classifier-de1_roberta

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

• Loss: 0.3372
• Accuracy: 0.8833
• Precision: 0.5359
• Recall: 0.4365
• F1: 0.4811

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: 1.5e-05
• 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: linear
• lr_scheduler_warmup_ratio: 0.1
• num_epochs: 6

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
0.3438	0.0513	500	0.3804	0.8760	0.0	0.0	0.0
0.3152	0.1025	1000	0.3564	0.8760	0.0	0.0	0.0
0.2733	0.1538	1500	0.3493	0.8641	0.4201	0.2516	0.3147
0.2512	0.2050	2000	0.3356	0.8740	0.4853	0.2719	0.3485
0.2508	0.2563	2500	0.3366	0.8850	0.6206	0.1865	0.2868
0.2355	0.3075	3000	0.3326	0.8865	0.6478	0.1864	0.2895
0.2345	0.3588	3500	0.3157	0.8849	0.5887	0.2378	0.3388
0.2178	0.4100	4000	0.3392	0.8872	0.6468	0.1983	0.3035
0.2127	0.4613	4500	0.3141	0.8874	0.6065	0.2622	0.3661
0.2047	0.5125	5000	0.3483	0.8889	0.6783	0.1985	0.3071
0.2052	0.5638	5500	0.3345	0.8872	0.6006	0.2696	0.3722
0.2107	0.6150	6000	0.3218	0.8891	0.6591	0.2185	0.3282
0.1984	0.6663	6500	0.3035	0.8823	0.5412	0.3366	0.4151
0.1827	0.7175	7000	0.3234	0.8882	0.6177	0.2586	0.3646
0.1793	0.7688	7500	0.3269	0.8878	0.5983	0.2898	0.3904
0.1831	0.8200	8000	0.3357	0.8898	0.6267	0.2743	0.3816
0.1789	0.8713	8500	0.3173	0.8879	0.5894	0.3160	0.4114
0.1669	0.9225	9000	0.3233	0.8840	0.5511	0.3461	0.4252
0.1711	0.9738	9500	0.3129	0.8891	0.6062	0.3012	0.4024
0.1745	1.0250	10000	0.3120	0.8893	0.6091	0.2996	0.4016
0.1656	1.0763	10500	0.3257	0.8890	0.5918	0.3381	0.4303
0.1668	1.1275	11000	0.3153	0.8864	0.5722	0.3321	0.4203
0.1505	1.1788	11500	0.3281	0.8840	0.5454	0.3882	0.4535
0.156	1.2300	12000	0.3344	0.8898	0.6064	0.3169	0.4162
0.155	1.2813	12500	0.3131	0.8896	0.6016	0.3249	0.4219
0.1729	1.3325	13000	0.3272	0.8911	0.6177	0.3204	0.4219
0.1558	1.3838	13500	0.3506	0.8911	0.6308	0.2942	0.4012
0.1587	1.4350	14000	0.3281	0.8916	0.6270	0.3113	0.4160
0.1609	1.4863	14500	0.3158	0.8900	0.6071	0.3195	0.4187
0.1638	1.5375	15000	0.3101	0.8937	0.6589	0.2958	0.4083
0.1372	1.5888	15500	0.3230	0.8928	0.6656	0.2729	0.3871
0.1594	1.6400	16000	0.3206	0.8909	0.6096	0.3347	0.4321
0.1589	1.6913	16500	0.3155	0.8924	0.6320	0.3169	0.4221
0.1515	1.7425	17000	0.3077	0.8908	0.6107	0.3301	0.4285
0.157	1.7938	17500	0.3080	0.8875	0.5690	0.3822	0.4573
0.1531	1.8450	18000	0.3258	0.8929	0.6694	0.2695	0.3843
0.1552	1.8963	18500	0.3241	0.8876	0.5742	0.3626	0.4445
0.1546	1.9475	19000	0.3167	0.8876	0.5755	0.3559	0.4398
0.14	1.9988	19500	0.3388	0.8876	0.5764	0.3537	0.4384
0.1579	2.0500	20000	0.3307	0.8911	0.6242	0.3068	0.4114
0.1402	2.1013	20500	0.3213	0.8906	0.6165	0.3105	0.4130
0.1349	2.1525	21000	0.3446	0.8902	0.6054	0.3281	0.4256
0.1506	2.2038	21500	0.3214	0.8909	0.6154	0.3208	0.4218
0.1219	2.2550	22000	0.3467	0.8893	0.5943	0.3376	0.4306
0.1469	2.3063	22500	0.3270	0.8922	0.6260	0.3244	0.4273
0.1487	2.3575	23000	0.3187	0.8890	0.5809	0.3765	0.4569
0.1341	2.4088	23500	0.3306	0.8887	0.5815	0.3646	0.4482
0.135	2.4600	24000	0.3415	0.8892	0.5908	0.3464	0.4367
0.1293	2.5113	24500	0.3546	0.8935	0.6468	0.3100	0.4192
0.1278	2.5625	25000	0.3618	0.8914	0.6185	0.3231	0.4245
0.1116	2.6138	25500	0.3505	0.8906	0.6033	0.3446	0.4386
0.1296	2.6650	26000	0.3494	0.8909	0.6195	0.3118	0.4148
0.1487	2.7163	26500	0.3328	0.8908	0.5979	0.3637	0.4523
0.1246	2.7675	27000	0.3481	0.8889	0.5771	0.3902	0.4656
0.1269	2.8188	27500	0.3844	0.8899	0.5929	0.3569	0.4455
0.1447	2.8700	28000	0.3254	0.8931	0.6391	0.3161	0.4230
0.1354	2.9213	28500	0.3251	0.8892	0.5814	0.3806	0.4600
0.137	2.9725	29000	0.3319	0.8922	0.6215	0.3351	0.4354
0.1063	3.0238	29500	0.3667	0.8924	0.6221	0.3370	0.4372
0.1095	3.0750	30000	0.3773	0.8919	0.6185	0.3347	0.4344
0.1254	3.1263	30500	0.3353	0.8929	0.6310	0.3275	0.4312
0.1204	3.1775	31000	0.3372	0.8833	0.5359	0.4365	0.4811
0.1013	3.2288	31500	0.3612	0.8893	0.5896	0.3536	0.4421
0.124	3.2800	32000	0.3470	0.8877	0.5683	0.3937	0.4652
0.1123	3.3313	32500	0.3479	0.8879	0.5718	0.3836	0.4592
0.1402	3.3825	33000	0.3561	0.8909	0.6038	0.3503	0.4433
0.1323	3.4338	33500	0.3417	0.8908	0.6044	0.3453	0.4395
0.1124	3.4850	34000	0.3409	0.8910	0.6018	0.3573	0.4484
0.1108	3.5363	34500	0.3723	0.8930	0.6400	0.3131	0.4205
0.121	3.5875	35000	0.3461	0.8920	0.6104	0.3567	0.4503
0.1145	3.6388	35500	0.3612	0.8901	0.5941	0.3601	0.4484
0.1052	3.6900	36000	0.3812	0.8916	0.6086	0.3518	0.4459
0.1181	3.7413	36500	0.3448	0.8877	0.5683	0.3916	0.4637
0.114	3.7925	37000	0.3502	0.8917	0.6081	0.3569	0.4498
0.1106	3.8438	37500	0.3499	0.8939	0.6471	0.3174	0.4259
0.127	3.8950	38000	0.3441	0.8926	0.6245	0.3360	0.4369
0.116	3.9463	38500	0.3358	0.8923	0.6163	0.3488	0.4455
0.1188	3.9975	39000	0.3496	0.8929	0.6295	0.3311	0.4340
0.1042	4.0488	39500	0.3884	0.8933	0.6292	0.3403	0.4417
0.1248	4.1000	40000	0.3510	0.8907	0.5950	0.3723	0.4580
0.0918	4.1513	40500	0.3670	0.8922	0.6149	0.3498	0.4460
0.1193	4.2025	41000	0.3725	0.8923	0.6130	0.3559	0.4504
0.1104	4.2538	41500	0.3949	0.8930	0.6335	0.3244	0.4290
0.1096	4.3050	42000	0.3742	0.8929	0.6268	0.3360	0.4375
0.122	4.3563	42500	0.3601	0.8935	0.6406	0.3209	0.4276
0.1162	4.4075	43000	0.3617	0.8917	0.6161	0.3359	0.4347
0.1031	4.4588	43500	0.3752	0.8918	0.6150	0.3416	0.4392
0.123	4.5100	44000	0.3720	0.8930	0.6245	0.3442	0.4438
0.0985	4.5613	44500	0.3783	0.8932	0.6296	0.3359	0.4381
0.1204	4.6125	45000	0.3617	0.8911	0.5987	0.3682	0.4560
0.1164	4.6638	45500	0.3566	0.8899	0.5928	0.3581	0.4465
0.1112	4.7150	46000	0.3622	0.8917	0.6162	0.3367	0.4354
0.1077	4.7663	46500	0.3831	0.8935	0.6612	0.2889	0.4021
0.1117	4.8175	47000	0.3668	0.8921	0.6144	0.3481	0.4444
0.1079	4.8688	47500	0.3830	0.8922	0.6173	0.3443	0.4421
0.1068	4.9200	48000	0.3710	0.8918	0.6116	0.3488	0.4443
0.1056	4.9713	48500	0.3824	0.8932	0.6439	0.3111	0.4195
0.1141	5.0226	49000	0.3688	0.8925	0.6204	0.3427	0.4415
0.123	5.0738	49500	0.3849	0.8912	0.5993	0.3707	0.4581
0.0977	5.1251	50000	0.4041	0.8925	0.6206	0.3414	0.4405
0.1091	5.1763	50500	0.3838	0.8901	0.5889	0.3755	0.4586
0.1082	5.2276	51000	0.3753	0.8914	0.6002	0.3717	0.4591
0.1119	5.2788	51500	0.3746	0.8919	0.6188	0.3348	0.4345
0.1206	5.3301	52000	0.3770	0.8913	0.6048	0.3562	0.4484
0.1136	5.3813	52500	0.3863	0.8905	0.5937	0.3709	0.4566
0.108	5.4326	53000	0.3671	0.8901	0.5889	0.3765	0.4594
0.1143	5.4838	53500	0.3798	0.8923	0.6178	0.3448	0.4426
0.1125	5.5351	54000	0.3828	0.8922	0.6169	0.3454	0.4428
0.1119	5.5863	54500	0.3743	0.8919	0.6129	0.3473	0.4434
0.1132	5.6376	55000	0.3621	0.8909	0.5989	0.3646	0.4533
0.1054	5.6888	55500	0.3651	0.8916	0.6091	0.3519	0.4461
0.0993	5.7401	56000	0.3721	0.8911	0.5999	0.3651	0.4539
0.0904	5.7913	56500	0.3825	0.8915	0.6065	0.3549	0.4478
0.1278	5.8426	57000	0.3801	0.8921	0.6148	0.3472	0.4438
0.0983	5.8938	57500	0.3813	0.8919	0.6122	0.3501	0.4455
0.0945	5.9451	58000	0.3819	0.8919	0.6119	0.3504	0.4456
0.0956	5.9963	58500	0.3767	0.8918	0.6097	0.3550	0.4487

Framework versions

• Transformers 4.51.3
• Pytorch 2.7.0+cu126
• Datasets 3.5.0
• Tokenizers 0.21.1