qevasion_flat_1782203759

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

• Loss: 4.2461
• Accuracy: 0.3855
• Precision Macro: 0.4055
• Recall Macro: 0.3971
• F1 Macro: 0.3973

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: 3e-05
• train_batch_size: 12
• eval_batch_size: 12
• seed: 120
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 500
• num_epochs: 371

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision Macro	Recall Macro	F1 Macro
1.9698	1.0	259	1.8661	0.3101	0.1082	0.1142	0.0582
1.8544	2.0	518	1.8735	0.3043	0.0590	0.1121	0.0577
1.7369	3.0	777	1.7141	0.3246	0.2073	0.1710	0.1506
1.4946	4.0	1036	1.6863	0.3884	0.3766	0.3260	0.3152
1.2273	5.0	1295	1.8349	0.3884	0.3925	0.3707	0.3775
1.0288	6.0	1554	2.0424	0.3623	0.3614	0.3320	0.3367
0.8834	7.0	1813	2.0523	0.3768	0.4129	0.3807	0.3899
0.7934	8.0	2072	2.1803	0.3768	0.3900	0.3685	0.3722
0.7381	9.0	2331	2.3786	0.3739	0.3932	0.4074	0.3973
0.6725	10.0	2590	2.5463	0.3884	0.3854	0.3897	0.3833
0.6345	11.0	2849	2.5735	0.3652	0.3743	0.3582	0.3636
0.6078	12.0	3108	2.6990	0.3884	0.3958	0.3861	0.3873
0.5738	13.0	3367	2.5616	0.3884	0.4046	0.3825	0.3890
0.5426	14.0	3626	2.8499	0.3942	0.4010	0.3977	0.3923
0.5155	15.0	3885	2.9224	0.3797	0.3909	0.3717	0.3800
0.5067	16.0	4144	3.0926	0.3942	0.4175	0.3971	0.4033
0.4810	17.0	4403	3.2033	0.3971	0.4219	0.3926	0.4044
0.4644	18.0	4662	3.5388	0.3884	0.4178	0.4072	0.4096
0.4619	19.0	4921	3.0813	0.3855	0.3906	0.3635	0.3730
0.4544	20.0	5180	3.2106	0.3884	0.4017	0.3780	0.3822
0.4413	21.0	5439	3.5367	0.3884	0.4004	0.4134	0.4033
0.4355	22.0	5698	3.8628	0.3797	0.3870	0.3801	0.3756
0.4329	23.0	5957	3.6883	0.3797	0.3840	0.3654	0.3698
0.4242	24.0	6216	3.7416	0.3826	0.3770	0.3679	0.3661
0.4193	25.0	6475	3.6155	0.3913	0.4057	0.3846	0.3889
0.4213	26.0	6734	3.8832	0.3594	0.3567	0.3599	0.3479
0.4068	27.0	6993	3.9166	0.3768	0.3950	0.3713	0.3787
0.4133	28.0	7252	3.8232	0.3884	0.3964	0.3844	0.3839
0.4047	29.0	7511	3.8123	0.3942	0.4159	0.3866	0.3971
0.4048	30.0	7770	3.9749	0.3768	0.4059	0.3740	0.3837
0.3946	31.0	8029	4.0958	0.3884	0.4118	0.3824	0.3914
0.4041	32.0	8288	4.1780	0.3681	0.3753	0.3594	0.3618
0.3954	33.0	8547	3.9895	0.3884	0.4133	0.3772	0.3903
0.4002	34.0	8806	4.3709	0.3826	0.3881	0.3804	0.3818
0.3883	35.0	9065	4.0575	0.3855	0.3990	0.3747	0.3838
0.3980	36.0	9324	4.0883	0.3710	0.3792	0.3582	0.3670
0.4011	37.0	9583	3.8487	0.3797	0.3891	0.3565	0.3691
0.3939	38.0	9842	4.3257	0.3652	0.3641	0.3712	0.3647
0.3921	39.0	10101	4.3116	0.3942	0.3979	0.3852	0.3883
0.3866	40.0	10360	4.1959	0.3768	0.3697	0.3668	0.3645
0.3921	41.0	10619	4.0224	0.3739	0.3831	0.3509	0.3641
0.3837	42.0	10878	4.3801	0.3739	0.3732	0.3733	0.3707
0.3760	43.0	11137	4.1546	0.3681	0.3806	0.3605	0.3676
0.3778	44.0	11396	4.3165	0.3768	0.3754	0.3721	0.3683
0.3803	45.0	11655	4.2174	0.3826	0.3892	0.3799	0.3780
0.3831	46.0	11914	4.3867	0.3826	0.3796	0.3694	0.3722
0.3765	47.0	12173	4.4456	0.3826	0.3937	0.3895	0.3894
0.3787	48.0	12432	4.4603	0.3710	0.3929	0.3844	0.3854
0.3818	49.0	12691	4.4421	0.3739	0.3683	0.3607	0.3596
0.3800	50.0	12950	4.4826	0.3942	0.4170	0.4320	0.4205
0.3710	51.0	13209	4.3705	0.3768	0.3918	0.3693	0.3759
0.3784	52.0	13468	4.2461	0.3855	0.4055	0.3971	0.3973

Framework versions

• Transformers 5.12.1
• Pytorch 2.10.0+cu128
• Datasets 5.0.0
• Tokenizers 0.22.2