learning_rate: 0.0006
train_batch_size: 16
eval_batch_size: 16
seed: 5039
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
4.8383	0.2915	1000	4.7558	0.2542
4.3502	0.5830	2000	4.2937	0.2978
4.159	0.8745	3000	4.1105	0.3137
4.0024	1.1659	4000	3.9980	0.3235
3.934	1.4574	5000	3.9193	0.3308
3.8803	1.7488	6000	3.8613	0.3361
3.7462	2.0402	7000	3.8212	0.3402
3.7638	2.3317	8000	3.7902	0.3433
3.7417	2.6232	9000	3.7595	0.3458
3.7281	2.9147	10000	3.7337	0.3484
3.6335	3.2061	11000	3.7214	0.3504
3.6506	3.4976	12000	3.7033	0.3519
3.6466	3.7891	13000	3.6854	0.3536
3.5475	4.0805	14000	3.6771	0.3549
3.5705	4.3719	15000	3.6674	0.3561
3.5816	4.6634	16000	3.6531	0.3572
3.581	4.9549	17000	3.6402	0.3582
3.5114	5.2463	18000	3.6434	0.3590
3.5169	5.5378	19000	3.6340	0.3599
3.5455	5.8293	20000	3.6219	0.3607
3.4549	6.1207	21000	3.6238	0.3614
3.4907	6.4122	22000	3.6184	0.3617
3.4978	6.7037	23000	3.6073	0.3626
3.4903	6.9952	24000	3.5979	0.3634
3.4389	7.2865	25000	3.6068	0.3632
3.4533	7.5780	26000	3.5958	0.3640
3.4617	7.8695	27000	3.5867	0.3651
3.396	8.1609	28000	3.5967	0.3647
3.4335	8.4524	29000	3.5920	0.3654
3.4377	8.7439	30000	3.5817	0.3659
3.3391	9.0353	31000	3.5883	0.3658
3.391	9.3268	32000	3.5880	0.3660
3.4011	9.6183	33000	3.5788	0.3668
3.4194	9.9098	34000	3.5697	0.3674
3.3474	10.2011	35000	3.5824	0.3670
3.3782	10.4926	36000	3.5732	0.3673
3.3985	10.7841	37000	3.5691	0.3680
3.3072	11.0755	38000	3.5785	0.3678
3.3449	11.3670	39000	3.5734	0.3679
3.3625	11.6585	40000	3.5655	0.3687
3.3861	11.9500	41000	3.5597	0.3690
3.3226	12.2414	42000	3.5730	0.3682
3.3447	12.5329	43000	3.5662	0.3687
3.3569	12.8243	44000	3.5566	0.3696
3.2827	13.1157	45000	3.5688	0.3691
3.3217	13.4072	46000	3.5644	0.3695
3.3322	13.6987	47000	3.5546	0.3700
3.3484	13.9902	48000	3.5513	0.3704
3.2778	14.2816	49000	3.5679	0.3696
3.3084	14.5731	50000	3.5606	0.3699
3.3434	14.8646	51000	3.5491	0.3704
3.2604	15.1559	52000	3.5671	0.3700
3.2787	15.4474	53000	3.5616	0.3704
3.3124	15.7389	54000	3.5512	0.3709
3.2017	16.0303	55000	3.5634	0.3702
3.2663	16.3218	56000	3.5572	0.3707
3.2794	16.6133	57000	3.5543	0.3710
3.3029	16.9048	58000	3.5459	0.3715
3.2322	17.1962	59000	3.5609	0.3707
3.2661	17.4877	60000	3.5534	0.3709
3.274	17.7792	61000	3.5470	0.3718
3.1897	18.0705	62000	3.5593	0.3709
3.2371	18.3620	63000	3.5588	0.3713
3.2613	18.6535	64000	3.5479	0.3718
3.2801	18.9450	65000	3.5393	0.3721
3.2061	19.2364	66000	3.5584	0.3718
3.2422	19.5279	67000	3.5529	0.3717
3.253	19.8194	68000	3.5425	0.3725
3.1972	20.1108	69000	3.5601	0.3716
3.2244	20.4023	70000	3.5569	0.3718
3.2414	20.6938	71000	3.5487	0.3722
3.2673	20.9853	72000	3.5396	0.3728
3.2181	21.2766	73000	3.5576	0.3716
3.2163	21.5681	74000	3.5498	0.3721
3.2433	21.8596	75000	3.5446	0.3725
3.1705	22.1510	76000	3.5633	0.3717
3.213	22.4425	77000	3.5552	0.3724
3.2327	22.7340	78000	3.5481	0.3725
3.1432	23.0254	79000	3.5570	0.3720
3.1876	23.3169	80000	3.5583	0.3722
3.214	23.6083	81000	3.5502	0.3727
3.2326	23.8998	82000	3.5446	0.3729
3.1754	24.1912	83000	3.5593	0.3723
3.1833	24.4827	84000	3.5556	0.3726
3.2149	24.7742	85000	3.5444	0.3732

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_last_to_push_5039

exceptions_exp2_swap_0.3_last_to_push_5039

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

craa/
exceptions_exp2_swap_0.3_last_to_push_5039