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
4.841	0.2915	1000	4.7631	0.2533
4.3505	0.5830	2000	4.2915	0.2985
4.1558	0.8744	3000	4.1098	0.3141
4.0019	1.1659	4000	3.9982	0.3237
3.9366	1.4573	5000	3.9220	0.3304
3.89	1.7488	6000	3.8661	0.3357
3.7455	2.0402	7000	3.8246	0.3399
3.7588	2.3317	8000	3.7918	0.3432
3.7481	2.6232	9000	3.7619	0.3459
3.7295	2.9147	10000	3.7358	0.3484
3.6283	3.2061	11000	3.7214	0.3502
3.655	3.4976	12000	3.7034	0.3518
3.6384	3.7890	13000	3.6854	0.3537
3.5558	4.0804	14000	3.6784	0.3548
3.5619	4.3719	15000	3.6638	0.3562
3.5852	4.6634	16000	3.6534	0.3573
3.5842	4.9549	17000	3.6398	0.3583
3.5154	5.2463	18000	3.6418	0.3590
3.5217	5.5378	19000	3.6315	0.3598
3.5332	5.8293	20000	3.6174	0.3611
3.4521	6.1207	21000	3.6236	0.3613
3.4702	6.4121	22000	3.6167	0.3621
3.49	6.7036	23000	3.6065	0.3628
3.5115	6.9951	24000	3.5965	0.3634
3.4444	7.2865	25000	3.6041	0.3635
3.4648	7.5780	26000	3.5947	0.3641
3.4617	7.8695	27000	3.5867	0.3652
3.4129	8.1609	28000	3.5977	0.3646
3.4211	8.4524	29000	3.5880	0.3654
3.4325	8.7438	30000	3.5805	0.3659
3.3415	9.0353	31000	3.5888	0.3661
3.3921	9.3267	32000	3.5833	0.3661
3.3973	9.6182	33000	3.5749	0.3667
3.4176	9.9097	34000	3.5685	0.3676
3.34	10.2011	35000	3.5788	0.3670
3.3871	10.4926	36000	3.5739	0.3676
3.3913	10.7841	37000	3.5658	0.3682
3.298	11.0755	38000	3.5752	0.3680
3.3436	11.3670	39000	3.5678	0.3686
3.3567	11.6584	40000	3.5602	0.3690
3.3793	11.9499	41000	3.5569	0.3690
3.2949	12.2413	42000	3.5705	0.3687
3.3412	12.5328	43000	3.5638	0.3688
3.3429	12.8243	44000	3.5568	0.3696
3.2734	13.1157	45000	3.5693	0.3692
3.3124	13.4072	46000	3.5636	0.3693
3.3339	13.6987	47000	3.5566	0.3701
3.3414	13.9901	48000	3.5500	0.3704
3.2837	14.2816	49000	3.5617	0.3700
3.3158	14.5730	50000	3.5563	0.3701
3.3285	14.8645	51000	3.5488	0.3709
3.2584	15.1559	52000	3.5613	0.3702
3.2896	15.4474	53000	3.5572	0.3706
3.3147	15.7389	54000	3.5496	0.3710
3.2064	16.0303	55000	3.5601	0.3707
3.2627	16.3218	56000	3.5560	0.3707
3.2846	16.6133	57000	3.5507	0.3713
3.2952	16.9047	58000	3.5435	0.3718
3.2283	17.1962	59000	3.5609	0.3712
3.2687	17.4876	60000	3.5542	0.3712
3.2795	17.7791	61000	3.5451	0.3718
3.2077	18.0705	62000	3.5564	0.3713
3.2501	18.3620	63000	3.5531	0.3717
3.2764	18.6535	64000	3.5490	0.3718
3.2771	18.9450	65000	3.5411	0.3721
3.2105	19.2364	66000	3.5565	0.3717
3.2409	19.5279	67000	3.5516	0.3723
3.2611	19.8193	68000	3.5417	0.3726
3.182	20.1108	69000	3.5592	0.3719
3.2104	20.4022	70000	3.5541	0.3720
3.2474	20.6937	71000	3.5446	0.3725
3.277	20.9852	72000	3.5380	0.3731
3.2051	21.2766	73000	3.5549	0.3723
3.2172	21.5681	74000	3.5492	0.3724
3.225	21.8596	75000	3.5394	0.3729
3.1673	22.1510	76000	3.5577	0.3724
3.2181	22.4425	77000	3.5522	0.3727
3.2302	22.7339	78000	3.5432	0.3731
3.1338	23.0254	79000	3.5547	0.3728
3.1826	23.3168	80000	3.5547	0.3726
3.2116	23.6083	81000	3.5471	0.3729
3.2367	23.8998	82000	3.5386	0.3733
3.1722	24.1912	83000	3.5587	0.3726
3.1771	24.4827	84000	3.5479	0.3730
3.2175	24.7742	85000	3.5410	0.3736
3.1227	25.0656	86000	3.5585	0.3729
3.1775	25.3571	87000	3.5497	0.3731
3.1967	25.6485	88000	3.5460	0.3735
3.2307	25.9400	89000	3.5395	0.3740
3.149	26.2314	90000	3.5590	0.3731
3.1865	26.5229	91000	3.5469	0.3735
3.1781	26.8144	92000	3.5426	0.3736

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_carry_2128

exceptions_exp2_swap_last_to_carry_2128

Model description

Intended uses & limitations

Training and evaluation data

Training procedure

Training hyperparameters

Training results

Framework versions

craa/
exceptions_exp2_swap_last_to_carry_2128