exceptions_exp2_swap_0.7_last_to_hit_1032

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

• Loss: 3.5645
• Accuracy: 0.3684

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: 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
4.8264	0.2915	1000	0.2540	4.7631
4.3456	0.5830	2000	0.2982	4.2904
4.1536	0.8745	3000	0.3145	4.1023
3.9989	1.1659	4000	0.3233	3.9997
3.9556	1.4574	5000	0.3303	3.9253
3.8898	1.7489	6000	0.3357	3.8675
3.7598	2.0402	7000	0.3398	3.8239
3.7633	2.3317	8000	0.3427	3.7955
3.7479	2.6233	9000	0.3453	3.7658
3.7312	2.9148	10000	0.3482	3.7383
3.652	3.2061	11000	0.3499	3.7270
3.6516	3.4976	12000	0.3516	3.7079
3.655	3.7891	13000	0.3532	3.6881
3.5467	4.0805	14000	0.3547	3.6835
3.5772	4.3720	15000	0.3558	3.6701
3.5928	4.6635	16000	0.3569	3.6566
3.591	4.9550	17000	0.3585	3.6416
3.5226	5.2463	18000	0.3587	3.6458
3.531	5.5378	19000	0.3598	3.6327
3.5355	5.8293	20000	0.3606	3.6232
3.4393	6.1207	21000	0.3610	3.6265
3.4902	6.4122	22000	0.3618	3.6220
3.4981	6.7037	23000	0.3623	3.6088
3.5078	6.9952	24000	0.3633	3.6015
3.4379	7.2866	25000	0.3633	3.6110
3.4689	7.5781	26000	0.3640	3.6010
3.4649	7.8696	27000	0.3645	3.5894
3.3924	8.1609	28000	0.3644	3.5999
3.4298	8.4524	29000	0.3650	3.5923
3.4356	8.7439	30000	0.3657	3.5841
3.3413	9.0353	31000	0.3658	3.5889
3.3913	9.3268	32000	0.3658	3.5882
3.4026	9.6183	33000	0.3667	3.5803
3.4305	9.9098	34000	0.3675	3.5687
3.3488	10.2011	35000	0.3667	3.5826
3.3714	10.4927	36000	0.3672	3.5762
3.4075	10.7842	37000	0.3678	3.5671
3.2992	11.0755	38000	0.3677	3.5783
3.363	11.3670	39000	0.3678	3.5723
3.3685	11.6585	40000	0.3684	3.5645
3.3934	11.9500	41000	0.3689	3.5583
3.3171	12.2414	42000	0.3683	3.5734
3.3376	12.5329	43000	0.3688	3.5642
3.3564	12.8244	44000	0.3694	3.5563
3.2743	13.1157	45000	0.3691	3.5710
3.3177	13.4072	46000	0.3691	3.5664
3.3488	13.6988	47000	0.3698	3.5572
3.3578	13.9903	48000	0.3706	3.5490
3.2915	14.2816	49000	0.3697	3.5650
3.3163	14.5731	50000	0.3700	3.5551
3.3346	14.8646	51000	0.3706	3.5476
3.2686	15.1560	52000	0.3702	3.5659
3.2915	15.4475	53000	0.3702	3.5592
3.3269	15.7390	54000	0.3711	3.5483
3.223	16.0303	55000	0.3705	3.5625
3.2716	16.3218	56000	0.3709	3.5593
3.2915	16.6133	57000	0.3713	3.5524
3.3044	16.9049	58000	0.3714	3.5420
3.237	17.1962	59000	0.3707	3.5604
3.2744	17.4877	60000	0.3712	3.5546
3.2833	17.7792	61000	0.3718	3.5483
3.1982	18.0705	62000	0.3709	3.5628
3.243	18.3621	63000	0.3711	3.5550
3.2624	18.6536	64000	0.3717	3.5502
3.2745	18.9451	65000	0.3722	3.5414
3.214	19.2364	66000	0.3715	3.5596
3.2602	19.5279	67000	0.3718	3.5502
3.2724	19.8194	68000	0.3723	3.5425
3.1947	20.1108	69000	0.3716	3.5606
3.2187	20.4023	70000	0.3721	3.5550
3.2584	20.6938	71000	0.3725	3.5461
3.2653	20.9853	72000	0.3726	3.5400
3.2041	21.2766	73000	0.3721	3.5572
3.2204	21.5682	74000	0.3724	3.5469
3.2459	21.8597	75000	0.3728	3.5438
3.1756	22.1510	76000	0.3718	3.5618
3.2035	22.4425	77000	0.3725	3.5523
3.2424	22.7340	78000	0.3729	3.5444
3.1346	23.0254	79000	0.3724	3.5535
3.1904	23.3169	80000	0.3724	3.5584
3.2024	23.6084	81000	3.5618	0.3723
3.2233	23.8999	82000	3.5478	0.3729
3.1613	24.1915	83000	3.5616	0.3723
3.2017	24.4830	84000	3.5518	0.3727
3.2196	24.7745	85000	3.5434	0.3732
3.1465	25.0659	86000	3.5607	0.3724
3.185	25.3574	87000	3.5561	0.3727
3.2043	25.6489	88000	3.5475	0.3733
3.2088	25.9404	89000	3.5401	0.3738
3.1696	26.2318	90000	3.5576	0.3725
3.1896	26.5233	91000	3.5510	0.3731
3.1998	26.8148	92000	3.5430	0.3737
3.1361	27.1061	93000	3.5578	0.3729
3.1578	27.3976	94000	3.5541	0.3732
3.2005	27.6891	95000	3.5459	0.3736
3.2071	27.9806	96000	3.5379	0.3742
3.1321	28.2720	97000	3.5564	0.3730
3.1656	28.5635	98000	3.5456	0.3739
3.1883	28.8550	99000	3.5395	0.3742
3.1201	29.1463	100000	3.5608	0.3730
3.1421	29.4378	101000	3.5554	0.3733
3.1691	29.7294	102000	3.5465	0.3737
3.1002	30.0207	103000	3.5535	0.3736
3.1452	30.3122	104000	3.5560	0.3735
3.1515	30.6037	105000	3.5494	0.3741
3.174	30.8952	106000	3.5431	0.3743
3.1176	31.1866	107000	3.5574	0.3733
3.1255	31.4781	108000	3.5556	0.3738
3.1599	31.7696	109000	3.5442	0.3743
3.0861	32.0609	110000	3.5563	0.3739
3.1318	32.3524	111000	3.5581	0.3735
3.1483	32.6439	112000	3.5473	0.3743
3.1564	32.9355	113000	3.5440	0.3744
3.1023	33.2268	114000	3.5612	0.3738
3.1314	33.5183	115000	3.5516	0.3739
3.1402	33.8098	116000	3.5482	0.3746

Framework versions

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