learning_rate: 0.0006
train_batch_size: 16
eval_batch_size: 16
seed: 40817
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
4.8297	0.2915	1000	4.7499	0.2549
4.3407	0.5831	2000	4.2920	0.2987
4.1555	0.8746	3000	4.1048	0.3145
4.0077	1.1662	4000	3.9986	0.3241
3.9334	1.4577	5000	3.9222	0.3305
3.8927	1.7493	6000	3.8632	0.3358
3.7519	2.0408	7000	3.8225	0.3403
3.7643	2.3324	8000	3.7928	0.3431
3.7685	2.6239	9000	3.7653	0.3458
3.7324	2.9155	10000	3.7368	0.3485
3.6454	3.2070	11000	3.7232	0.3503
3.6495	3.4985	12000	3.7056	0.3517
3.6466	3.7901	13000	3.6867	0.3537
3.5468	4.0816	14000	3.6804	0.3549
3.5696	4.3732	15000	3.6699	0.3562
3.5799	4.6647	16000	3.6549	0.3573
3.589	4.9563	17000	3.6415	0.3586
3.5	5.2478	18000	3.6437	0.3593
3.5262	5.5394	19000	3.6316	0.3598
3.543	5.8309	20000	3.6218	0.3610
3.4553	6.1224	21000	3.6255	0.3615
3.4813	6.4140	22000	3.6167	0.3620
3.4949	6.7055	23000	3.6102	0.3629
3.5007	6.9971	24000	3.5975	0.3637
3.4453	7.2886	25000	3.6117	0.3636
3.4573	7.5802	26000	3.6007	0.3642
3.468	7.8717	27000	3.5877	0.3650
3.4009	8.1633	28000	3.5994	0.3646
3.4331	8.4548	29000	3.5912	0.3655
3.4368	8.7464	30000	3.5832	0.3662
3.3406	9.0379	31000	3.5875	0.3660
3.3939	9.3294	32000	3.5873	0.3662
3.4115	9.6210	33000	3.5801	0.3666
3.4242	9.9125	34000	3.5705	0.3674
3.3576	10.2041	35000	3.5811	0.3668
3.384	10.4956	36000	3.5745	0.3675
3.3938	10.7872	37000	3.5664	0.3680
3.2889	11.0787	38000	3.5776	0.3677
3.34	11.3703	39000	3.5715	0.3682
3.3743	11.6618	40000	3.5657	0.3686
3.3752	11.9534	41000	3.5586	0.3692
3.3133	12.2449	42000	3.5710	0.3689
3.3545	12.5364	43000	3.5626	0.3691
3.3537	12.8280	44000	3.5552	0.3696
3.2762	13.1195	45000	3.5729	0.3689
3.305	13.4111	46000	3.5670	0.3693
3.3392	13.7026	47000	3.5555	0.3701
3.3395	13.9942	48000	3.5500	0.3702
3.2782	14.2857	49000	3.5651	0.3699
3.3161	14.5773	50000	3.5574	0.3703
3.331	14.8688	51000	3.5510	0.3705
3.2589	15.1603	52000	3.5672	0.3700
3.2948	15.4519	53000	3.5598	0.3704
3.3086	15.7434	54000	3.5503	0.3711
3.2226	16.0350	55000	3.5612	0.3709
3.2584	16.3265	56000	3.5577	0.3706
3.2886	16.6181	57000	3.5541	0.3712
3.3092	16.9096	58000	3.5435	0.3717
3.2349	17.2012	59000	3.5600	0.3709
3.2662	17.4927	60000	3.5529	0.3714
3.2851	17.7843	61000	3.5456	0.3719
3.209	18.0758	62000	3.5612	0.3712
3.2441	18.3673	63000	3.5558	0.3713
3.2631	18.6589	64000	3.5499	0.3720
3.2744	18.9504	65000	3.5393	0.3724
3.2166	19.2420	66000	3.5609	0.3712
3.246	19.5335	67000	3.5522	0.3719
3.2624	19.8251	68000	3.5455	0.3725
3.1782	20.1166	69000	3.5587	0.3720
3.2205	20.4082	70000	3.5529	0.3721
3.2465	20.6997	71000	3.5456	0.3725
3.2586	20.9913	72000	3.5360	0.3730
3.2025	21.2828	73000	3.5580	0.3722
3.2286	21.5743	74000	3.5502	0.3724
3.2479	21.8659	75000	3.5406	0.3730
3.1793	22.1574	76000	3.5621	0.3721
3.2045	22.4490	77000	3.5501	0.3726
3.243	22.7405	78000	3.5449	0.3729
3.1363	23.0321	79000	3.5557	0.3727
3.1928	23.3236	80000	3.5540	0.3722
3.2047	23.6152	81000	3.5480	0.3731
3.2237	23.9067	82000	3.5423	0.3734
3.151	24.1983	83000	3.5589	0.3724
3.1982	24.4898	84000	3.5500	0.3731
3.2193	24.7813	85000	3.5422	0.3736
3.1411	25.0729	86000	3.5540	0.3729
3.1813	25.3644	87000	3.5534	0.3732
3.1907	25.6560	88000	3.5468	0.3734
3.2121	25.9475	89000	3.5401	0.3739
3.1561	26.2391	90000	3.5555	0.3731
3.1778	26.5306	91000	3.5527	0.3733
3.186	26.8222	92000	3.5417	0.3739

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_carry_40817

exceptions_exp2_swap_0.7_resemble_to_carry_40817

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_carry_40817