SignBart-KArSL02-190

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

• Loss: 0.0716
• Accuracy: 0.9934
• Precision: 0.9943
• Recall: 0.9934

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: 0.4
• num_epochs: 200

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall
5.3336	1.0	32	5.2987	0.0086	0.0012	0.0086
5.3237	2.0	64	5.2869	0.0086	0.0013	0.0086
5.3131	3.0	96	5.2682	0.0099	0.0018	0.0099
5.2946	4.0	128	5.2422	0.0105	0.0023	0.0105
5.2673	5.0	160	5.2088	0.0145	0.0021	0.0145
5.2436	6.0	192	5.1689	0.0243	0.0036	0.0243
5.2079	7.0	224	5.1188	0.0296	0.0144	0.0296
5.1626	8.0	256	5.0595	0.0395	0.0248	0.0395
5.1057	9.0	288	4.9851	0.0539	0.0456	0.0539
5.0297	10.0	320	4.8976	0.075	0.0740	0.075
4.9676	11.0	352	4.7984	0.0882	0.0749	0.0882
4.8651	12.0	384	4.6843	0.0993	0.0663	0.0993
4.7767	13.0	416	4.5587	0.1454	0.1080	0.1454
4.6935	14.0	448	4.4318	0.1888	0.1385	0.1888
4.5998	15.0	480	4.3051	0.2447	0.2278	0.2447
4.4664	16.0	512	4.1649	0.2987	0.2676	0.2987
4.3658	17.0	544	4.0262	0.3342	0.3072	0.3342
4.2487	18.0	576	3.8854	0.375	0.3683	0.375
4.1063	19.0	608	3.7287	0.4322	0.4424	0.4322
3.9612	20.0	640	3.5731	0.4711	0.4690	0.4711
3.8803	21.0	672	3.4290	0.5125	0.5394	0.5125
3.6752	22.0	704	3.2657	0.5487	0.5642	0.5487
3.6902	23.0	736	3.1158	0.5724	0.6011	0.5724
3.499	24.0	768	2.9687	0.6112	0.6397	0.6112
3.3281	25.0	800	2.7928	0.6697	0.7011	0.6697
3.147	26.0	832	2.6164	0.7026	0.7276	0.7026
3.1176	27.0	864	2.4676	0.7336	0.7787	0.7336
2.9645	28.0	896	2.3092	0.7757	0.8045	0.7757
2.805	29.0	928	2.1415	0.8	0.8311	0.8
2.6921	30.0	960	1.9922	0.8257	0.8445	0.8257
2.5579	31.0	992	1.8445	0.8526	0.8716	0.8526
2.3409	32.0	1024	1.6978	0.8757	0.8821	0.8757
2.3109	33.0	1056	1.5622	0.8921	0.8991	0.8921
2.2235	34.0	1088	1.4396	0.8980	0.9080	0.8980
2.0948	35.0	1120	1.3258	0.9125	0.9205	0.9125
2.0076	36.0	1152	1.2190	0.9263	0.9294	0.9263
1.8034	37.0	1184	1.1180	0.9276	0.9372	0.9276
1.7826	38.0	1216	1.0216	0.9329	0.9491	0.9329
1.7587	39.0	1248	0.9280	0.9375	0.9464	0.9375
1.7431	40.0	1280	0.8627	0.9395	0.9521	0.9395
1.5773	41.0	1312	0.8022	0.9480	0.9575	0.9480
1.4618	42.0	1344	0.7288	0.9566	0.9648	0.9566
1.4851	43.0	1376	0.6780	0.9493	0.9582	0.9493
1.5139	44.0	1408	0.6270	0.9487	0.9616	0.9487
1.2987	45.0	1440	0.5705	0.9566	0.9670	0.9566
1.2984	46.0	1472	0.5395	0.9586	0.9677	0.9586
1.2926	47.0	1504	0.4812	0.9671	0.9722	0.9671
1.1754	48.0	1536	0.4504	0.9664	0.9727	0.9664
1.0912	49.0	1568	0.4167	0.9671	0.9746	0.9671
0.9942	50.0	1600	0.3912	0.9717	0.9765	0.9717
0.9692	51.0	1632	0.3609	0.9717	0.9774	0.9717
1.079	52.0	1664	0.3364	0.9776	0.9815	0.9776
0.9258	53.0	1696	0.3146	0.9717	0.9768	0.9717
0.8585	54.0	1728	0.2941	0.9770	0.9822	0.9770
0.8138	55.0	1760	0.2830	0.9743	0.9796	0.9743
0.9074	56.0	1792	0.2643	0.9770	0.9816	0.9770
0.8005	57.0	1824	0.2482	0.9809	0.9842	0.9809
0.7716	58.0	1856	0.2336	0.9789	0.9833	0.9789
0.8067	59.0	1888	0.2145	0.9849	0.9872	0.9849
0.6962	60.0	1920	0.2021	0.9816	0.9851	0.9816
0.8509	61.0	1952	0.1907	0.9829	0.9864	0.9829
0.7205	62.0	1984	0.1795	0.9796	0.9836	0.9796
0.6458	63.0	2016	0.1798	0.9849	0.9877	0.9849
0.7943	64.0	2048	0.1609	0.9829	0.9858	0.9829
0.6831	65.0	2080	0.1494	0.9882	0.9905	0.9882
0.6027	66.0	2112	0.1413	0.9849	0.9876	0.9849
0.6674	67.0	2144	0.1322	0.9882	0.9900	0.9882
0.5492	68.0	2176	0.1310	0.9882	0.9898	0.9882
0.6182	69.0	2208	0.1280	0.9855	0.9878	0.9855
0.5138	70.0	2240	0.1125	0.9875	0.9899	0.9875
0.7093	71.0	2272	0.1155	0.9888	0.9904	0.9888
0.6398	72.0	2304	0.1080	0.9901	0.9916	0.9901
0.602	73.0	2336	0.1036	0.9882	0.9901	0.9882
0.5343	74.0	2368	0.1048	0.9888	0.9905	0.9888
0.5908	75.0	2400	0.0974	0.9914	0.9926	0.9914
0.4642	76.0	2432	0.0927	0.9888	0.9911	0.9888
0.548	77.0	2464	0.0906	0.9908	0.9921	0.9908
0.454	78.0	2496	0.0892	0.9901	0.9918	0.9901
0.4374	79.0	2528	0.0853	0.9901	0.9915	0.9901
0.5177	80.0	2560	0.0834	0.9921	0.9930	0.9921
0.4313	81.0	2592	0.0790	0.9882	0.9899	0.9882
0.4908	82.0	2624	0.0746	0.9928	0.9937	0.9928
0.347	83.0	2656	0.0712	0.9921	0.9934	0.9921
0.4243	84.0	2688	0.0716	0.9934	0.9943	0.9934

Framework versions

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