vinh120203/Bert_SA_rwBK_20e

finBERT_sentiment_analysis_20e

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

• Loss: 0.8337
• Accuracy: 0.9040
• F1: 0.9040
• Precision: 0.9038
• Recall: 0.9044

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: 64
• eval_batch_size: 64
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 20
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1	Precision	Recall
1.2898	0.1323	50	0.5537	0.7885	0.7864	0.7889	0.7896
0.5067	0.2646	100	0.4689	0.8270	0.8216	0.8290	0.8281
0.4212	0.3968	150	0.4237	0.8467	0.8458	0.8475	0.8476
0.3969	0.5291	200	0.4118	0.8502	0.8477	0.8500	0.8511
0.3891	0.6614	250	0.3767	0.8545	0.8559	0.8586	0.8546
0.3927	0.7937	300	0.3542	0.8696	0.8695	0.8692	0.8701
0.378	0.9259	350	0.3563	0.8703	0.8689	0.8694	0.8710
0.3537	1.0582	400	0.3472	0.8686	0.8696	0.8709	0.8689
0.32	1.1905	450	0.3591	0.8721	0.8710	0.8710	0.8727
0.3119	1.3228	500	0.3545	0.8709	0.8704	0.8707	0.8714
0.2992	1.4550	550	0.3378	0.8753	0.8763	0.8777	0.8756
0.3025	1.5873	600	0.3320	0.8785	0.8779	0.8778	0.8790
0.2913	1.7196	650	0.3835	0.8729	0.8702	0.8737	0.8738
0.3103	1.8519	700	0.3321	0.8812	0.8805	0.8804	0.8817
0.2847	1.9841	750	0.3337	0.8832	0.8836	0.8839	0.8835
0.2037	2.1164	800	0.3848	0.8809	0.8811	0.8812	0.8812
0.2199	2.2487	850	0.4087	0.8708	0.8690	0.8720	0.8714
0.2103	2.3810	900	0.3562	0.8794	0.8784	0.8787	0.8799
0.2178	2.5132	950	0.3473	0.8848	0.8847	0.8846	0.8853
0.2067	2.6455	1000	0.3554	0.8864	0.8863	0.8861	0.8869
0.2327	2.7778	1050	0.3667	0.8761	0.8778	0.8826	0.8760
0.2139	2.9101	1100	0.3657	0.8813	0.8825	0.8847	0.8814
0.1998	3.0423	1150	0.3761	0.8802	0.8815	0.8843	0.8804
0.1387	3.1746	1200	0.3918	0.8870	0.8874	0.8876	0.8873
0.1355	3.3069	1250	0.4323	0.8810	0.8791	0.8818	0.8818
0.1546	3.4392	1300	0.4122	0.8853	0.8842	0.8848	0.8860
0.135	3.5714	1350	0.3948	0.8862	0.8857	0.8857	0.8867
0.1827	3.7037	1400	0.3676	0.8846	0.8852	0.8858	0.8848
0.1551	3.8360	1450	0.3910	0.8893	0.8887	0.8888	0.8898
0.1415	3.9683	1500	0.3669	0.8913	0.8919	0.8926	0.8916
0.1122	4.1005	1550	0.4385	0.8876	0.8882	0.8897	0.8879
0.0996	4.2328	1600	0.4151	0.8926	0.8922	0.8921	0.8931
0.1077	4.3651	1650	0.4277	0.8902	0.8902	0.8899	0.8906
0.1207	4.4974	1700	0.4166	0.8859	0.8852	0.8853	0.8864
0.0993	4.6296	1750	0.4141	0.8931	0.8934	0.8934	0.8933
0.1172	4.7619	1800	0.4173	0.8929	0.8933	0.8937	0.8931
0.121	4.8942	1850	0.4067	0.8934	0.8926	0.8928	0.8939
0.1134	5.0265	1900	0.4496	0.8936	0.8930	0.8931	0.8942
0.0677	5.1587	1950	0.4808	0.8896	0.8901	0.8906	0.8899
0.0722	5.2910	2000	0.4848	0.8881	0.8889	0.8903	0.8881
0.0917	5.4233	2050	0.4863	0.8927	0.8931	0.8937	0.8928
0.0871	5.5556	2100	0.4359	0.8973	0.8969	0.8967	0.8977
0.078	5.6878	2150	0.4410	0.8926	0.8925	0.8924	0.8931
0.0761	5.8201	2200	0.4724	0.8949	0.8945	0.8944	0.8954
0.0925	5.9524	2250	0.4932	0.8953	0.8944	0.8950	0.8958
0.076	6.0847	2300	0.5118	0.8885	0.8885	0.8889	0.8887
0.049	6.2169	2350	0.5233	0.8928	0.8930	0.8930	0.8930
0.0628	6.3492	2400	0.5108	0.9001	0.8998	0.8997	0.9006
0.0661	6.4815	2450	0.5096	0.8952	0.8952	0.8950	0.8955
0.0625	6.6138	2500	0.5538	0.8917	0.8920	0.8921	0.8919
0.0689	6.7460	2550	0.5341	0.8929	0.8931	0.8930	0.8932
0.0614	6.8783	2600	0.5080	0.8966	0.8968	0.8968	0.8969
0.0699	7.0106	2650	0.5037	0.8987	0.8987	0.8986	0.8991
0.0527	7.1429	2700	0.5176	0.9002	0.9002	0.9001	0.9006
0.0553	7.2751	2750	0.5412	0.8973	0.8980	0.8988	0.8976
0.0601	7.4074	2800	0.5279	0.8905	0.8916	0.8939	0.8906
0.0519	7.5397	2850	0.5628	0.9008	0.9006	0.9008	0.9013
0.0418	7.6720	2900	0.5653	0.8977	0.8974	0.8973	0.8982
0.0499	7.8042	2950	0.5412	0.8970	0.8972	0.8973	0.8974
0.0424	7.9365	3000	0.5626	0.8977	0.8969	0.8973	0.8982
0.0324	8.0688	3050	0.6073	0.9001	0.9000	0.8999	0.9005
0.0309	8.2011	3100	0.6108	0.8982	0.8983	0.8982	0.8984
0.03	8.3333	3150	0.6021	0.8975	0.8973	0.8971	0.8979
0.0429	8.4656	3200	0.6003	0.8953	0.8955	0.8955	0.8956
0.0455	8.5979	3250	0.6162	0.8947	0.8953	0.8961	0.8948
0.037	8.7302	3300	0.5923	0.8957	0.8961	0.8962	0.8959
0.0462	8.8624	3350	0.5522	0.8979	0.8977	0.8975	0.8983
0.0356	8.9947	3400	0.5926	0.9010	0.9009	0.9011	0.9014
0.0243	9.1270	3450	0.6353	0.8972	0.8968	0.8967	0.8976
0.0341	9.2593	3500	0.6161	0.8925	0.8931	0.8939	0.8926
0.0271	9.3915	3550	0.6381	0.9008	0.9007	0.9006	0.9012
0.0344	9.5238	3600	0.6282	0.9001	0.9000	0.8998	0.9005
0.0236	9.6561	3650	0.7047	0.8982	0.8968	0.8989	0.8988
0.035	9.7884	3700	0.6561	0.8975	0.8974	0.8974	0.8979
0.0308	9.9206	3750	0.6754	0.8973	0.8968	0.8968	0.8978
0.0404	10.0529	3800	0.6452	0.8994	0.8988	0.8989	0.8999
0.0176	10.1852	3850	0.6636	0.8993	0.8989	0.8988	0.8998
0.0233	10.3175	3900	0.6820	0.8953	0.8948	0.8947	0.8957
0.0192	10.4497	3950	0.6954	0.8989	0.8981	0.8986	0.8995
0.0175	10.5820	4000	0.6959	0.8985	0.8982	0.8980	0.8989
0.0339	10.7143	4050	0.6624	0.8990	0.8993	0.8995	0.8993
0.0259	10.8466	4100	0.6787	0.8997	0.8993	0.8994	0.9002
0.0236	10.9788	4150	0.6708	0.8987	0.8989	0.8992	0.8990
0.0175	11.1111	4200	0.6893	0.9021	0.9021	0.9023	0.9026
0.0233	11.2434	4250	0.6769	0.8999	0.8998	0.8996	0.9003
0.0112	11.3757	4300	0.6949	0.8990	0.8990	0.8988	0.8993
0.017	11.5079	4350	0.6952	0.9019	0.9015	0.9015	0.9023
0.0159	11.6402	4400	0.6913	0.9031	0.9032	0.9032	0.9035
0.0214	11.7725	4450	0.7120	0.8996	0.8988	0.8992	0.9001
0.0257	11.9048	4500	0.6963	0.9032	0.9031	0.9031	0.9036
0.0189	12.0370	4550	0.6746	0.9032	0.9031	0.9031	0.9036
0.0138	12.1693	4600	0.7145	0.8996	0.9001	0.9008	0.8998
0.0095	12.3016	4650	0.7094	0.9018	0.9017	0.9017	0.9022
0.0189	12.4339	4700	0.7084	0.9000	0.9001	0.9000	0.9002
0.0159	12.5661	4750	0.7567	0.8937	0.8941	0.8948	0.8938
0.0127	12.6984	4800	0.7099	0.9013	0.9011	0.9009	0.9017
0.0147	12.8307	4850	0.7231	0.9032	0.9032	0.9030	0.9036
0.0134	12.9630	4900	0.7168	0.9008	0.9009	0.9008	0.9011
0.0121	13.0952	4950	0.7427	0.9030	0.9027	0.9027	0.9035
0.0114	13.2275	5000	0.7568	0.8998	0.8999	0.8998	0.9001
0.0157	13.3598	5050	0.7427	0.9024	0.9019	0.9020	0.9029
0.0104	13.4921	5100	0.7503	0.9020	0.9014	0.9015	0.9024
0.0129	13.6243	5150	0.7438	0.9020	0.9018	0.9017	0.9024
0.0152	13.7566	5200	0.7613	0.8984	0.8987	0.8987	0.8986
0.0072	13.8889	5250	0.7603	0.9030	0.9026	0.9026	0.9034
0.0103	14.0212	5300	0.7771	0.9000	0.9003	0.9004	0.9003
0.0115	14.1534	5350	0.7600	0.9031	0.9031	0.9030	0.9035
0.006	14.2857	5400	0.7614	0.9034	0.9031	0.9030	0.9038
0.0067	14.4180	5450	0.7912	0.9021	0.9023	0.9023	0.9023
0.0089	14.5503	5500	0.7771	0.9030	0.9031	0.9030	0.9034
0.0103	14.6825	5550	0.7795	0.9031	0.9031	0.9029	0.9035
0.0159	14.8148	5600	0.7478	0.9040	0.9039	0.9037	0.9043
0.0089	14.9471	5650	0.7904	0.8973	0.8978	0.8983	0.8974
0.0115	15.0794	5700	0.7904	0.8987	0.8990	0.8989	0.8990
0.0063	15.2116	5750	0.7864	0.9033	0.9032	0.9030	0.9037
0.0078	15.3439	5800	0.7965	0.9001	0.9005	0.9006	0.9004
0.0026	15.4762	5850	0.7972	0.9027	0.9026	0.9024	0.9030
0.0109	15.6085	5900	0.7800	0.9031	0.9029	0.9030	0.9036
0.0075	15.7407	5950	0.7770	0.9049	0.9047	0.9046	0.9053
0.008	15.8730	6000	0.7980	0.9013	0.9017	0.9019	0.9015
0.0039	16.0053	6050	0.7939	0.9045	0.9044	0.9043	0.9049
0.0048	16.1376	6100	0.8197	0.9003	0.9006	0.9007	0.9005
0.0077	16.2698	6150	0.8159	0.9030	0.9028	0.9027	0.9035
0.0047	16.4021	6200	0.8150	0.9018	0.9019	0.9017	0.9021
0.0044	16.5344	6250	0.8150	0.9018	0.9020	0.9019	0.9021
0.0057	16.6667	6300	0.8151	0.9025	0.9024	0.9023	0.9028
0.0089	16.7989	6350	0.8155	0.9026	0.9022	0.9021	0.9030
0.0027	16.9312	6400	0.8215	0.9028	0.9029	0.9029	0.9031
0.0041	17.0635	6450	0.8356	0.9011	0.9011	0.9009	0.9015
0.0058	17.1958	6500	0.8291	0.9018	0.9018	0.9016	0.9022
0.003	17.3280	6550	0.8411	0.9017	0.9016	0.9014	0.9021
0.0086	17.4603	6600	0.8326	0.9010	0.9010	0.9008	0.9013
0.0041	17.5926	6650	0.8296	0.9015	0.9015	0.9013	0.9018
0.0055	17.7249	6700	0.8302	0.9014	0.9014	0.9012	0.9017
0.005	17.8571	6750	0.8357	0.9021	0.9019	0.9017	0.9025
0.0038	17.9894	6800	0.8310	0.9015	0.9014	0.9012	0.9018
0.0065	18.1217	6850	0.8276	0.9026	0.9027	0.9026	0.9029
0.005	18.2540	6900	0.8336	0.9011	0.9013	0.9012	0.9014
0.002	18.3862	6950	0.8343	0.9014	0.9014	0.9012	0.9017
0.0022	18.5185	7000	0.8368	0.9033	0.9033	0.9032	0.9036
0.0045	18.6508	7050	0.8339	0.9032	0.9032	0.9031	0.9036
0.0055	18.7831	7100	0.8346	0.9040	0.9038	0.9037	0.9044
0.0034	18.9153	7150	0.8320	0.9038	0.9035	0.9034	0.9042
0.0037	19.0476	7200	0.8382	0.9039	0.9035	0.9035	0.9043
0.0024	19.1799	7250	0.8398	0.9040	0.9037	0.9038	0.9045
0.0041	19.3122	7300	0.8356	0.9035	0.9034	0.9034	0.9040
0.0037	19.4444	7350	0.8332	0.9036	0.9036	0.9034	0.9040
0.0052	19.5767	7400	0.8342	0.9036	0.9036	0.9034	0.9040
0.0051	19.7090	7450	0.8331	0.9039	0.9038	0.9036	0.9043
0.0043	19.8413	7500	0.8334	0.9042	0.9041	0.9040	0.9046
0.0022	19.9735	7550	0.8337	0.9040	0.9040	0.9038	0.9044

Framework versions

• Transformers 4.44.0
• Pytorch 2.2.1+cu121
• Tokenizers 0.19.1