138	0.125	4.0388	0.0640	0.125
3.9655	24.0	144	0.1	3.9950	0.0346	0.1
3.9254	25.0	150	0.135	3.9532	0.0573	0.135
3.9579	26.0	156	0.15	3.9087	0.0719	0.15
3.8965	27.0	162	0.155	3.8768	0.0835	0.155
3.843	28.0	168	0.16	3.8324	0.0905	0.16
3.7373	29.0	174	0.15	3.8015	0.0759	0.15
3.7614	30.0	180	0.175	3.7630	0.0837	0.175
3.6968	31.0	186	0.215	3.7246	0.1210	0.215
3.7135	32.0	192	0.215	3.6867	0.1268	0.215
3.6779	33.0	198	0.22	3.6611	0.1250	0.22
3.6548	34.0	204	0.245	3.6102	0.1357	0.245
3.5786	35.0	210	0.245	3.5803	0.1312	0.245
3.5998	36.0	216	0.22	3.5474	0.1240	0.22
3.5238	37.0	222	0.27	3.5202	0.1498	0.27
3.4736	38.0	228	0.26	3.4935	0.1346	0.26
3.4239	39.0	234	0.235	3.4449	0.1225	0.235
3.4223	40.0	240	0.3	3.4191	0.1804	0.3
3.4023	41.0	246	0.245	3.3846	0.1462	0.245
3.4277	42.0	252	0.29	3.3555	0.1728	0.29
3.501	43.0	258	0.28	3.3154	0.1822	0.28
3.2925	44.0	264	0.265	3.3009	0.1623	0.265
3.3913	45.0	270	0.28	3.2685	0.1769	0.28
3.2456	46.0	276	0.3	3.2331	0.1872	0.3
3.182	47.0	282	0.315	3.2161	0.2002	0.315
3.1807	48.0	288	0.33	3.1669	0.2145	0.33
3.1399	49.0	294	0.32	3.1431	0.2029	0.32
3.3169	50.0	300	0.32	3.1204	0.1982	0.32
3.1367	51.0	306	0.31	3.0864	0.1975	0.31
3.2824	52.0	312	0.325	3.0645	0.1887	0.325
3.1501	53.0	318	0.32	3.0479	0.2120	0.32
3.0395	54.0	324	0.33	3.0220	0.2164	0.33
3.0884	55.0	330	0.33	2.9982	0.2120	0.33
3.0406	56.0	336	0.335	2.9811	0.2142	0.335
2.9801	57.0	342	0.34	2.9381	0.2130	0.34
2.9503	58.0	348	0.375	2.9224	0.2583	0.375
3.099	59.0	354	0.36	2.9058	0.2277	0.36
2.8996	60.0	360	0.365	2.8744	0.2436	0.365
2.9776	61.0	366	0.385	2.8375	0.2569	0.385
2.8564	62.0	372	0.38	2.8375	0.2617	0.38
2.9198	63.0	378	0.385	2.7962	0.2676	0.385
2.9856	64.0	384	0.405	2.7725	0.2871	0.405
2.8922	65.0	390	0.415	2.7818	0.3039	0.415
2.88	66.0	396	0.395	2.7437	0.2785	0.395
2.83	67.0	402	0.42	2.7202	0.2915	0.42
2.8292	68.0	408	0.41	2.7009	0.3004	0.41
2.7326	69.0	414	0.445	2.6925	0.3229	0.445
2.7876	70.0	420	0.42	2.6812	0.3152	0.42
2.9079	71.0	426	0.435	2.6309	0.3187	0.435
2.598	72.0	432	0.49	2.6130	0.3724	0.49
2.733	73.0	438	0.475	2.5932	0.3665	0.475
2.6207	74.0	444	0.445	2.5658	0.3417	0.445
2.5644	75.0	450	0.46	2.5580	0.3519	0.46
2.5451	76.0	456	0.475	2.5283	0.3655	0.475
2.5322	77.0	462	0.505	2.5012	0.3910	0.505
2.6572	78.0	468	0.485	2.4966	0.3902	0.485
2.6936	79.0	474	0.5	2.4734	0.3832	0.5
2.4608	80.0	480	0.5	2.4425	0.3827	0.5
2.4683	81.0	486	0.495	2.4316	0.3745	0.495
2.5201	82.0	492	0.485	2.3955	0.3630	0.485
2.7546	83.0	498	0.5	2.3787	0.3845	0.5
2.7336	84.0	504	0.5	2.3700	0.3785	0.5
2.3704	85.0	510	0.515	2.3543	0.3982	0.515
2.5186	86.0	516	0.51	2.3338	0.3955	0.51
2.3039	87.0	522	0.51	2.2979	0.3994	0.51
2.3963	88.0	528	0.525	2.3010	0.3942	0.525
2.3421	89.0	534	0.51	2.2681	0.3871	0.51
2.2704	90.0	540	0.515	2.2597	0.3839	0.515
2.4131	91.0	546	0.535	2.2290	0.4152	0.535
2.3811	92.0	552	0.535	2.2067	0.4024	0.535
2.3056	93.0	558	0.585	2.1941	0.4653	0.585
2.3986	94.0	564	0.545	2.1842	0.4210	0.545
2.2206	95.0	570	0.56	2.1725	0.4255	0.56
2.2062	96.0	576	0.59	2.1497	0.4707	0.59
2.2706	97.0	582	0.545	2.1617	0.4193	0.545
2.3096	98.0	588	0.56	2.1203	0.455	0.56
2.2011	99.0	594	0.555	2.1191	0.4378	0.555
2.1684	100.0	600	0.565	2.0828	0.4528	0.565
2.4284	101.0	606	0.585	2.1053	0.477	0.585
2.1035	102.0	612	0.55	2.0606	0.4334	0.55
2.0871	103.0	618	0.59	2.0717	0.4759	0.59
2.0805	104.0	624	0.595	2.0255	0.4874	0.595
2.1787	105.0	630	0.545	2.0744	0.4260	0.545
2.0334	106.0	636	0.595	2.0002	0.485	0.595
1.9427	107.0	642	0.59	2.0116	0.4725	0.59
2.382	108.0	648	0.62	1.9687	0.5113	0.62
2.308	109.0	654	0.575	1.9729	0.4480	0.575
2.0161	110.0	660	0.59	1.9531	0.4927	0.59
2.082	111.0	666	0.63	1.9344	0.5343	0.63
2.2083	112.0	672	0.635	1.9182	0.5460	0.635
1.9757	113.0	678	0.615	1.9075	0.5253	0.615
2.128	114.0	684	0.625	1.8839	0.5355	0.625
2.1757	115.0	690	0.635	1.8928	0.5117	0.635
2.1565	116.0	696	0.625	1.8499	0.5327	0.625
1.8744	117.0	702	0.67	1.8631	0.5789	0.67
2.0423	118.0	708	0.635	1.8490	0.5437	0.635
1.8725	119.0	714	0.665	1.8267	0.5670	0.665
1.8861	120.0	720	0.665	1.8062	0.571	0.665
2.0352	121.0	726	0.65	1.7875	0.5620	0.65
1.8213	122.0	732	0.64	1.7902	0.5413	0.64
1.8251	123.0	738	0.65	1.7903	0.5557	0.65
1.8493	124.0	744	0.68	1.7514	0.5873	0.68
1.6917	125.0	750	0.695	1.7363	0.615	0.695
1.7612	126.0	756	0.675	1.7167	0.584	0.675
1.8076	127.0	762	0.665	1.7312	0.568	0.665
2.414	128.0	768	0.66	1.6981	0.5707	0.66
2.1288	129.0	774	0.71	1.6856	0.6313	0.71
1.9291	130.0	780	0.66	1.7015	0.5687	0.66
1.8468	131.0	786	0.705	1.6648	0.623	0.705
2.0147	132.0	792	0.705	1.6471	0.6259	0.705
1.9928	133.0	798	0.725	1.6293	0.638	0.725
1.8756	134.0	804	0.71	1.6247	0.6297	0.71
1.9056	135.0	810	0.675	1.6222	0.5733	0.675
1.7642	136.0	816	0.695	1.6210	0.6169	0.695
1.7536	137.0	822	0.675	1.6224	0.5792	0.675
1.8607	138.0	828	0.71	1.5810	0.6292	0.71
1.562	139.0	834	0.71	1.5714	0.6257	0.71
1.9451	140.0	840	0.71	1.5916	0.627	0.71
1.812	141.0	846	0.655	1.5766	0.5635	0.655
1.656	142.0	852	0.695	1.5512	0.5997	0.695
1.7792	143.0	858	0.715	1.5488	0.6312	0.715
1.7944	144.0	864	0.685	1.5584	0.5923	0.685
1.6757	145.0	870	0.715	1.5074	0.6233	0.715
1.6519	146.0	876	0.71	1.5170	0.6353	0.71
1.6211	147.0	882	0.725	1.4860	0.635	0.725
1.7501	148.0	888	0.725	1.4868	0.6575	0.725
1.5653	149.0	894	0.72	1.4813	0.6362	0.72
1.8336	150.0	900	0.73	1.4826	0.6463	0.73
1.8838	151.0	906	0.72	1.4810	0.6397	0.72
1.5399	152.0	912	0.735	1.4701	0.657	0.735
1.723	153.0	918	0.725	1.4678	0.6445	0.725
1.5305	154.0	924	0.725	1.4434	0.6415	0.725
1.4943	155.0	930	0.75	1.4164	0.6810	0.75
1.5646	156.0	936	0.725	1.4331	0.6433	0.725
1.5382	157.0	942	0.755	1.4091	0.6863	0.755
1.627	158.0	948	0.735	1.4264	0.6580	0.735
1.4762	159.0	954	0.705	1.3982	0.6013	0.705
1.5063	160.0	960	0.705	1.3832	0.6213	0.705
1.5869	161.0	966	0.715	1.3863	0.6262	0.715
1.9302	162.0	972	0.735	1.3584	0.6457	0.735
1.378	163.0	978	0.72	1.3744	0.6395	0.72
1.7616	164.0	984	0.725	1.3423	0.6307	0.725
1.6735	165.0	990	0.755	1.3338	0.6840	0.755
1.4282	166.0	996	0.73	1.3135	0.6593	0.73
1.5056	167.0	1002	0.725	1.3506	0.6457	0.725
1.5995	168.0	1008	0.745	1.2977	0.6660	0.745
1.4362	169.0	1014	0.715	1.3200	0.6240	0.715
1.3997	170.0	1020	0.725	1.2853	0.649	0.725
1.5701	171.0	1026	0.75	1.3088	0.6830	0.75
1.3818	172.0	1032	0.725	1.3047	0.6357	0.725
1.5829	173.0	1038	0.77	1.2654	0.7057	0.77
1.5523	174.0	1044	0.73	1.2629	0.6567	0.73
1.8276	175.0	1050	0.77	1.2461	0.6947	0.77
1.3251	176.0	1056	0.715	1.2612	0.643	0.715
1.373	177.0	1062	0.73	1.2579	0.6540	0.73
1.5401	178.0	1068	0.75	1.2286	0.6927	0.75
1.5521	179.0	1074	0.73	1.2305	0.6613	0.73
1.6234	180.0	1080	0.76	1.2573	0.6937	0.76
1.5218	181.0	1086	0.755	1.2149	0.688	0.755
1.3136	182.0	1092	0.76	1.2223	0.6897	0.76
1.3794	183.0	1098	0.73	1.2080	0.6563	0.73
1.5216	184.0	1104	0.74	1.2023	0.6637	0.74
1.7363	185.0	1110	0.76	1.1980	0.6923	0.76
1.3141	186.0	1116	0.76	1.1972	0.6957	0.76
1.6765	187.0	1122	0.72	1.1820	0.6477	0.72
1.3	188.0	1128	0.76	1.1560	0.7063	0.76
1.5483	189.0	1134	0.75	1.1707	0.6723	0.75
1.2051	190.0	1140	0.735	1.1763	0.6563	0.735
1.9507	191.0	1146	0.78	1.1784	0.7237	0.78
1.305	192.0	1152	0.73	1.1539	0.6603	0.73
1.2795	193.0	1158	0.75	1.1403	0.6797	0.75
1.4111	194.0	1164	0.75	1.1507	0.6823	0.75
1.1497	195.0	1170	0.74	1.1406	0.6647	0.74
1.4996	196.0	1176	0.76	1.1158	0.6963	0.76
1.4444	197.0	1182	0.73	1.1355	0.6603	0.73
1.3417	198.0	1188	0.755	1.1324	0.6797	0.755
1.1406	199.0	1194	0.76	1.1591	0.6920	0.76
1.1757	200.0	1200	0.715	1.1135	0.6363	0.715
1.3617	201.0	1206	0.735	1.1023	0.6597	0.735
1.2367	202.0	1212	0.75	1.1295	0.6747	0.75
1.3171	203.0	1218	0.74	1.1078	0.6680	0.74
1.2745	204.0	1224	0.73	1.1112	0.6563	0.73
1.1156	205.0	1230	0.75	1.0996	0.6723	0.75
1.3313	206.0	1236	0.74	1.0861	0.6580	0.74
1.1574	207.0	1242	0.74	1.0804	0.6703	0.74
1.1958	208.0	1248	0.75	1.0807	0.6730	0.75
1.2248	209.0	1254	0.755	1.0895	0.6863	0.755
1.2119	210.0	1260	0.73	1.0966	0.6570	0.73
1.3926	211.0	1266	0.78	1.0557	0.7197	0.78
1.622	212.0	1272	0.76	1.0554	0.6847	0.76
1.0935	213.0	1278	0.77	1.0435	0.7153	0.77
1.3827	214.0	1284	0.775	1.0450	0.7153	0.775
1.1396	215.0	1290	0.78	1.0312	0.722	0.78
1.1396	216.0	1296	0.76	1.0262	0.6990	0.76
1.3323	217.0	1302	0.77	1.0134	0.708	0.77
1.528	218.0	1308	0.755	1.0120	0.6840	0.755
1.3792	219.0	1314	0.75	1.0258	0.6747	0.75
1.1346	220.0	1320	0.76	1.0285	0.6830	0.76
1.3889	221.0	1326	0.755	1.0195	0.6897	0.755
1.05	222.0	1332	0.77	1.0092	0.698	0.77
0.9776	223.0	1338	0.765	0.9792	0.7083	0.765
1.0452	224.0	1344	0.775	0.9796	0.7090	0.775
1.1201	225.0	1350	0.765	0.9833	0.6920	0.765
1.1193	226.0	1356	0.79	0.9759	0.7287	0.79
1.5232	227.0	1362	0.77	0.9979	0.6890	0.77
1.396	228.0	1368	0.77	0.9695	0.7047	0.77
1.2484	229.0	1374	0.76	0.9747	0.6830	0.76
1.0507	230.0	1380	0.76	0.9871	0.6810	0.76
1.3088	231.0	1386	0.76	0.9829	0.6823	0.76
1.0486	232.0	1392	0.79	0.9811	0.7297	0.79
0.9835	233.0	1398	0.76	0.9728	0.6970	0.76
1.0444	234.0	1404	0.9399	0.78	0.7123	0.78
1.1453	235.0	1410	0.9437	0.755	0.6830	0.755
1.1698	236.0	1416	0.9584	0.775	0.7047	0.775
0.9383	237.0	1422	0.9601	0.75	0.6769	0.75
1.0332	238.0	1428	0.9365	0.785	0.718	0.785
1.2459	239.0	1434	0.9300	0.78	0.7163	0.78
1.2756	240.0	1440	0.9475	0.75	0.6747	0.75
0.9732	241.0	1446	0.9400	0.79	0.723	0.79
1.211	242.0	1452	0.9266	0.78	0.7070	0.78
1.2219	243.0	1458	0.9242	0.77	0.7013	0.77
1.0283	244.0	1464	0.9027	0.775	0.7053	0.775
1.0291	245.0	1470	0.9043	0.775	0.7113	0.775
1.0978	246.0	1476	0.9127	0.785	0.7253	0.785
0.9759	247.0	1482	0.8754	0.79	0.7307	0.79
1.2333	248.0	1488	0.8764	0.765	0.6923	0.765
0.9991	249.0	1494	0.8965	0.78	0.7147	0.78
0.9778	250.0	1500	0.8986	0.765	0.6970	0.765
1.087	251.0	1506	0.8999	0.775	0.7053	0.775
1.2273	252.0	1512	0.8875	0.78	0.7083	0.78

Framework versions

• Transformers 4.47.0
• Pytorch 2.5.1+cu121
• Datasets 3.2.0
• Tokenizers 0.21.0