exceptions_exp2_swap_0.7_cost_to_carry_2128

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

• Loss: 3.5809
• Accuracy: 0.3660

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
4.8196	0.2917	1000	4.7522	0.2546
4.352	0.5834	2000	4.2930	0.2985
4.1567	0.8750	3000	4.1055	0.3145
3.981	1.1665	4000	3.9968	0.3246
3.9322	1.4582	5000	3.9242	0.3307
3.8923	1.7499	6000	3.8637	0.3360
3.7477	2.0414	7000	3.8228	0.3401
3.7587	2.3331	8000	3.7924	0.3433
3.7414	2.6248	9000	3.7628	0.3457
3.7316	2.9165	10000	3.7354	0.3483
3.6378	3.2080	11000	3.7241	0.3500
3.6542	3.4996	12000	3.7076	0.3517
3.6496	3.7913	13000	3.6888	0.3536
3.5471	4.0828	14000	3.6780	0.3547
3.5676	4.3745	15000	3.6662	0.3560
3.5909	4.6662	16000	3.6547	0.3571
3.5768	4.9579	17000	3.6410	0.3585
3.5331	5.2494	18000	3.6406	0.3590
3.5281	5.5411	19000	3.6345	0.3599
3.5368	5.8327	20000	3.6194	0.3608
3.445	6.1243	21000	3.6234	0.3614
3.4715	6.4159	22000	3.6190	0.3619
3.4999	6.7076	23000	3.6075	0.3626
3.5098	6.9993	24000	3.5987	0.3633
3.4423	7.2908	25000	3.6048	0.3636
3.4451	7.5825	26000	3.5982	0.3642
3.4745	7.8742	27000	3.5886	0.3646
3.3882	8.1657	28000	3.5982	0.3647
3.4272	8.4574	29000	3.5897	0.3654
3.4265	8.7490	30000	3.5809	0.3660
3.3367	9.0405	31000	3.5872	0.3660
3.3911	9.3322	32000	3.5857	0.3662
3.4057	9.6239	33000	3.5775	0.3669
3.421	9.9156	34000	3.5705	0.3676
3.345	10.2071	35000	3.5826	0.3669
3.3668	10.4988	36000	3.5744	0.3676
3.3984	10.7905	37000	3.5642	0.3680
3.299	11.0820	38000	3.5760	0.3684
3.3509	11.3736	39000	3.5711	0.3679
3.3598	11.6653	40000	3.5647	0.3684
3.3698	11.9570	41000	3.5552	0.3693
3.3069	12.2485	42000	3.5723	0.3687
3.335	12.5402	43000	3.5615	0.3692
3.3598	12.8319	44000	3.5557	0.3696
3.2664	13.1234	45000	3.5684	0.3693
3.3075	13.4151	46000	3.5627	0.3693
3.3367	13.7067	47000	3.5561	0.3701
3.3365	13.9984	48000	3.5441	0.3706
3.2835	14.2899	49000	3.5642	0.3698
3.3114	14.5816	50000	3.5530	0.3703
3.3349	14.8733	51000	3.5494	0.3708
3.2604	15.1648	52000	3.5625	0.3702
3.2767	15.4565	53000	3.5601	0.3702
3.3059	15.7482	54000	3.5493	0.3710
3.215	16.0397	55000	3.5603	0.3706
3.2628	16.3313	56000	3.5587	0.3708
3.2891	16.6230	57000	3.5499	0.3712
3.3014	16.9147	58000	3.5451	0.3717
3.2314	17.2062	59000	3.5585	0.3711
3.2655	17.4979	60000	3.5539	0.3710
3.2824	17.7896	61000	3.5434	0.3719
3.2046	18.0811	62000	3.5582	0.3713
3.241	18.3728	63000	3.5560	0.3714
3.2721	18.6644	64000	3.5489	0.3717
3.2826	18.9561	65000	3.5421	0.3723
3.2107	19.2476	66000	3.5551	0.3714
3.2404	19.5393	67000	3.5510	0.3719
3.2669	19.8310	68000	3.5407	0.3727
3.1954	20.1225	69000	3.5586	0.3714
3.232	20.4142	70000	3.5542	0.3719
3.2662	20.7059	71000	3.5453	0.3721
3.254	20.9975	72000	3.5379	0.3731
3.1993	21.2891	73000	3.5558	0.3725
3.2249	21.5807	74000	3.5483	0.3721
3.2397	21.8724	75000	3.5421	0.3731
3.1778	22.1639	76000	3.5575	0.3724
3.2168	22.4556	77000	3.5507	0.3725
3.2322	22.7473	78000	3.5447	0.3729
3.1333	23.0388	79000	3.5584	0.3722
3.1918	23.3305	80000	3.5544	0.3725
3.2181	23.6222	81000	3.5488	0.3728
3.22	23.9138	82000	3.5395	0.3734
3.1686	24.2053	83000	3.5583	0.3725
3.1963	24.4970	84000	3.5507	0.3731
3.2088	24.7887	85000	3.5448	0.3731
3.1363	25.0802	86000	3.5595	0.3725
3.1864	25.3719	87000	3.5562	0.3729
3.2064	25.6636	88000	3.5437	0.3736
3.2085	25.9553	89000	3.5407	0.3736
3.1601	26.2468	90000	3.5590	0.3729
3.1765	26.5384	91000	3.5512	0.3733
3.2048	26.8301	92000	3.5444	0.3740

Framework versions

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