Intent-classification-BERT-Large-Ashuv3

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

• Loss: 0.2610
• Accuracy: 0.8951
• F1: 0.8807
• Precision: 0.8812
• Recall: 0.8820

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: 5e-05
• train_batch_size: 16
• eval_batch_size: 32
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.6762	0.24	10	1.3120	0.5280	0.4993	0.6178	0.5370
0.9717	0.49	20	0.7487	0.8571	0.8402	0.8670	0.8455
0.6375	0.73	30	0.4393	0.8509	0.8479	0.8862	0.8548
0.4006	0.98	40	0.2427	0.9068	0.9005	0.9228	0.9075
0.2291	1.22	50	0.1875	0.9068	0.8940	0.9106	0.8902
0.2634	1.46	60	0.2204	0.9068	0.8977	0.9135	0.9051
0.1916	1.71	70	0.1730	0.9130	0.9053	0.9232	0.9123
0.1881	1.95	80	0.1676	0.9130	0.9051	0.9232	0.9133
0.2692	2.2	90	0.1728	0.9068	0.8958	0.9423	0.8790
0.1425	2.44	100	0.1757	0.9068	0.8958	0.9423	0.8790
0.2674	2.68	110	0.3307	0.8758	0.8608	0.8756	0.8713
0.2385	2.93	120	0.1878	0.9006	0.8901	0.9059	0.8988
0.1868	3.17	130	0.1679	0.9130	0.9027	0.9147	0.9097
0.2281	3.41	140	0.1796	0.9130	0.9057	0.9274	0.9133
0.1459	3.66	150	0.1982	0.9068	0.8960	0.9077	0.9049
0.161	3.9	160	0.2266	0.8944	0.8772	0.9012	0.8765
0.1441	4.15	170	0.2062	0.8944	0.8889	0.9115	0.8935
0.172	4.39	180	0.2208	0.9006	0.8922	0.9216	0.8988
0.1365	4.63	190	0.2088	0.9068	0.8974	0.9244	0.9045
0.1795	4.88	200	0.2011	0.8820	0.8682	0.8936	0.8569
0.204	5.12	210	0.2377	0.8820	0.8642	0.8656	0.8721
0.1409	5.37	220	0.2178	0.8944	0.8852	0.9003	0.8776
0.1771	5.61	230	0.2284	0.8758	0.8624	0.8871	0.8511
0.1926	5.85	240	0.2211	0.8944	0.8815	0.8990	0.8761
0.2142	6.1	250	0.2217	0.9193	0.9082	0.9306	0.9130
0.1125	6.34	260	0.2321	0.9006	0.8889	0.9420	0.8702
0.1473	6.59	270	0.2129	0.9130	0.9057	0.9274	0.9133
0.1468	6.83	280	0.2318	0.9130	0.9057	0.9274	0.9133
0.1951	7.07	290	0.1957	0.9006	0.8879	0.9061	0.8788
0.1659	7.32	300	0.1961	0.9006	0.8872	0.9143	0.8752
0.1265	7.56	310	0.2058	0.9130	0.9049	0.9226	0.9097
0.1774	7.8	320	0.2223	0.9068	0.8974	0.9244	0.9045
0.2609	8.05	330	0.2218	0.8944	0.8833	0.8906	0.8811
0.1079	8.29	340	0.3312	0.8820	0.8675	0.8672	0.8680
0.1729	8.54	350	0.3627	0.8696	0.8500	0.8540	0.8554
0.2337	8.78	360	0.2526	0.9006	0.8872	0.9143	0.8752
0.1573	9.02	370	0.2072	0.9130	0.9049	0.9226	0.9097
0.1843	9.27	380	0.2605	0.9068	0.8991	0.9210	0.9085
0.1521	9.51	390	0.2695	0.9006	0.8920	0.9081	0.8966
0.193	9.76	400	0.3340	0.9130	0.9039	0.9187	0.9061
0.1034	10.0	410	0.3391	0.9068	0.8948	0.9025	0.9049
0.1348	10.24	420	0.3377	0.9006	0.8902	0.8998	0.8930
0.0856	10.49	430	0.3274	0.8882	0.8768	0.8920	0.8692
0.1877	10.73	440	0.3401	0.8696	0.8498	0.8504	0.8514
0.1775	10.98	450	0.4162	0.8882	0.8708	0.8716	0.8799
0.1357	11.22	460	0.3992	0.8820	0.8652	0.8622	0.8716
0.0878	11.46	470	0.3920	0.8944	0.8803	0.8772	0.8883
0.1892	11.71	480	0.3148	0.8696	0.8499	0.8472	0.8549
0.1712	11.95	490	0.3028	0.8758	0.8589	0.8585	0.8597
0.0914	12.2	500	0.3450	0.8820	0.8688	0.8705	0.8680
0.1793	12.44	510	0.3617	0.8882	0.8758	0.8872	0.8692
0.1355	12.68	520	0.4130	0.8820	0.8688	0.8705	0.8680
0.1518	12.93	530	0.5015	0.8944	0.8798	0.8808	0.8878
0.1778	13.17	540	0.3596	0.8882	0.8716	0.8709	0.8804
0.1662	13.41	550	0.3716	0.9006	0.8864	0.8868	0.8930
0.1105	13.66	560	0.3452	0.9006	0.8874	0.8903	0.8966
0.1369	13.9	570	0.3606	0.8944	0.8807	0.8824	0.8883
0.2051	14.15	580	0.3497	0.8882	0.8750	0.8784	0.8728
0.1441	14.39	590	0.4031	0.8820	0.8664	0.8649	0.8680
0.1586	14.63	600	0.3853	0.8820	0.8664	0.8649	0.8680
0.0974	14.88	610	0.4037	0.8820	0.8664	0.8649	0.8680
0.0799	15.12	620	0.5252	0.8820	0.8688	0.8705	0.8680
0.0969	15.37	630	0.5702	0.8820	0.8691	0.8699	0.8716
0.1664	15.61	640	0.5281	0.8820	0.8688	0.8705	0.8680
0.175	15.85	650	0.4865	0.8820	0.8688	0.8705	0.8680
0.1904	16.1	660	0.3893	0.8696	0.8528	0.8520	0.8549
0.1054	16.34	670	0.4320	0.8758	0.8612	0.8636	0.8597
0.1657	16.59	680	0.5669	0.8820	0.8688	0.8705	0.8680
0.1089	16.83	690	0.5642	0.8820	0.8677	0.8649	0.8716
0.0831	17.07	700	0.4782	0.8820	0.8709	0.8744	0.8716
0.1518	17.32	710	0.5122	0.8820	0.8695	0.8720	0.8680
0.1203	17.56	720	0.5720	0.8820	0.8695	0.8720	0.8680
0.1185	17.8	730	0.5798	0.8820	0.8698	0.8703	0.8716
0.1065	18.05	740	0.5495	0.8820	0.8685	0.8701	0.8716
0.13	18.29	750	0.6271	0.8820	0.8687	0.8696	0.8716
0.1382	18.54	760	0.6307	0.8758	0.8585	0.8556	0.8633
0.0979	18.78	770	0.6167	0.8758	0.8585	0.8556	0.8633
0.1328	19.02	780	0.6011	0.8758	0.8585	0.8556	0.8633
0.1561	19.27	790	0.5938	0.8696	0.8517	0.8495	0.8549
0.1638	19.51	800	0.6397	0.8696	0.8528	0.8520	0.8549
0.1358	19.76	810	0.6917	0.8758	0.8614	0.8649	0.8597
0.1298	20.0	820	0.6769	0.8696	0.8528	0.8489	0.8585
0.1102	20.24	830	0.6891	0.8758	0.8610	0.8594	0.8669
0.127	20.49	840	0.6950	0.8820	0.8685	0.8701	0.8716
0.1719	20.73	850	0.6719	0.8882	0.8754	0.8773	0.8799
0.1503	20.98	860	0.6462	0.8820	0.8675	0.8666	0.8716
0.1118	21.22	870	0.6405	0.8820	0.8690	0.8705	0.8680
0.0991	21.46	880	0.6492	0.8758	0.8614	0.8600	0.8633
0.1288	21.71	890	0.7045	0.8820	0.8688	0.8705	0.8680
0.1414	21.95	900	0.7439	0.8820	0.8688	0.8705	0.8680
0.1744	22.2	910	0.7353	0.8820	0.8688	0.8705	0.8680
0.1072	22.44	920	0.7524	0.8820	0.8688	0.8705	0.8680
0.0931	22.68	930	0.7671	0.8758	0.8614	0.8649	0.8597
0.0775	22.93	940	0.7442	0.8758	0.8614	0.8649	0.8597
0.0713	23.17	950	0.7456	0.8758	0.8614	0.8649	0.8597
0.1027	23.41	960	0.7528	0.8820	0.8664	0.8649	0.8680
0.1163	23.66	970	0.7503	0.8820	0.8664	0.8649	0.8680
0.1067	23.9	980	0.7359	0.8758	0.8622	0.8660	0.8597
0.0955	24.15	990	0.7457	0.8820	0.8676	0.8687	0.8680
0.0874	24.39	1000	0.7663	0.8820	0.8685	0.8701	0.8716
0.0865	24.63	1010	0.7761	0.8820	0.8685	0.8701	0.8716
0.1378	24.88	1020	0.7761	0.8820	0.8691	0.8699	0.8716
0.1411	25.12	1030	0.7714	0.8820	0.8676	0.8687	0.8680
0.1034	25.37	1040	0.7662	0.8820	0.8685	0.8700	0.8680
0.0709	25.61	1050	0.7720	0.8820	0.8670	0.8681	0.8680
0.1286	25.85	1060	0.7809	0.8820	0.8670	0.8681	0.8680
0.1191	26.1	1070	0.7861	0.8820	0.8676	0.8687	0.8680
0.0902	26.34	1080	0.7888	0.8820	0.8691	0.8699	0.8716
0.1054	26.59	1090	0.7894	0.8820	0.8698	0.8703	0.8716
0.1142	26.83	1100	0.7914	0.8820	0.8691	0.8699	0.8716
0.1175	27.07	1110	0.7923	0.8820	0.8691	0.8699	0.8716
0.1319	27.32	1120	0.7938	0.8820	0.8685	0.8701	0.8716
0.1181	27.56	1130	0.7967	0.8820	0.8685	0.8701	0.8716
0.0858	27.8	1140	0.8003	0.8820	0.8685	0.8701	0.8716
0.0697	28.05	1150	0.8025	0.8820	0.8685	0.8701	0.8716
0.0644	28.29	1160	0.8050	0.8820	0.8685	0.8701	0.8716
0.1123	28.54	1170	0.8063	0.8820	0.8685	0.8701	0.8716
0.0998	28.78	1180	0.8078	0.8820	0.8685	0.8701	0.8716
0.1297	29.02	1190	0.8095	0.8820	0.8685	0.8701	0.8716
0.1133	29.27	1200	0.8094	0.8820	0.8685	0.8701	0.8716
0.1122	29.51	1210	0.8095	0.8820	0.8685	0.8701	0.8716
0.1115	29.76	1220	0.8096	0.8820	0.8685	0.8701	0.8716
0.0692	30.0	1230	0.8095	0.8820	0.8685	0.8701	0.8716

Framework versions

• Transformers 4.38.2
• Pytorch 2.1.2
• Datasets 2.1.0
• Tokenizers 0.15.2