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	Validation Loss	Accuracy
4.8154	0.2915	1000	4.7468	0.2555
4.3398	0.5831	2000	4.2887	0.2989
4.1459	0.8746	3000	4.0995	0.3145
3.9942	1.1662	4000	3.9970	0.3240
3.9408	1.4577	5000	3.9194	0.3310
3.8834	1.7493	6000	3.8622	0.3360
3.7542	2.0408	7000	3.8225	0.3402
3.7622	2.3324	8000	3.7875	0.3435
3.7387	2.6239	9000	3.7590	0.3464
3.7479	2.9155	10000	3.7325	0.3488
3.6562	3.2070	11000	3.7215	0.3503
3.648	3.4985	12000	3.7012	0.3524
3.652	3.7901	13000	3.6857	0.3537
3.5543	4.0816	14000	3.6762	0.3553
3.5735	4.3732	15000	3.6669	0.3564
3.5839	4.6647	16000	3.6540	0.3575
3.5775	4.9563	17000	3.6385	0.3588
3.5206	5.2478	18000	3.6426	0.3593
3.5208	5.5394	19000	3.6315	0.3602
3.5373	5.8309	20000	3.6199	0.3612
3.4493	6.1224	21000	3.6253	0.3615
3.4733	6.4140	22000	3.6153	0.3622
3.4973	6.7055	23000	3.6064	0.3629
3.495	6.9971	24000	3.5967	0.3638
3.4392	7.2886	25000	3.6064	0.3636
3.4491	7.5802	26000	3.5945	0.3643
3.4575	7.8717	27000	3.5890	0.3652
3.3965	8.1633	28000	3.5952	0.3650
3.4216	8.4548	29000	3.5914	0.3650
3.4314	8.7464	30000	3.5821	0.3657
3.3311	9.0379	31000	3.5840	0.3662
3.3829	9.3294	32000	3.5874	0.3662
3.4194	9.6210	33000	3.5780	0.3669
3.4116	9.9125	34000	3.5707	0.3671
3.3385	10.2041	35000	3.5826	0.3669
3.3877	10.4956	36000	3.5742	0.3672
3.3867	10.7872	37000	3.5651	0.3683
3.2908	11.0787	38000	3.5789	0.3677
3.3411	11.3703	39000	3.5783	0.3681
3.3667	11.6618	40000	3.5668	0.3685
3.3829	11.9534	41000	3.5574	0.3692
3.3135	12.2449	42000	3.5715	0.3686
3.3483	12.5364	43000	3.5633	0.3692
3.3626	12.8280	44000	3.5581	0.3698
3.2807	13.1195	45000	3.5701	0.3692
3.3153	13.4111	46000	3.5640	0.3694
3.3408	13.7026	47000	3.5567	0.3698
3.3432	13.9942	48000	3.5479	0.3708
3.296	14.2857	49000	3.5622	0.3698
3.3051	14.5773	50000	3.5563	0.3705
3.3203	14.8688	51000	3.5499	0.3708
3.2497	15.1603	52000	3.5628	0.3702
3.2881	15.4519	53000	3.5587	0.3705
3.3246	15.7434	54000	3.5496	0.3711
3.2017	16.0350	55000	3.5590	0.3709
3.2606	16.3265	56000	3.5604	0.3709
3.2928	16.6181	57000	3.5540	0.3714
3.2918	16.9096	58000	3.5432	0.3718
3.2278	17.2012	59000	3.5660	0.3710
3.261	17.4927	60000	3.5560	0.3714
3.2845	17.7843	61000	3.5428	0.3721
3.2045	18.0758	62000	3.5605	0.3713
3.25	18.3673	63000	3.5554	0.3715
3.256	18.6589	64000	3.5517	0.3720
3.2858	18.9504	65000	3.5418	0.3725
3.2071	19.2420	66000	3.5585	0.3715
3.2309	19.5335	67000	3.5523	0.3719
3.269	19.8251	68000	3.5471	0.3723
3.2104	20.1166	69000	3.5597	0.3718
3.2172	20.4082	70000	3.5510	0.3723
3.2547	20.6997	71000	3.5475	0.3723
3.264	20.9913	72000	3.5382	0.3731
3.2117	21.2828	73000	3.5537	0.3724
3.2199	21.5743	74000	3.5509	0.3725
3.2503	21.8659	75000	3.5386	0.3733
3.1741	22.1574	76000	3.5594	0.3723
3.2084	22.4490	77000	3.5498	0.3728
3.2316	22.7405	78000	3.5434	0.3732
3.131	23.0321	79000	3.5552	0.3728
3.1765	23.3236	80000	3.5569	0.3727
3.2131	23.6152	81000	3.5461	0.3731
3.2304	23.9067	82000	3.5406	0.3734
3.166	24.1983	83000	3.5578	0.3728
3.1946	24.4898	84000	3.5527	0.3730
3.2078	24.7813	85000	3.5396	0.3739
3.1337	25.0729	86000	3.5544	0.3729
3.1774	25.3644	87000	3.5551	0.3730
3.1918	25.6560	88000	3.5467	0.3736
3.2087	25.9475	89000	3.5416	0.3738
3.1479	26.2391	90000	3.5586	0.3731
3.1807	26.5306	91000	3.5498	0.3734
3.1987	26.8222	92000	3.5458	0.3737

Framework versions

Transformers 4.55.2
Pytorch 2.8.0+cu128
Datasets 4.0.0
Tokenizers 0.21.4

craa/exceptions_exp2_swap_0.7_resemble_to_hit_1032

exceptions_exp2_swap_0.7_resemble_to_hit_1032

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

craa/
exceptions_exp2_swap_0.7_resemble_to_hit_1032