exceptions_exp2_swap_0.7_last_to_drop_2128

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

• Loss: 3.5649
• Accuracy: 0.3685

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
4.835	0.2915	1000	4.7451	0.2552
4.3392	0.5830	2000	4.2907	0.2987
4.1498	0.8745	3000	4.1057	0.3146
4.0068	1.1659	4000	4.0004	0.3239
3.9399	1.4574	5000	3.9233	0.3305
3.8838	1.7489	6000	3.8674	0.3352
3.7669	2.0402	7000	3.8247	0.3398
3.7653	2.3317	8000	3.7941	0.3429
3.7487	2.6233	9000	3.7636	0.3456
3.7258	2.9148	10000	3.7388	0.3482
3.6525	3.2061	11000	3.7265	0.3497
3.66	3.4976	12000	3.7072	0.3515
3.6561	3.7891	13000	3.6887	0.3532
3.5433	4.0805	14000	3.6818	0.3547
3.5692	4.3720	15000	3.6704	0.3557
3.6003	4.6635	16000	3.6583	0.3569
3.5864	4.9550	17000	3.6432	0.3585
3.501	5.2463	18000	3.6454	0.3589
3.5274	5.5378	19000	3.6352	0.3597
3.5393	5.8293	20000	3.6224	0.3608
3.4556	6.1207	21000	3.6273	0.3613
3.4698	6.4122	22000	3.6174	0.3619
3.4991	6.7037	23000	3.6070	0.3626
3.5032	6.9952	24000	3.5986	0.3636
3.4559	7.2866	25000	3.6083	0.3632
3.4554	7.5781	26000	3.5996	0.3642
3.46	7.8696	27000	3.5902	0.3647
3.39	8.1609	28000	3.5981	0.3645
3.4215	8.4524	29000	3.5932	0.3651
3.431	8.7439	30000	3.5843	0.3657
3.3411	9.0353	31000	3.5885	0.3659
3.374	9.3268	32000	3.5844	0.3661
3.4097	9.6183	33000	3.5762	0.3664
3.416	9.9098	34000	3.5697	0.3673
3.3498	10.2011	35000	3.5812	0.3668
3.3851	10.4927	36000	3.5751	0.3672
3.3935	10.7842	37000	3.5671	0.3680
3.2982	11.0755	38000	3.5750	0.3677
3.3398	11.3670	39000	3.5737	0.3676
3.3823	11.6585	40000	3.5649	0.3685
3.3804	11.9500	41000	3.5576	0.3691
3.3093	12.2414	42000	3.5698	0.3687
3.3451	12.5329	43000	3.5659	0.3692
3.3521	12.8244	44000	3.5556	0.3695
3.2693	13.1157	45000	3.5709	0.3692
3.3104	13.4072	46000	3.5658	0.3694
3.3526	13.6988	47000	3.5563	0.3701
3.3386	13.9903	48000	3.5494	0.3703
3.2844	14.2816	49000	3.5644	0.3700
3.3194	14.5731	50000	3.5596	0.3702
3.3342	14.8646	51000	3.5490	0.3706
3.2611	15.1560	52000	3.5651	0.3700
3.2973	15.4475	53000	3.5577	0.3703
3.3001	15.7390	54000	3.5516	0.3709
3.214	16.0303	55000	3.5626	0.3704
3.2651	16.3218	56000	3.5581	0.3707
3.2914	16.6133	57000	3.5506	0.3712
3.3116	16.9049	58000	3.5449	0.3717
3.2426	17.1962	59000	3.5620	0.3708
3.2743	17.4877	60000	3.5532	0.3714
3.2881	17.7792	61000	3.5483	0.3717
3.1966	18.0705	62000	3.5614	0.3712
3.2479	18.3621	63000	3.5575	0.3714
3.2603	18.6536	64000	3.5470	0.3721
3.2901	18.9451	65000	3.5428	0.3723
3.2231	19.2364	66000	3.5566	0.3714
3.2365	19.5279	67000	3.5505	0.3719
3.2645	19.8194	68000	3.5435	0.3726
3.1919	20.1108	69000	3.5586	0.3717
3.2132	20.4023	70000	3.5564	0.3720
3.2433	20.6938	71000	3.5453	0.3723
3.2711	20.9853	72000	3.5393	0.3730
3.2121	21.2766	73000	3.5550	0.3722
3.23	21.5682	74000	3.5468	0.3728
3.2567	21.8597	75000	3.5400	0.3729
3.1766	22.1510	76000	3.5583	0.3721
3.2005	22.4425	77000	3.5535	0.3724
3.2244	22.7340	78000	3.5421	0.3731
3.1494	23.0254	79000	3.5552	0.3724
3.1781	23.3169	80000	3.5548	0.3725
3.2131	23.6084	81000	3.5438	0.3730
3.2351	23.8999	82000	3.5386	0.3735
3.1522	24.1912	83000	3.5568	0.3727
3.1877	24.4827	84000	3.5524	0.3731
3.2152	24.7743	85000	3.5430	0.3733
3.1475	25.0656	86000	3.5565	0.3728
3.1773	25.3571	87000	3.5514	0.3730
3.1967	25.6486	88000	3.5456	0.3735
3.2104	25.9401	89000	3.5381	0.3741
3.169	26.2315	90000	3.5567	0.3728
3.1796	26.5230	91000	3.5500	0.3732
3.2059	26.8145	92000	3.5410	0.3739
3.1235	27.1058	93000	3.5583	0.3733
3.1552	27.3973	94000	3.5548	0.3732
3.1796	27.6888	95000	3.5437	0.3738
3.2061	27.9804	96000	3.5371	0.3742
3.1366	28.2717	97000	3.5549	0.3733
3.1694	28.5632	98000	3.5494	0.3735
3.1718	28.8547	99000	3.5415	0.3741
3.1236	29.1460	100000	3.5575	0.3734
3.1376	29.4376	101000	3.5544	0.3736
3.1525	29.7291	102000	3.5459	0.3741
3.0821	30.0204	103000	3.5547	0.3738
3.1334	30.3119	104000	3.5565	0.3736
3.1443	30.6034	105000	3.5472	0.3740
3.1652	30.8949	106000	3.5425	0.3742
3.1073	31.1863	107000	3.5573	0.3740
3.1223	31.4778	108000	3.5525	0.3739
3.1587	31.7693	109000	3.5436	0.3744
3.0771	32.0606	110000	3.5599	0.3738
3.1064	32.3521	111000	3.5584	0.3739
3.1283	32.6437	112000	3.5501	0.3741
3.1471	32.9352	113000	3.5469	0.3745
3.0986	33.2265	114000	3.5578	0.3741
3.1017	33.5180	115000	3.5537	0.3741
3.1393	33.8095	116000	3.5414	0.3750

Framework versions

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