SignBart-ASL-200

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

• Loss: 0.7109
• Accuracy: 0.8221
• Precision: 0.8311
• Recall: 0.8221

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
• lr_scheduler_warmup_ratio: 0.4
• num_epochs: 200

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall
5.3549	1.0	19	5.3199	0.0047	0.0001	0.0047
5.3645	2.0	38	5.3187	0.0047	0.0001	0.0047
5.3424	3.0	57	5.3168	0.0047	0.0002	0.0047
5.3543	4.0	76	5.3143	0.0057	0.0002	0.0057
5.3376	5.0	95	5.3115	0.0057	0.0002	0.0057
5.3359	6.0	114	5.3081	0.0052	0.0002	0.0052
5.3501	7.0	133	5.3048	0.0052	0.0002	0.0052
5.3429	8.0	152	5.3012	0.0052	0.0004	0.0052
5.3396	9.0	171	5.2972	0.0071	0.0010	0.0071
5.3373	10.0	190	5.2930	0.0090	0.0012	0.0090
5.3249	11.0	209	5.2879	0.0100	0.0020	0.0100
5.3222	12.0	228	5.2816	0.0081	0.0019	0.0081
5.3205	13.0	247	5.2735	0.0104	0.0018	0.0104
5.3071	14.0	266	5.2614	0.0142	0.0025	0.0142
5.2991	15.0	285	5.2444	0.0152	0.0045	0.0152
5.2737	16.0	304	5.2162	0.0147	0.0039	0.0147
5.2555	17.0	323	5.1744	0.0180	0.0050	0.0180
5.2234	18.0	342	5.1180	0.0270	0.0056	0.0270
5.1801	19.0	361	5.0365	0.0308	0.0134	0.0308
5.1194	20.0	380	4.9496	0.0365	0.0166	0.0365
5.0521	21.0	399	4.8696	0.0323	0.0187	0.0323
4.9802	22.0	418	4.7886	0.0351	0.0243	0.0351
4.9402	23.0	437	4.7132	0.0403	0.0224	0.0403
4.8529	24.0	456	4.6353	0.0536	0.0374	0.0536
4.7494	25.0	475	4.5609	0.0588	0.0389	0.0588
4.6729	26.0	494	4.4885	0.0702	0.0748	0.0702
4.6316	27.0	513	4.4167	0.0778	0.0728	0.0778
4.5378	28.0	532	4.3467	0.0935	0.0907	0.0935
4.5376	29.0	551	4.2839	0.1129	0.1314	0.1129
4.3979	30.0	570	4.2131	0.1319	0.1312	0.1319
4.3222	31.0	589	4.1410	0.1850	0.1846	0.1850
4.3148	32.0	608	4.0735	0.1945	0.1944	0.1945
4.2017	33.0	627	3.9994	0.2301	0.2378	0.2301
4.1925	34.0	646	3.9264	0.2623	0.2861	0.2623
4.0518	35.0	665	3.8486	0.2880	0.2875	0.2880
4.0531	36.0	684	3.7749	0.3093	0.3281	0.3093
3.9153	37.0	703	3.6932	0.3321	0.3480	0.3321
3.9242	38.0	722	3.6332	0.3316	0.3543	0.3316
3.8218	39.0	741	3.5509	0.3681	0.4009	0.3681
3.7898	40.0	760	3.4834	0.3980	0.4284	0.3980
3.6222	41.0	779	3.4139	0.4075	0.4153	0.4075
3.5352	42.0	798	3.3201	0.4374	0.4748	0.4374
3.5514	43.0	817	3.2447	0.4488	0.4727	0.4488
3.6046	44.0	836	3.1651	0.4706	0.5061	0.4706
3.4516	45.0	855	3.1013	0.4763	0.5130	0.4763
3.4509	46.0	874	3.0329	0.4905	0.5349	0.4905
3.2986	47.0	893	2.9758	0.5009	0.5488	0.5009
3.2178	48.0	912	2.9014	0.5076	0.5326	0.5076
3.2297	49.0	931	2.8158	0.5308	0.5672	0.5308
3.1389	50.0	950	2.7522	0.5508	0.5783	0.5508
3.0557	51.0	969	2.7001	0.5394	0.5578	0.5394
2.9371	52.0	988	2.6214	0.5745	0.6102	0.5745
3.0186	53.0	1007	2.5843	0.5617	0.5954	0.5617
2.8527	54.0	1026	2.5029	0.5811	0.6218	0.5811
2.9163	55.0	1045	2.4460	0.5787	0.6054	0.5787
2.7863	56.0	1064	2.4063	0.6020	0.6250	0.6020
2.8894	57.0	1083	2.3272	0.6162	0.6424	0.6162
2.6293	58.0	1102	2.2795	0.6176	0.6508	0.6176
2.767	59.0	1121	2.2302	0.6153	0.6451	0.6153
2.7323	60.0	1140	2.1798	0.6257	0.6613	0.6257
2.4808	61.0	1159	2.1117	0.6471	0.6648	0.6471
2.618	62.0	1178	2.0771	0.6490	0.6723	0.6490
2.4898	63.0	1197	2.0241	0.6651	0.6912	0.6651
2.4625	64.0	1216	1.9684	0.6656	0.6791	0.6656
2.4547	65.0	1235	1.9293	0.6679	0.6904	0.6679
2.3561	66.0	1254	1.8772	0.6803	0.7051	0.6803
2.2464	67.0	1273	1.8578	0.6803	0.7143	0.6803
2.2122	68.0	1292	1.7968	0.6945	0.7243	0.6945
2.1056	69.0	1311	1.7516	0.7011	0.7299	0.7011
1.9995	70.0	1330	1.7363	0.6959	0.7253	0.6959
2.2053	71.0	1349	1.6867	0.7092	0.7319	0.7092
2.3178	72.0	1368	1.6469	0.7002	0.7212	0.7002
2.0398	73.0	1387	1.6208	0.7059	0.7353	0.7059
1.8561	74.0	1406	1.5764	0.7154	0.7386	0.7154
2.0418	75.0	1425	1.5329	0.7220	0.7420	0.7220
1.981	76.0	1444	1.5162	0.7163	0.7349	0.7163
1.9567	77.0	1463	1.4606	0.7239	0.7440	0.7239
1.8852	78.0	1482	1.4240	0.7386	0.7531	0.7386
1.7001	79.0	1501	1.4042	0.7343	0.7564	0.7343
1.6829	80.0	1520	1.3866	0.7367	0.7549	0.7367
1.8055	81.0	1539	1.3634	0.7348	0.7563	0.7348
1.836	82.0	1558	1.3024	0.7562	0.7678	0.7562
1.6973	83.0	1577	1.2858	0.7509	0.7645	0.7509
1.6283	84.0	1596	1.2776	0.7543	0.7704	0.7543
1.718	85.0	1615	1.2341	0.7585	0.7702	0.7585
1.5066	86.0	1634	1.2086	0.7638	0.7778	0.7638
1.4972	87.0	1653	1.2002	0.7609	0.7772	0.7609
1.6668	88.0	1672	1.1755	0.7609	0.7737	0.7609
1.4476	89.0	1691	1.1497	0.7666	0.7831	0.7666
1.3613	90.0	1710	1.1639	0.7547	0.7716	0.7547
1.4138	91.0	1729	1.1015	0.7728	0.7869	0.7728
1.5279	92.0	1748	1.0970	0.7690	0.7796	0.7690
1.3739	93.0	1767	1.0887	0.7652	0.7795	0.7652
1.7574	94.0	1786	1.0629	0.7704	0.7819	0.7704
1.4224	95.0	1805	1.0320	0.7770	0.7905	0.7770
1.4949	96.0	1824	1.0447	0.7742	0.7875	0.7742
1.243	97.0	1843	1.0287	0.7785	0.7913	0.7785
1.3567	98.0	1862	1.0163	0.7799	0.7907	0.7799
1.459	99.0	1881	1.0027	0.7804	0.7932	0.7804
1.4693	100.0	1900	1.0012	0.7856	0.7989	0.7856
1.3065	101.0	1919	0.9809	0.7761	0.7904	0.7761
1.2348	102.0	1938	0.9719	0.7799	0.7900	0.7799
1.2244	103.0	1957	0.9564	0.7851	0.7977	0.7851
1.2975	104.0	1976	0.9586	0.7789	0.7949	0.7789
1.3115	105.0	1995	0.9143	0.7993	0.8109	0.7993
1.4709	106.0	2014	0.9389	0.7818	0.7946	0.7818
1.1515	107.0	2033	0.9290	0.7827	0.7974	0.7827
1.1417	108.0	2052	0.9212	0.7898	0.8057	0.7898
0.9945	109.0	2071	0.9114	0.7917	0.8018	0.7917
1.1865	110.0	2090	0.9060	0.7889	0.8026	0.7889
1.2328	111.0	2109	0.9071	0.7832	0.7991	0.7832
0.9823	112.0	2128	0.8837	0.7951	0.8048	0.7951
1.048	113.0	2147	0.8754	0.7870	0.7975	0.7870
1.2036	114.0	2166	0.8514	0.8050	0.8151	0.8050
0.9754	115.0	2185	0.8641	0.7951	0.8092	0.7951
1.1285	116.0	2204	0.8523	0.8017	0.8146	0.8017
1.1034	117.0	2223	0.8427	0.7993	0.8092	0.7993
0.9819	118.0	2242	0.8544	0.7955	0.8075	0.7955
1.5943	119.0	2261	0.8278	0.7998	0.8097	0.7998
1.2826	120.0	2280	0.8418	0.7998	0.8159	0.7998
0.9552	121.0	2299	0.8384	0.7998	0.8108	0.7998
1.0865	122.0	2318	0.8308	0.8065	0.8202	0.8065
0.9937	123.0	2337	0.8239	0.8074	0.8184	0.8074
1.4032	124.0	2356	0.8122	0.8017	0.8158	0.8017
0.958	125.0	2375	0.8067	0.8055	0.8176	0.8055
1.2128	126.0	2394	0.8123	0.7960	0.8085	0.7960
0.9514	127.0	2413	0.8034	0.8008	0.8127	0.8008
1.0942	128.0	2432	0.7860	0.8102	0.8219	0.8102
0.7591	129.0	2451	0.7975	0.8065	0.8196	0.8065
0.8035	130.0	2470	0.7892	0.8046	0.8146	0.8046
0.9076	131.0	2489	0.7738	0.8098	0.8210	0.8098
0.9893	132.0	2508	0.7734	0.8121	0.8218	0.8121
0.7718	133.0	2527	0.7843	0.8046	0.8172	0.8046
0.8829	134.0	2546	0.7644	0.8069	0.8192	0.8069
1.124	135.0	2565	0.7750	0.8126	0.8261	0.8126
1.0116	136.0	2584	0.7686	0.8055	0.8162	0.8055
0.7288	137.0	2603	0.7533	0.8107	0.8190	0.8107
1.0708	138.0	2622	0.7552	0.8102	0.8214	0.8102
1.0291	139.0	2641	0.7522	0.8112	0.8209	0.8112
0.7209	140.0	2660	0.7486	0.8121	0.8213	0.8121
0.7913	141.0	2679	0.7413	0.8098	0.8186	0.8098
0.8083	142.0	2698	0.7370	0.8107	0.8179	0.8107
0.8165	143.0	2717	0.7395	0.8088	0.8206	0.8088
0.7434	144.0	2736	0.7446	0.8031	0.8135	0.8031
0.9177	145.0	2755	0.7353	0.8174	0.8274	0.8174
1.0394	146.0	2774	0.7324	0.8140	0.8234	0.8140
1.3435	147.0	2793	0.7397	0.8102	0.8213	0.8102
0.8562	148.0	2812	0.7298	0.8107	0.8201	0.8107
0.8174	149.0	2831	0.7312	0.8074	0.8183	0.8074
0.8868	150.0	2850	0.7361	0.8159	0.8253	0.8159
0.9154	151.0	2869	0.7315	0.8159	0.8254	0.8159
0.8896	152.0	2888	0.7197	0.8159	0.8255	0.8159
0.8435	153.0	2907	0.7216	0.8212	0.8292	0.8212
0.8209	154.0	2926	0.7255	0.8169	0.8282	0.8169
0.8216	155.0	2945	0.7150	0.8231	0.8326	0.8231
1.0481	156.0	2964	0.7187	0.8183	0.8259	0.8183
0.641	157.0	2983	0.7170	0.8131	0.8214	0.8131
0.6464	158.0	3002	0.7228	0.8207	0.8315	0.8207
0.7827	159.0	3021	0.7175	0.8174	0.8254	0.8174
0.5899	160.0	3040	0.7108	0.8150	0.8238	0.8150
0.8651	161.0	3059	0.7192	0.8178	0.8257	0.8178
0.7751	162.0	3078	0.7138	0.8145	0.8233	0.8145
0.7911	163.0	3097	0.7154	0.8193	0.8287	0.8193
0.9019	164.0	3116	0.7125	0.8169	0.8249	0.8169
0.6788	165.0	3135	0.7109	0.8221	0.8311	0.8221

Framework versions

• Transformers 4.46.3
• Pytorch 2.4.0
• Datasets 3.1.0
• Tokenizers 0.20.3