exceptions_exp2_swap_0.7_cost_to_push_2128

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

• Loss: 3.5656
• 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: 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
4.8338	0.2917	1000	4.7557	0.2539
4.3402	0.5834	2000	4.2894	0.2985
4.1517	0.8750	3000	4.1054	0.3145
3.9903	1.1665	4000	3.9977	0.3236
3.9523	1.4582	5000	3.9219	0.3308
3.8827	1.7499	6000	3.8654	0.3359
3.7572	2.0414	7000	3.8223	0.3402
3.7646	2.3331	8000	3.7894	0.3436
3.7509	2.6248	9000	3.7617	0.3459
3.7233	2.9165	10000	3.7339	0.3486
3.6461	3.2080	11000	3.7247	0.3504
3.6611	3.4996	12000	3.7017	0.3521
3.6356	3.7913	13000	3.6864	0.3537
3.5425	4.0828	14000	3.6793	0.3550
3.581	4.3745	15000	3.6677	0.3558
3.5895	4.6662	16000	3.6532	0.3574
3.5848	4.9579	17000	3.6402	0.3585
3.5148	5.2494	18000	3.6451	0.3590
3.5316	5.5411	19000	3.6331	0.3597
3.5259	5.8327	20000	3.6204	0.3609
3.4457	6.1243	21000	3.6269	0.3611
3.4804	6.4159	22000	3.6194	0.3617
3.5013	6.7076	23000	3.6067	0.3627
3.5029	6.9993	24000	3.5982	0.3636
3.4429	7.2908	25000	3.6079	0.3634
3.459	7.5825	26000	3.5971	0.3643
3.4528	7.8742	27000	3.5894	0.3649
3.3868	8.1657	28000	3.5951	0.3650
3.4173	8.4574	29000	3.5888	0.3654
3.4277	8.7490	30000	3.5781	0.3658
3.3269	9.0405	31000	3.5897	0.3658
3.3931	9.3322	32000	3.5842	0.3665
3.4108	9.6239	33000	3.5775	0.3672
3.4116	9.9156	34000	3.5693	0.3673
3.3557	10.2071	35000	3.5818	0.3670
3.3691	10.4988	36000	3.5733	0.3676
3.3801	10.7905	37000	3.5659	0.3682
3.3021	11.0820	38000	3.5760	0.3679
3.3552	11.3736	39000	3.5738	0.3680
3.3689	11.6653	40000	3.5656	0.3686
3.374	11.9570	41000	3.5593	0.3690
3.3049	12.2485	42000	3.5733	0.3684
3.3449	12.5402	43000	3.5647	0.3689
3.3538	12.8319	44000	3.5564	0.3695
3.2757	13.1234	45000	3.5710	0.3691
3.3181	13.4151	46000	3.5626	0.3695
3.3385	13.7067	47000	3.5551	0.3700
3.343	13.9984	48000	3.5502	0.3704
3.2808	14.2899	49000	3.5654	0.3698
3.3103	14.5816	50000	3.5559	0.3702
3.3366	14.8733	51000	3.5506	0.3707
3.256	15.1648	52000	3.5647	0.3702
3.2911	15.4565	53000	3.5606	0.3705
3.3124	15.7482	54000	3.5516	0.3709
3.2155	16.0397	55000	3.5626	0.3703
3.2553	16.3313	56000	3.5593	0.3708
3.2917	16.6230	57000	3.5537	0.3713
3.2971	16.9147	58000	3.5450	0.3715
3.2404	17.2062	59000	3.5573	0.3709
3.2517	17.4979	60000	3.5582	0.3710
3.3012	17.7896	61000	3.5440	0.3718
3.2051	18.0811	62000	3.5600	0.3711
3.2447	18.3728	63000	3.5573	0.3712
3.2502	18.6644	64000	3.5509	0.3720
3.2732	18.9561	65000	3.5395	0.3724
3.2232	19.2476	66000	3.5611	0.3715
3.2381	19.5393	67000	3.5524	0.3720
3.2685	19.8310	68000	3.5455	0.3724
3.1802	20.1225	69000	3.5588	0.3716
3.2236	20.4142	70000	3.5541	0.3721
3.2489	20.7059	71000	3.5452	0.3724
3.2577	20.9975	72000	3.5435	0.3728
3.1989	21.2891	73000	3.5603	0.3718
3.2185	21.5807	74000	3.5502	0.3726
3.2527	21.8724	75000	3.5416	0.3727
3.1685	22.1639	76000	3.5620	0.3721
3.2139	22.4556	77000	3.5563	0.3722
3.2165	22.7473	78000	3.5456	0.3730
3.1322	23.0388	79000	3.5607	0.3722
3.1914	23.3305	80000	3.5560	0.3723
3.2002	23.6222	81000	3.5512	0.3724
3.2265	23.9138	82000	3.5439	0.3731
3.157	24.2053	83000	3.5597	0.3724
3.1921	24.4970	84000	3.5508	0.3728
3.2222	24.7887	85000	3.5455	0.3733

Framework versions

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