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	Accuracy	Validation Loss
4.8496	0.2915	1000	0.2521	4.7696
4.3538	0.5831	2000	0.2984	4.2922
4.1511	0.8746	3000	0.3142	4.1058
4.0148	1.1662	4000	0.3239	3.9965
3.9338	1.4577	5000	0.3312	3.9220
3.8851	1.7493	6000	0.3358	3.8631
3.7683	2.0408	7000	0.3401	3.8218
3.7601	2.3324	8000	0.3431	3.7933
3.7486	2.6239	9000	0.3459	3.7603
3.7265	2.9155	10000	0.3483	3.7352
3.6403	3.2070	11000	0.3505	3.7237
3.661	3.4985	12000	0.3521	3.7054
3.6578	3.7901	13000	0.3534	3.6875
3.5509	4.0816	14000	0.3548	3.6793
3.5819	4.3732	15000	0.3560	3.6683
3.5821	4.6647	16000	0.3572	3.6550
3.5663	4.9563	17000	0.3581	3.6446
3.5166	5.2478	18000	0.3589	3.6430
3.5411	5.5394	19000	0.3599	3.6336
3.534	5.8309	20000	0.3610	3.6227
3.4542	6.1224	21000	0.3611	3.6269
3.4911	6.4140	22000	0.3619	3.6180
3.5021	6.7055	23000	0.3628	3.6087
3.5138	6.9971	24000	0.3634	3.5984
3.4399	7.2886	25000	0.3633	3.6074
3.4766	7.5802	26000	0.3640	3.6005
3.47	7.8717	27000	0.3646	3.5890
3.3861	8.1633	28000	0.3647	3.6009
3.4296	8.4548	29000	0.3650	3.5930
3.4542	8.7464	30000	0.3657	3.5820
3.34	9.0379	31000	0.3656	3.5907
3.3813	9.3294	32000	0.3662	3.5873
3.4204	9.6210	33000	0.3668	3.5781
3.4332	9.9125	34000	0.3673	3.5695
3.358	10.2041	35000	0.3667	3.5827
3.3841	10.4956	36000	0.3673	3.5766
3.3953	10.7872	37000	0.3680	3.5661
3.3095	11.0787	38000	0.3677	3.5803
3.3484	11.3703	39000	0.3679	3.5738
3.3807	11.6618	40000	0.3685	3.5648
3.3826	11.9534	41000	0.3690	3.5599
3.301	12.2449	42000	0.3687	3.5735
3.3463	12.5364	43000	0.3689	3.5653
3.361	12.8280	44000	0.3694	3.5575
3.2808	13.1195	45000	0.3688	3.5714
3.3277	13.4111	46000	0.3694	3.5636
3.332	13.7026	47000	0.3700	3.5585
3.3478	13.9942	48000	0.3705	3.5460
3.2935	14.2857	49000	0.3698	3.5619
3.3178	14.5773	50000	0.3702	3.5560
3.3417	14.8688	51000	0.3705	3.5489
3.2587	15.1603	52000	0.3699	3.5665
3.2861	15.4519	53000	0.3706	3.5588
3.3157	15.7434	54000	0.3708	3.5517
3.2147	16.0350	55000	0.3708	3.5603
3.2652	16.3265	56000	0.3707	3.5600
3.287	16.6181	57000	0.3709	3.5551
3.3072	16.9096	58000	0.3714	3.5454
3.2329	17.2012	59000	0.3709	3.5584
3.2811	17.4927	60000	0.3712	3.5536
3.291	17.7843	61000	0.3716	3.5486
3.2135	18.0758	62000	0.3710	3.5614
3.2485	18.3673	63000	0.3715	3.5551
3.2739	18.6589	64000	0.3718	3.5492
3.2757	18.9504	65000	0.3724	3.5397
3.2183	19.2420	66000	0.3711	3.5601
3.2446	19.5335	67000	0.3718	3.5514
3.2819	19.8251	68000	0.3723	3.5408
3.1849	20.1166	69000	0.3720	3.5598
3.2363	20.4082	70000	0.3716	3.5544
3.255	20.6997	71000	0.3724	3.5463
3.2608	20.9913	72000	0.3729	3.5402
3.2107	21.2828	73000	0.3719	3.5586
3.241	21.5743	74000	0.3725	3.5470
3.2468	21.8659	75000	0.3730	3.5394
3.1848	22.1574	76000	0.3720	3.5601
3.2153	22.4490	77000	0.3725	3.5524
3.2333	22.7405	78000	0.3729	3.5445
3.1473	23.0321	79000	0.3724	3.5561
3.1939	23.3236	80000	0.3727	3.5533
3.1872	23.6152	81000	3.5600	0.3720
3.2145	23.9067	82000	3.5495	0.3730
3.1711	24.1983	83000	3.5614	0.3721
3.2094	24.4898	84000	3.5529	0.3729
3.2298	24.7813	85000	3.5435	0.3733
3.1346	25.0729	86000	3.5546	0.3729
3.1677	25.3644	87000	3.5563	0.3728
3.1998	25.6560	88000	3.5452	0.3731
3.2121	25.9475	89000	3.5416	0.3736
3.1529	26.2391	90000	3.5602	0.3728
3.1839	26.5306	91000	3.5498	0.3734
3.2028	26.8222	92000	3.5407	0.3738
3.1244	27.1137	93000	3.5571	0.3730
3.1632	27.4052	94000	3.5507	0.3734
3.1768	27.6968	95000	3.5447	0.3736
3.1984	27.9883	96000	3.5422	0.3741
3.14	28.2799	97000	3.5594	0.3730
3.1614	28.5714	98000	3.5521	0.3733
3.1733	28.8630	99000	3.5450	0.3738
3.1134	29.1545	100000	3.5620	0.3730

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_push_1032

exceptions_exp2_swap_0.7_resemble_to_push_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_push_1032