learning_rate: 0.0006
train_batch_size: 16
eval_batch_size: 16
seed: 2128
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	Validation Loss	Accuracy
4.8157	0.2915	1000	4.7660	0.2535
4.3392	0.5830	2000	4.2830	0.2994
4.1551	0.8745	3000	4.1019	0.3143
3.9993	1.1659	4000	3.9990	0.3237
3.9561	1.4574	5000	3.9265	0.3302
3.8858	1.7488	6000	3.8663	0.3356
3.7502	2.0402	7000	3.8235	0.3403
3.7641	2.3317	8000	3.7918	0.3432
3.7433	2.6232	9000	3.7616	0.3458
3.7296	2.9147	10000	3.7356	0.3482
3.6443	3.2061	11000	3.7219	0.3502
3.6549	3.4976	12000	3.7028	0.3521
3.649	3.7891	13000	3.6838	0.3537
3.5526	4.0805	14000	3.6787	0.3551
3.5829	4.3719	15000	3.6681	0.3559
3.5695	4.6634	16000	3.6525	0.3573
3.5887	4.9549	17000	3.6409	0.3583
3.496	5.2463	18000	3.6424	0.3588
3.5339	5.5378	19000	3.6322	0.3600
3.5248	5.8293	20000	3.6197	0.3612
3.4491	6.1207	21000	3.6261	0.3613
3.4867	6.4122	22000	3.6175	0.3619
3.4819	6.7037	23000	3.6061	0.3628
3.5039	6.9952	24000	3.5966	0.3634
3.4283	7.2865	25000	3.6067	0.3635
3.4657	7.5780	26000	3.5963	0.3641
3.4601	7.8695	27000	3.5862	0.3648
3.3878	8.1609	28000	3.5983	0.3646
3.433	8.4524	29000	3.5891	0.3652
3.4348	8.7439	30000	3.5821	0.3659
3.3442	9.0353	31000	3.5863	0.3661
3.3836	9.3268	32000	3.5859	0.3662
3.4156	9.6183	33000	3.5774	0.3668
3.4167	9.9098	34000	3.5674	0.3676
3.3384	10.2011	35000	3.5799	0.3675
3.3625	10.4926	36000	3.5744	0.3675
3.3805	10.7841	37000	3.5632	0.3683
3.2897	11.0755	38000	3.5753	0.3679
3.3385	11.3670	39000	3.5727	0.3678
3.3726	11.6585	40000	3.5638	0.3688
3.3756	11.9500	41000	3.5553	0.3693
3.312	12.2414	42000	3.5691	0.3686
3.351	12.5329	43000	3.5635	0.3691
3.3496	12.8243	44000	3.5533	0.3698
3.2693	13.1157	45000	3.5689	0.3692
3.3063	13.4072	46000	3.5634	0.3694
3.3365	13.6987	47000	3.5573	0.3698
3.3392	13.9902	48000	3.5472	0.3707
3.2769	14.2816	49000	3.5606	0.3701
3.3022	14.5731	50000	3.5554	0.3704
3.3385	14.8646	51000	3.5451	0.3710
3.2523	15.1559	52000	3.5655	0.3702
3.2825	15.4474	53000	3.5553	0.3711
3.3105	15.7389	54000	3.5495	0.3713
3.2081	16.0303	55000	3.5590	0.3709
3.263	16.3218	56000	3.5583	0.3710
3.2883	16.6133	57000	3.5518	0.3710
3.3067	16.9048	58000	3.5430	0.3718
3.2422	17.1962	59000	3.5577	0.3710
3.262	17.4877	60000	3.5479	0.3713
3.2824	17.7792	61000	3.5449	0.3720
3.2024	18.0705	62000	3.5552	0.3715
3.241	18.3620	63000	3.5538	0.3714
3.2603	18.6535	64000	3.5461	0.3721
3.285	18.9450	65000	3.5365	0.3725
3.211	19.2364	66000	3.5558	0.3719
3.2489	19.5279	67000	3.5486	0.3723
3.2722	19.8194	68000	3.5425	0.3726
3.1891	20.1108	69000	3.5586	0.3718
3.2327	20.4023	70000	3.5524	0.3721
3.2557	20.6938	71000	3.5433	0.3727
3.2644	20.9853	72000	3.5353	0.3733
3.2031	21.2766	73000	3.5550	0.3723
3.2293	21.5681	74000	3.5439	0.3729
3.2566	21.8596	75000	3.5382	0.3732
3.1681	22.1510	76000	3.5566	0.3726
3.2126	22.4425	77000	3.5485	0.3729
3.2391	22.7340	78000	3.5423	0.3730
3.1311	23.0254	79000	3.5510	0.3730
3.1876	23.3169	80000	3.5519	0.3726
3.2135	23.6083	81000	3.5463	0.3729
3.2247	23.8998	82000	3.5389	0.3735
3.1656	24.1912	83000	3.5545	0.3730
3.1925	24.4827	84000	3.5481	0.3734
3.2149	24.7742	85000	3.5402	0.3735
3.1314	25.0656	86000	3.5590	0.3730
3.1727	25.3571	87000	3.5492	0.3733
3.1956	25.6486	88000	3.5456	0.3736
3.2213	25.9401	89000	3.5349	0.3741
3.144	26.2314	90000	3.5541	0.3733
3.1896	26.5229	91000	3.5467	0.3735
3.1893	26.8144	92000	3.5407	0.3741
3.1241	27.1058	93000	3.5572	0.3734
3.1633	27.3973	94000	3.5542	0.3733
3.1783	27.6888	95000	3.5423	0.3737
3.2036	27.9803	96000	3.5367	0.3744
3.1337	28.2717	97000	3.5550	0.3736
3.1695	28.5632	98000	3.5487	0.3739
3.1675	28.8547	99000	3.5410	0.3742
3.1202	29.1460	100000	3.5586	0.3734
3.1342	29.4375	101000	3.5531	0.3737
3.1739	29.7290	102000	3.5422	0.3744
3.0872	30.0204	103000	3.5524	0.3742
3.1255	30.3119	104000	3.5528	0.3738
3.1618	30.6034	105000	3.5464	0.3743
3.1524	30.8949	106000	3.5418	0.3747
3.1228	31.1863	107000	3.5586	0.3738
3.1281	31.4778	108000	3.5502	0.3743
3.143	31.7693	109000	3.5428	0.3745

Framework versions

Transformers 4.55.2
Pytorch 2.8.0+cu128
Datasets 4.0.0
Tokenizers 0.21.4

craa/exceptions_exp2_swap_0.3_resemble_to_drop_2128

exceptions_exp2_swap_0.3_resemble_to_drop_2128

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

craa/
exceptions_exp2_swap_0.3_resemble_to_drop_2128