exceptions_exp2_swap_0.7_last_to_push_5039

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

• Loss: 3.5793
• Accuracy: 0.3662

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: 5039
• 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
4.808	0.2915	1000	0.2566	4.7399
4.3421	0.5830	2000	0.2993	4.2834
4.1444	0.8745	3000	0.3155	4.0976
4.0018	1.1659	4000	0.3245	3.9943
3.9302	1.4574	5000	0.3313	3.9176
3.8701	1.7489	6000	0.3365	3.8609
3.7414	2.0402	7000	0.3407	3.8190
3.7581	2.3317	8000	0.3438	3.7858
3.7401	2.6233	9000	0.3466	3.7566
3.7345	2.9148	10000	0.3487	3.7316
3.646	3.2061	11000	0.3509	3.7185
3.6457	3.4976	12000	0.3525	3.7005
3.6494	3.7891	13000	0.3543	3.6803
3.5513	4.0805	14000	0.3555	3.6764
3.5712	4.3720	15000	0.3565	3.6646
3.5652	4.6635	16000	0.3578	3.6520
3.5727	4.9550	17000	0.3585	3.6395
3.4975	5.2463	18000	0.3594	3.6414
3.5181	5.5378	19000	0.3602	3.6309
3.5394	5.8293	20000	0.3611	3.6191
3.4473	6.1207	21000	0.3618	3.6240
3.4634	6.4122	22000	0.3621	3.6155
3.4966	6.7037	23000	0.3628	3.6071
3.4898	6.9952	24000	0.3637	3.5951
3.4365	7.2866	25000	0.3636	3.6033
3.4459	7.5781	26000	0.3643	3.5944
3.4577	7.8696	27000	0.3652	3.5877
3.3857	8.1609	28000	0.3651	3.5955
3.4119	8.4524	29000	0.3656	3.5882
3.4199	8.7439	30000	0.3662	3.5793
3.3211	9.0353	31000	0.3664	3.5845
3.3799	9.3268	32000	0.3665	3.5841
3.3861	9.6183	33000	0.3671	3.5752
3.4258	9.9098	34000	0.3679	3.5655
3.3346	10.2011	35000	0.3673	3.5778
3.3768	10.4927	36000	0.3676	3.5711
3.3908	10.7842	37000	0.3682	3.5624
3.2957	11.0755	38000	0.3681	3.5743
3.3414	11.3670	39000	0.3680	3.5714
3.3675	11.6585	40000	0.3685	3.5630
3.373	11.9500	41000	0.3695	3.5541
3.3066	12.2414	42000	0.3690	3.5681
3.3441	12.5329	43000	0.3694	3.5649
3.3471	12.8244	44000	0.3698	3.5537
3.2738	13.1157	45000	0.3690	3.5698
3.3214	13.4072	46000	0.3697	3.5634
3.3399	13.6988	47000	0.3700	3.5547
3.3465	13.9903	48000	0.3710	3.5480
3.2856	14.2816	49000	0.3700	3.5661
3.3087	14.5731	50000	0.3704	3.5559
3.3084	14.8646	51000	0.3706	3.5494
3.255	15.1560	52000	0.3698	3.5644
3.29	15.4475	53000	0.3707	3.5581
3.3065	15.7390	54000	0.3710	3.5503
3.2043	16.0303	55000	0.3705	3.5610
3.2589	16.3218	56000	0.3707	3.5618
3.2869	16.6133	57000	0.3713	3.5539
3.2945	16.9049	58000	0.3717	3.5450
3.2203	17.1962	59000	0.3707	3.5620
3.2626	17.4877	60000	0.3712	3.5546
3.2893	17.7792	61000	0.3718	3.5505
3.1948	18.0705	62000	0.3710	3.5648
3.2341	18.3621	63000	0.3718	3.5538
3.2475	18.6536	64000	0.3719	3.5513
3.2883	18.9451	65000	0.3725	3.5423
3.2009	19.2364	66000	0.3715	3.5601
3.2452	19.5279	67000	0.3719	3.5564
3.2689	19.8194	68000	0.3725	3.5440
3.1962	20.1108	69000	0.3718	3.5547
3.2246	20.4023	70000	0.3721	3.5526
3.2484	20.6938	71000	0.3725	3.5480
3.2648	20.9853	72000	0.3730	3.5375
3.2029	21.2766	73000	0.3720	3.5564
3.2161	21.5682	74000	0.3721	3.5505
3.2343	21.8597	75000	0.3728	3.5432
3.1751	22.1510	76000	0.3720	3.5577
3.2019	22.4425	77000	0.3725	3.5501
3.2193	22.7340	78000	0.3730	3.5453
3.1355	23.0254	79000	0.3726	3.5590
3.1879	23.3169	80000	0.3724	3.5575
3.179	23.6084	81000	3.5616	0.3721
3.208	23.8999	82000	3.5486	0.3728
3.1582	24.1915	83000	3.5644	0.3722
3.1921	24.4830	84000	3.5517	0.3728
3.2206	24.7745	85000	3.5467	0.3735
3.1331	25.0659	86000	3.5616	0.3723
3.1676	25.3574	87000	3.5599	0.3726
3.1953	25.6489	88000	3.5496	0.3733
3.2072	25.9404	89000	3.5419	0.3737
3.1404	26.2318	90000	3.5621	0.3726
3.1779	26.5233	91000	3.5511	0.3733
3.1964	26.8148	92000	3.5473	0.3736
3.1097	27.1061	93000	3.5605	0.3731
3.1633	27.3976	94000	3.5554	0.3730
3.1803	27.6891	95000	3.5472	0.3735
3.184	27.9806	96000	3.5425	0.3739
3.1397	28.2720	97000	3.5601	0.3732
3.1659	28.5635	98000	3.5515	0.3736
3.1839	28.8550	99000	3.5437	0.3740
3.1123	29.1463	100000	3.5621	0.3730

Framework versions

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