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
4.8339	0.2915	1000	4.7491	0.2550
4.3432	0.5830	2000	4.2854	0.2986
4.1502	0.8744	3000	4.1021	0.3146
3.9879	1.1659	4000	3.9960	0.3238
3.937	1.4573	5000	3.9217	0.3305
3.881	1.7488	6000	3.8612	0.3360
3.7642	2.0402	7000	3.8185	0.3404
3.754	2.3317	8000	3.7890	0.3435
3.726	2.6232	9000	3.7587	0.3463
3.741	2.9147	10000	3.7343	0.3482
3.6372	3.2061	11000	3.7199	0.3507
3.6435	3.4976	12000	3.7006	0.3523
3.6556	3.7890	13000	3.6828	0.3541
3.5445	4.0804	14000	3.6766	0.3550
3.5822	4.3719	15000	3.6657	0.3559
3.5808	4.6634	16000	3.6501	0.3575
3.5743	4.9549	17000	3.6411	0.3586
3.5086	5.2463	18000	3.6438	0.3588
3.5221	5.5378	19000	3.6306	0.3598
3.5426	5.8293	20000	3.6203	0.3609
3.4418	6.1207	21000	3.6252	0.3613
3.4786	6.4121	22000	3.6181	0.3620
3.4878	6.7036	23000	3.6069	0.3627
3.4978	6.9951	24000	3.5989	0.3634
3.4395	7.2865	25000	3.6066	0.3635
3.4558	7.5780	26000	3.5966	0.3645
3.4579	7.8695	27000	3.5866	0.3649
3.3924	8.1609	28000	3.5977	0.3649
3.4084	8.4524	29000	3.5923	0.3651
3.4383	8.7438	30000	3.5805	0.3658
3.3292	9.0353	31000	3.5850	0.3660
3.3831	9.3267	32000	3.5866	0.3660
3.3934	9.6182	33000	3.5752	0.3669
3.4109	9.9097	34000	3.5707	0.3672
3.3428	10.2011	35000	3.5808	0.3669
3.3606	10.4926	36000	3.5775	0.3673
3.3946	10.7841	37000	3.5680	0.3682
3.2964	11.0755	38000	3.5790	0.3676
3.3485	11.3670	39000	3.5732	0.3681
3.3626	11.6584	40000	3.5690	0.3684
3.3696	11.9499	41000	3.5557	0.3694
3.3037	12.2413	42000	3.5703	0.3686
3.3497	12.5328	43000	3.5629	0.3693
3.3502	12.8243	44000	3.5567	0.3697
3.2639	13.1157	45000	3.5686	0.3688
3.3081	13.4072	46000	3.5662	0.3693
3.3386	13.6987	47000	3.5571	0.3696
3.3385	13.9901	48000	3.5484	0.3702
3.2811	14.2816	49000	3.5630	0.3698
3.3164	14.5730	50000	3.5584	0.3698
3.3276	14.8645	51000	3.5515	0.3707
3.2448	15.1559	52000	3.5628	0.3701
3.2932	15.4474	53000	3.5605	0.3701
3.2937	15.7389	54000	3.5498	0.3711
3.2117	16.0303	55000	3.5632	0.3703
3.2603	16.3218	56000	3.5598	0.3706
3.2934	16.6133	57000	3.5505	0.3712
3.3022	16.9047	58000	3.5462	0.3716
3.2375	17.1962	59000	3.5623	0.3707
3.2588	17.4876	60000	3.5560	0.3714
3.2821	17.7791	61000	3.5438	0.3716
3.1896	18.0705	62000	3.5653	0.3713
3.2406	18.3620	63000	3.5603	0.3713
3.2635	18.6535	64000	3.5513	0.3715
3.2701	18.9450	65000	3.5428	0.3725
3.2163	19.2364	66000	3.5582	0.3717
3.2405	19.5279	67000	3.5511	0.3720
3.2554	19.8193	68000	3.5478	0.3721
3.1662	20.1108	69000	3.5609	0.3715
3.2246	20.4022	70000	3.5598	0.3717
3.2392	20.6937	71000	3.5473	0.3723
3.2569	20.9852	72000	3.5428	0.3728
3.2044	21.2766	73000	3.5561	0.3718
3.2311	21.5681	74000	3.5508	0.3724
3.2468	21.8596	75000	3.5424	0.3727
3.1742	22.1510	76000	3.5607	0.3721
3.2095	22.4425	77000	3.5559	0.3719
3.2252	22.7339	78000	3.5484	0.3724
3.133	23.0254	79000	3.5577	0.3724
3.1897	23.3168	80000	3.5593	0.3724
3.2067	23.6083	81000	3.5488	0.3728
3.2239	23.8998	82000	3.5425	0.3731
3.1573	24.1912	83000	3.5604	0.3725
3.1998	24.4827	84000	3.5549	0.3725
3.2108	24.7742	85000	3.5450	0.3733
3.1299	25.0656	86000	3.5647	0.3722
3.1636	25.3571	87000	3.5598	0.3726
3.1826	25.6485	88000	3.5535	0.3731
3.2198	25.9400	89000	3.5437	0.3736
3.1401	26.2314	90000	3.5631	0.3725
3.1722	26.5229	91000	3.5547	0.3729
3.1945	26.8144	92000	3.5441	0.3735
3.1136	27.1058	93000	3.5593	0.3729
3.1688	27.3973	94000	3.5625	0.3728
3.1743	27.6888	95000	3.5505	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_last_to_push_1032

exceptions_exp2_swap_last_to_push_1032

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

craa/
exceptions_exp2_swap_last_to_push_1032