exceptions_exp2_swap_0.3_cost_to_carry_3591

This model is a fine-tuned version of on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 3.5646
• Accuracy: 0.3686

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 0.0006
• train_batch_size: 16
• eval_batch_size: 16
• seed: 3591
• 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
4.8325	0.2915	1000	0.2536	4.7556
4.3322	0.5831	2000	0.2984	4.2877
4.1494	0.8746	3000	0.3143	4.1042
3.9924	1.1662	4000	0.3237	3.9976
3.9382	1.4577	5000	0.3306	3.9242
3.8834	1.7493	6000	0.3358	3.8639
3.7502	2.0408	7000	0.3400	3.8238
3.7663	2.3324	8000	0.3434	3.7890
3.7385	2.6239	9000	0.3456	3.7604
3.7394	2.9155	10000	0.3485	3.7344
3.6429	3.2070	11000	0.3501	3.7232
3.6462	3.4985	12000	0.3521	3.7042
3.6394	3.7901	13000	0.3536	3.6854
3.5434	4.0816	14000	0.3550	3.6766
3.5706	4.3732	15000	0.3561	3.6678
3.5799	4.6647	16000	0.3573	3.6538
3.5828	4.9563	17000	0.3586	3.6394
3.5236	5.2478	18000	0.3587	3.6434
3.5335	5.5394	19000	0.3600	3.6322
3.539	5.8309	20000	0.3608	3.6219
3.4416	6.1224	21000	0.3613	3.6257
3.485	6.4140	22000	0.3615	3.6173
3.5007	6.7055	23000	0.3627	3.6065
3.4915	6.9971	24000	0.3633	3.5983
3.4286	7.2886	25000	0.3635	3.6058
3.4512	7.5802	26000	0.3639	3.5998
3.459	7.8717	27000	0.3648	3.5882
3.4008	8.1633	28000	0.3650	3.5963
3.4198	8.4548	29000	0.3654	3.5887
3.4241	8.7464	30000	0.3660	3.5794
3.3219	9.0379	31000	0.3658	3.5886
3.3886	9.3294	32000	0.3662	3.5851
3.4079	9.6210	33000	0.3668	3.5773
3.411	9.9125	34000	0.3672	3.5695
3.3508	10.2041	35000	0.3671	3.5802
3.3789	10.4956	36000	0.3673	3.5777
3.3985	10.7872	37000	0.3678	3.5659
3.3045	11.0787	38000	0.3676	3.5789
3.3475	11.3703	39000	0.3680	3.5722
3.3657	11.6618	40000	0.3686	3.5646
3.3842	11.9534	41000	0.3693	3.5569
3.3177	12.2449	42000	0.3685	3.5733
3.3409	12.5364	43000	0.3691	3.5632
3.3609	12.8280	44000	0.3697	3.5544
3.2721	13.1195	45000	0.3689	3.5692
3.3129	13.4111	46000	0.3696	3.5616
3.3339	13.7026	47000	0.3698	3.5560
3.3547	13.9942	48000	0.3703	3.5481
3.2931	14.2857	49000	0.3698	3.5630
3.3007	14.5773	50000	0.3703	3.5551
3.3365	14.8688	51000	0.3706	3.5486
3.2529	15.1603	52000	0.3703	3.5602
3.2856	15.4519	53000	0.3704	3.5566
3.3168	15.7434	54000	0.3708	3.5503
3.2138	16.0350	55000	0.3705	3.5617
3.2648	16.3265	56000	0.3708	3.5590
3.2852	16.6181	57000	0.3712	3.5505
3.2964	16.9096	58000	0.3715	3.5424
3.2284	17.2012	59000	0.3711	3.5593
3.2694	17.4927	60000	0.3712	3.5538
3.2848	17.7843	61000	0.3717	3.5449
3.1896	18.0758	62000	0.3707	3.5651
3.2437	18.3673	63000	0.3716	3.5527
3.2646	18.6589	64000	0.3721	3.5488
3.2773	18.9504	65000	0.3726	3.5386
3.2216	19.2420	66000	0.3716	3.5559
3.2507	19.5335	67000	0.3718	3.5523
3.2619	19.8251	68000	0.3724	3.5422
3.1822	20.1166	69000	0.3717	3.5555
3.2251	20.4082	70000	0.3722	3.5507
3.2546	20.6997	71000	0.3725	3.5437
3.2556	20.9913	72000	0.3731	3.5370
3.2062	21.2828	73000	0.3724	3.5510
3.2305	21.5743	74000	0.3725	3.5497
3.2454	21.8659	75000	0.3731	3.5394
3.1779	22.1574	76000	0.3724	3.5573
3.211	22.4490	77000	0.3726	3.5475
3.234	22.7405	78000	0.3729	3.5424
3.1314	23.0321	79000	0.3726	3.5546
3.1801	23.3236	80000	0.3726	3.5540
3.1808	23.6152	81000	3.5595	0.3722
3.2122	23.9067	82000	3.5496	0.3730
3.174	24.1983	83000	3.5594	0.3726
3.197	24.4898	84000	3.5525	0.3730
3.217	24.7813	85000	3.5436	0.3735
3.1305	25.0729	86000	3.5610	0.3728
3.1713	25.3644	87000	3.5550	0.3732
3.1786	25.6560	88000	3.5467	0.3735
3.2141	25.9475	89000	3.5366	0.3742
3.1414	26.2391	90000	3.5570	0.3730
3.1904	26.5306	91000	3.5502	0.3734
3.1951	26.8222	92000	3.5431	0.3738
3.1156	27.1137	93000	3.5601	0.3731
3.1662	27.4052	94000	3.5558	0.3730
3.1762	27.6968	95000	3.5469	0.3736
3.1881	27.9883	96000	3.5415	0.3741
3.1349	28.2799	97000	3.5551	0.3731
3.1703	28.5714	98000	3.5482	0.3737
3.1847	28.8630	99000	3.5425	0.3742
3.1126	29.1545	100000	3.5597	0.3734
3.1553	29.4461	101000	3.5534	0.3736
3.1708	29.7376	102000	3.5452	0.3742
3.0826	30.0292	103000	3.5592	0.3735
3.1414	30.3207	104000	3.5560	0.3736
3.1349	30.6122	105000	3.5504	0.3739
3.1615	30.9038	106000	3.5423	0.3745
3.1044	31.1953	107000	3.5584	0.3739
3.1225	31.4869	108000	3.5538	0.3742
3.1434	31.7784	109000	3.5446	0.3745

Framework versions

• Transformers 4.55.2
• Pytorch 2.8.0+cu128
• Datasets 4.0.0
• Tokenizers 0.21.4