exceptions_exp2_swap_0.7_last_to_drop_1032

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

• Loss: 3.5637
• Accuracy: 0.3685

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: 0.0006
• train_batch_size: 16
• eval_batch_size: 16
• seed: 1032
• gradient_accumulation_steps: 5
• total_train_batch_size: 80
• optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.98) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
• lr_scheduler_type: linear
• lr_scheduler_warmup_steps: 100
• num_epochs: 50.0
• mixed_precision_training: Native AMP

Training results

Training Loss	Epoch	Step	Accuracy	Validation Loss
4.8334	0.2915	1000	0.2538	4.7653
4.3432	0.5830	2000	0.2984	4.2905
4.1536	0.8745	3000	0.3143	4.1021
3.995	1.1659	4000	0.3237	3.9967
3.9514	1.4574	5000	0.3312	3.9191
3.8843	1.7489	6000	0.3361	3.8622
3.7559	2.0402	7000	0.3404	3.8209
3.7585	2.3317	8000	0.3433	3.7903
3.7423	2.6233	9000	0.3457	3.7613
3.7272	2.9148	10000	0.3488	3.7338
3.6478	3.2061	11000	0.3504	3.7210
3.6472	3.4976	12000	0.3521	3.7040
3.6516	3.7891	13000	0.3536	3.6848
3.543	4.0805	14000	0.3552	3.6779
3.5733	4.3720	15000	0.3562	3.6663
3.5895	4.6635	16000	0.3572	3.6546
3.5878	4.9550	17000	0.3588	3.6382
3.5192	5.2463	18000	0.3590	3.6444
3.528	5.5378	19000	0.3600	3.6313
3.5324	5.8293	20000	0.3610	3.6199
3.436	6.1207	21000	0.3609	3.6246
3.4874	6.4122	22000	0.3620	3.6199
3.495	6.7037	23000	0.3628	3.6065
3.505	6.9952	24000	0.3634	3.5992
3.4347	7.2866	25000	0.3635	3.6087
3.4662	7.5781	26000	0.3640	3.5990
3.4609	7.8696	27000	0.3649	3.5871
3.3894	8.1609	28000	0.3647	3.5998
3.4282	8.4524	29000	0.3652	3.5920
3.4342	8.7439	30000	0.3658	3.5820
3.3384	9.0353	31000	0.3659	3.5885
3.3887	9.3268	32000	0.3657	3.5878
3.4005	9.6183	33000	0.3668	3.5784
3.4281	9.9098	34000	0.3675	3.5698
3.3461	10.2011	35000	0.3668	3.5807
3.3689	10.4927	36000	0.3673	3.5759
3.4026	10.7842	37000	0.3679	3.5656
3.2978	11.0755	38000	0.3678	3.5771
3.3608	11.3670	39000	0.3679	3.5725
3.3649	11.6585	40000	0.3685	3.5637
3.3894	11.9500	41000	0.3690	3.5592
3.3135	12.2414	42000	0.3683	3.5720
3.3346	12.5329	43000	0.3687	3.5648
3.3542	12.8244	44000	0.3693	3.5562
3.2718	13.1157	45000	0.3693	3.5683
3.3142	13.4072	46000	0.3689	3.5673
3.3455	13.6988	47000	0.3698	3.5549
3.3557	13.9903	48000	0.3706	3.5467
3.288	14.2816	49000	0.3698	3.5625
3.313	14.5731	50000	0.3703	3.5545
3.3325	14.8646	51000	0.3707	3.5467
3.2656	15.1560	52000	0.3703	3.5630
3.2885	15.4475	53000	0.3704	3.5575
3.3244	15.7390	54000	0.3711	3.5487
3.2198	16.0303	55000	0.3705	3.5613
3.2689	16.3218	56000	0.3708	3.5587
3.2886	16.6133	57000	0.3712	3.5549
3.3007	16.9049	58000	0.3716	3.5430
3.236	17.1962	59000	0.3705	3.5601
3.2693	17.4877	60000	0.3714	3.5534
3.2808	17.7792	61000	0.3718	3.5475
3.1944	18.0705	62000	0.3707	3.5628
3.2397	18.3621	63000	0.3711	3.5543
3.26	18.6536	64000	0.3717	3.5492
3.2709	18.9451	65000	0.3721	3.5419
3.2119	19.2364	66000	0.3716	3.5583
3.2563	19.5279	67000	0.3720	3.5489
3.2697	19.8194	68000	0.3723	3.5417
3.1929	20.1108	69000	0.3716	3.5584
3.2158	20.4023	70000	0.3722	3.5533
3.2561	20.6938	71000	0.3723	3.5448
3.2597	20.9853	72000	0.3726	3.5397
3.2006	21.2766	73000	0.3722	3.5564
3.2171	21.5682	74000	0.3725	3.5464
3.2422	21.8597	75000	0.3729	3.5411
3.1721	22.1510	76000	0.3720	3.5582
3.1991	22.4425	77000	0.3727	3.5507
3.2398	22.7340	78000	0.3729	3.5431
3.1305	23.0254	79000	0.3722	3.5554
3.1851	23.3169	80000	0.3722	3.5564
3.197	23.6084	81000	3.5617	0.3725
3.2191	23.8999	82000	3.5461	0.3728
3.1566	24.1915	83000	3.5584	0.3724
3.1969	24.4830	84000	3.5514	0.3728
3.2163	24.7745	85000	3.5425	0.3732
3.142	25.0659	86000	3.5598	0.3724
3.1786	25.3574	87000	3.5538	0.3728
3.1995	25.6489	88000	3.5452	0.3733
3.2032	25.9404	89000	3.5370	0.3738
3.1651	26.2318	90000	3.5582	0.3726
3.1855	26.5233	91000	3.5505	0.3731
3.1965	26.8148	92000	3.5430	0.3737
3.1315	27.1061	93000	3.5580	0.3729
3.1529	27.3976	94000	3.5540	0.3731
3.1944	27.6891	95000	3.5449	0.3735
3.2029	27.9806	96000	3.5407	0.3740
3.1261	28.2720	97000	3.5553	0.3731
3.1616	28.5635	98000	3.5457	0.3740
3.1823	28.8550	99000	3.5398	0.3742
3.1149	29.1463	100000	3.5600	0.3729
3.1365	29.4378	101000	3.5508	0.3736
3.1626	29.7294	102000	3.5458	0.3738
3.0949	30.0207	103000	3.5542	0.3735
3.14	30.3122	104000	3.5554	0.3735
3.1466	30.6037	105000	3.5483	0.3739
3.1679	30.8952	106000	3.5419	0.3745
3.1108	31.1866	107000	3.5570	0.3734
3.1211	31.4781	108000	3.5552	0.3737
3.1553	31.7696	109000	3.5458	0.3744

Framework versions

• Transformers 4.55.2
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4