Modelpublic

ManojAlexender/Trail_run_final_roberta

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes4downloads

Model Card

Trailrunfinal_roberta

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

—Loss: 0.2322
—Accuracy: 0.9163
—F1: 0.9159
—Precision: 0.9181
—Recall: 0.9163

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
—trainbatchsize: 16
—evalbatchsize: 64
—seed: 42
—gradientaccumulationsteps: 2
—totaltrainbatch_size: 32
—optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
—lrschedulertype: linear
—lrschedulerwarmup_steps: 500
—num_epochs: 3
—mixedprecisiontraining: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
0.4057	0.01	100	0.3391	0.8808	0.8795	0.8865	0.8808
0.3447	0.01	200	0.3667	0.8470	0.8417	0.8728	0.8470
0.3202	0.02	300	0.3871	0.8689	0.8656	0.8869	0.8689
0.2354	0.03	400	0.4020	0.8597	0.8557	0.8805	0.8597
0.3268	0.04	500	0.3679	0.8318	0.8252	0.8610	0.8318
0.2771	0.04	600	0.2474	0.8924	0.8912	0.8982	0.8924
0.2288	0.05	700	0.2297	0.9103	0.9103	0.9103	0.9103
0.2307	0.06	800	0.2633	0.8944	0.8939	0.8957	0.8944
0.3375	0.06	900	0.2458	0.8988	0.8979	0.9024	0.8988
0.26	0.07	1000	0.2428	0.9071	0.9065	0.9099	0.9071
0.274	0.08	1100	0.2395	0.9035	0.9036	0.9036	0.9035
0.2513	0.09	1200	0.4167	0.8569	0.8532	0.8751	0.8569
0.2281	0.09	1300	0.3968	0.8633	0.8598	0.8815	0.8633
0.249	0.1	1400	0.2548	0.8804	0.8783	0.8920	0.8804
0.1986	0.11	1500	0.2590	0.9020	0.9020	0.9021	0.9020
0.26	0.11	1600	0.3084	0.8804	0.8784	0.8913	0.8804
0.2272	0.12	1700	0.2827	0.8884	0.8870	0.8956	0.8884
0.2312	0.13	1800	0.2373	0.9067	0.9068	0.9068	0.9067
0.2563	0.14	1900	0.2628	0.9008	0.9008	0.9011	0.9008
0.1876	0.14	2000	0.2744	0.8852	0.8840	0.8906	0.8852
0.284	0.15	2100	0.2751	0.8928	0.8914	0.9002	0.8928
0.203	0.16	2200	0.2406	0.9031	0.9034	0.9054	0.9031
0.2278	0.16	2300	0.2378	0.9115	0.9112	0.9123	0.9115
0.2204	0.17	2400	0.4288	0.8677	0.8646	0.8837	0.8677
0.2323	0.18	2500	0.2331	0.9115	0.9113	0.9118	0.9115
0.2508	0.19	2600	0.2932	0.8956	0.8955	0.8955	0.8956
0.2838	0.19	2700	0.2454	0.9035	0.9036	0.9037	0.9035
0.221	0.2	2800	0.3153	0.8800	0.8783	0.8881	0.8800
0.2167	0.21	2900	0.3200	0.8745	0.8724	0.8838	0.8745
0.2336	0.21	3000	0.2842	0.8880	0.8866	0.8947	0.8880
0.2653	0.22	3100	0.2353	0.9059	0.9059	0.9059	0.9059
0.2953	0.23	3200	0.2374	0.9051	0.9044	0.9087	0.9051
0.174	0.24	3300	0.2810	0.8964	0.8954	0.9006	0.8964
0.2184	0.24	3400	0.2127	0.9127	0.9125	0.9131	0.9127
0.2519	0.25	3500	0.2286	0.9083	0.9085	0.9126	0.9083
0.2326	0.26	3600	0.2904	0.8948	0.8944	0.8956	0.8948
0.1862	0.26	3700	0.2203	0.9259	0.9258	0.9259	0.9259
0.2098	0.27	3800	0.2350	0.9075	0.9074	0.9077	0.9075
0.2152	0.28	3900	0.2319	0.9063	0.9063	0.9063	0.9063
0.3154	0.29	4000	0.2184	0.9071	0.9070	0.9072	0.9071
0.1679	0.29	4100	0.4091	0.8764	0.8740	0.8892	0.8764
0.1535	0.3	4200	0.2574	0.9091	0.9090	0.9092	0.9091
0.1487	0.31	4300	0.2510	0.9063	0.9060	0.9072	0.9063
0.2337	0.31	4400	0.2163	0.9131	0.9128	0.9138	0.9131
0.3144	0.32	4500	0.2627	0.9051	0.9047	0.9062	0.9051
0.2487	0.33	4600	0.2557	0.8992	0.8985	0.9014	0.8992
0.2194	0.34	4700	0.2363	0.9159	0.9157	0.9163	0.9159
0.2602	0.34	4800	0.2374	0.9051	0.9053	0.9058	0.9051
0.2353	0.35	4900	0.2482	0.9059	0.9057	0.9062	0.9059
0.2107	0.36	5000	0.2903	0.9008	0.8998	0.9052	0.9008
0.2364	0.36	5100	0.2901	0.8760	0.8746	0.8815	0.8760
0.2009	0.37	5200	0.2491	0.9091	0.9086	0.9116	0.9091
0.2469	0.38	5300	0.3049	0.8992	0.8988	0.9000	0.8992
0.162	0.39	5400	0.2847	0.9059	0.9055	0.9071	0.9059
0.24	0.39	5500	0.2146	0.9135	0.9132	0.9143	0.9135
0.2667	0.4	5600	0.2379	0.9075	0.9072	0.9085	0.9075
0.2165	0.41	5700	0.2662	0.8844	0.8829	0.8915	0.8844
0.2007	0.41	5800	0.2539	0.9047	0.9039	0.9087	0.9047
0.221	0.42	5900	0.2272	0.9047	0.9046	0.9047	0.9047
0.2028	0.43	6000	0.3618	0.8669	0.8638	0.8826	0.8669
0.3003	0.44	6100	0.2454	0.9071	0.9071	0.9071	0.9071
0.2025	0.44	6200	0.2103	0.9175	0.9175	0.9175	0.9175
0.253	0.45	6300	0.2470	0.8992	0.8981	0.9044	0.8992
0.1955	0.46	6400	0.2887	0.9000	0.8992	0.9031	0.9000
0.1621	0.46	6500	0.2245	0.9151	0.9149	0.9155	0.9151
0.2532	0.47	6600	0.2493	0.8912	0.8907	0.8924	0.8912
0.1898	0.48	6700	0.2313	0.9083	0.9082	0.9083	0.9083
0.1858	0.49	6800	0.2514	0.9031	0.9026	0.9049	0.9031
0.1977	0.49	6900	0.2155	0.9167	0.9166	0.9167	0.9167
0.2247	0.5	7000	0.2280	0.9059	0.9056	0.9070	0.9059
0.1931	0.51	7100	0.2431	0.9047	0.9042	0.9066	0.9047
0.1746	0.51	7200	0.2400	0.9155	0.9152	0.9164	0.9155
0.2579	0.52	7300	0.2707	0.9107	0.9102	0.9125	0.9107
0.2139	0.53	7400	0.2625	0.8920	0.8910	0.8965	0.8920
0.2703	0.54	7500	0.2500	0.8980	0.8972	0.9013	0.8980
0.1412	0.54	7600	0.2210	0.9159	0.9158	0.9160	0.9159
0.2382	0.55	7700	0.2712	0.9028	0.9020	0.9064	0.9028
0.2498	0.56	7800	0.2200	0.9195	0.9193	0.9197	0.9195
0.2002	0.56	7900	0.3254	0.8832	0.8813	0.8935	0.8832
0.2359	0.57	8000	0.3023	0.8928	0.8918	0.8973	0.8928
0.2193	0.58	8100	0.2837	0.8892	0.8875	0.8988	0.8892
0.2436	0.59	8200	0.2221	0.9143	0.9142	0.9143	0.9143
0.1704	0.59	8300	0.2402	0.9123	0.9119	0.9136	0.9123
0.1979	0.6	8400	0.2722	0.8912	0.8896	0.9003	0.8912
0.2476	0.61	8500	0.2165	0.9211	0.9209	0.9216	0.9211
0.1996	0.61	8600	0.2374	0.9151	0.9148	0.9163	0.9151
0.2278	0.62	8700	0.2357	0.9079	0.9080	0.9083	0.9079
0.1625	0.63	8800	0.2205	0.9231	0.9228	0.9237	0.9231
0.2197	0.64	8900	0.3041	0.9020	0.9011	0.9063	0.9020
0.1868	0.64	9000	0.2280	0.9207	0.9205	0.9212	0.9207
0.2979	0.65	9100	0.2931	0.8948	0.8935	0.9011	0.8948
0.1973	0.66	9200	0.2322	0.9163	0.9159	0.9181	0.9163

Framework versions

—Transformers 4.39.3
—Pytorch 2.1.2+cu121
—Datasets 2.16.1
—Tokenizers 0.15.1