SignBart-WLASL-2000

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

• Loss: 1.4912
• Accuracy: 0.68
• Precision: 0.6019
• Recall: 0.68

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	Validation Loss	Accuracy	Precision	Recall
7.6489	1.0	110	7.5022	0.0013	0.0000	0.0013
7.3735	2.0	220	7.2033	0.0008	0.0000	0.0008
7.1295	3.0	330	6.9478	0.007	0.0016	0.007
6.8839	4.0	440	6.6738	0.0125	0.0019	0.0125
6.627	5.0	550	6.4258	0.0168	0.0036	0.0168
6.3866	6.0	660	6.1761	0.0395	0.0147	0.0395
6.1698	7.0	770	5.9318	0.0495	0.0177	0.0495
5.9667	8.0	880	5.7072	0.0843	0.0388	0.0843
5.7255	9.0	990	5.4857	0.1	0.0456	0.1
5.5576	10.0	1100	5.2878	0.1305	0.0719	0.1305
5.4192	11.0	1210	5.0944	0.1417	0.0780	0.1417
5.2082	12.0	1320	4.9029	0.16	0.0885	0.16
5.0117	13.0	1430	4.7469	0.1817	0.1019	0.1817
4.8283	14.0	1540	4.5616	0.2157	0.1267	0.2157
4.691	15.0	1650	4.3981	0.2323	0.1424	0.2323
4.542	16.0	1760	4.2604	0.245	0.1545	0.245
4.3522	17.0	1870	4.1121	0.2697	0.1786	0.2697
4.226	18.0	1980	3.9804	0.28	0.1836	0.28
4.0573	19.0	2090	3.8366	0.3073	0.2070	0.3073
4.0087	20.0	2200	3.7192	0.3152	0.2196	0.3152
3.7773	21.0	2310	3.5762	0.3438	0.2410	0.3438
3.8235	22.0	2420	3.4872	0.3453	0.2438	0.3453
3.6151	23.0	2530	3.3964	0.356	0.2555	0.356
3.5224	24.0	2640	3.2593	0.372	0.2646	0.372
3.3423	25.0	2750	3.1617	0.3835	0.2776	0.3835
3.2568	26.0	2860	3.0623	0.3982	0.2932	0.3982
3.2763	27.0	2970	2.9700	0.4108	0.3072	0.4108
3.0283	28.0	3080	2.9304	0.406	0.3039	0.406
2.9101	29.0	3190	2.7948	0.4263	0.3239	0.4263
2.9604	30.0	3300	2.7174	0.44	0.3355	0.44
2.7793	31.0	3410	2.6641	0.4387	0.3450	0.4387
2.8075	32.0	3520	2.5902	0.4472	0.3489	0.4472
2.8177	33.0	3630	2.5332	0.4597	0.3537	0.4597
2.5544	34.0	3740	2.4934	0.4693	0.3702	0.4693
2.5066	35.0	3850	2.4003	0.4795	0.3816	0.4795
2.4216	36.0	3960	2.3750	0.4833	0.3842	0.4833
2.4033	37.0	4070	2.2923	0.4865	0.3857	0.4865
2.4433	38.0	4180	2.2590	0.494	0.3964	0.494
2.3229	39.0	4290	2.2439	0.502	0.4022	0.502
2.1739	40.0	4400	2.1781	0.5075	0.4114	0.5075
2.2153	41.0	4510	2.1279	0.519	0.4237	0.519
2.3144	42.0	4620	2.0859	0.5208	0.4223	0.5208
2.0518	43.0	4730	2.0583	0.533	0.4355	0.533
2.0481	44.0	4840	2.0337	0.527	0.4292	0.527
2.054	45.0	4950	1.9905	0.5335	0.4408	0.5335
2.0334	46.0	5060	1.9681	0.5503	0.4608	0.5503
1.8618	47.0	5170	1.9568	0.5467	0.4553	0.5467
1.9496	48.0	5280	1.9172	0.548	0.4559	0.548
1.8307	49.0	5390	1.8857	0.5585	0.4680	0.5585
2.1095	50.0	5500	1.8818	0.5617	0.4701	0.5617
1.7956	51.0	5610	1.8739	0.565	0.4741	0.565
1.9301	52.0	5720	1.8454	0.5757	0.4884	0.5757
1.8489	53.0	5830	1.8089	0.5723	0.4829	0.5723
1.8966	54.0	5940	1.7888	0.5885	0.4994	0.5885
1.7394	55.0	6050	1.7789	0.5773	0.4878	0.5773
1.7855	56.0	6160	1.7540	0.5877	0.5028	0.5877
1.8291	57.0	6270	1.7320	0.591	0.5037	0.591
1.6508	58.0	6380	1.7414	0.5917	0.5078	0.5917
1.7013	59.0	6490	1.7093	0.6088	0.5202	0.6088
1.7326	60.0	6600	1.6983	0.6072	0.5202	0.6072
1.6144	61.0	6710	1.6868	0.619	0.5283	0.619
1.5571	62.0	6820	1.6737	0.6125	0.5268	0.6125
1.5713	63.0	6930	1.6658	0.6225	0.5346	0.6225
1.5142	64.0	7040	1.6568	0.6135	0.5284	0.6135
1.5322	65.0	7150	1.6504	0.6235	0.5396	0.6235
1.5453	66.0	7260	1.6422	0.6195	0.5334	0.6195
1.6717	67.0	7370	1.6085	0.6312	0.5456	0.6312
1.4665	68.0	7480	1.6141	0.6312	0.5463	0.6312
1.4348	69.0	7590	1.6070	0.6375	0.5540	0.6375
1.4072	70.0	7700	1.6027	0.6318	0.5520	0.6318
1.4678	71.0	7810	1.6120	0.6352	0.5550	0.6352
1.5865	72.0	7920	1.5795	0.646	0.5606	0.646
1.4616	73.0	8030	1.5613	0.6438	0.5625	0.6438
1.4366	74.0	8140	1.5789	0.6388	0.5560	0.6388
1.4672	75.0	8250	1.5609	0.6442	0.5614	0.6442
1.5142	76.0	8360	1.5847	0.6408	0.5563	0.6408
1.4293	77.0	8470	1.5253	0.6595	0.5752	0.6595
1.3751	78.0	8580	1.5312	0.6573	0.5753	0.6573
1.4565	79.0	8690	1.5411	0.657	0.5720	0.657
1.4108	80.0	8800	1.5292	0.6545	0.5780	0.6545
1.3375	81.0	8910	1.5380	0.6558	0.5772	0.6558
1.537	82.0	9020	1.5216	0.6645	0.5857	0.6645
1.4765	83.0	9130	1.5259	0.6567	0.5795	0.6567
1.3163	84.0	9240	1.5263	0.656	0.5807	0.656
1.4499	85.0	9350	1.5158	0.6633	0.5832	0.6633
1.2989	86.0	9460	1.5037	0.6645	0.5809	0.6645
1.2658	87.0	9570	1.5058	0.6685	0.5875	0.6685
1.4123	88.0	9680	1.5158	0.6583	0.5789	0.6583
1.3052	89.0	9790	1.5227	0.6713	0.5919	0.6713
1.1476	90.0	9900	1.5025	0.6655	0.5892	0.6655
1.3008	91.0	10010	1.5131	0.6823	0.6061	0.6823
1.3031	92.0	10120	1.4897	0.6745	0.5929	0.6745
1.2874	93.0	10230	1.5153	0.669	0.5885	0.669
1.2294	94.0	10340	1.4878	0.679	0.6003	0.679
1.1178	95.0	10450	1.4735	0.6727	0.5954	0.6727
1.4336	96.0	10560	1.5133	0.6683	0.5898	0.6683
1.2048	97.0	10670	1.4928	0.6777	0.6014	0.6777
1.2698	98.0	10780	1.4600	0.6873	0.6087	0.6873
1.2215	99.0	10890	1.5048	0.6807	0.6053	0.6807
1.2245	100.0	11000	1.4827	0.6845	0.6112	0.6845
1.2975	101.0	11110	1.4915	0.6765	0.6008	0.6765
1.2672	102.0	11220	1.4956	0.6813	0.6042	0.6813
1.3949	103.0	11330	1.4912	0.68	0.6019	0.68

Framework versions

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