exceptions_exp2_swap_0.7_last_to_hit_3591

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

• Loss: 3.5821
• Accuracy: 0.3658

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: 3591
• 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
4.8343	0.2915	1000	4.7581	0.2535
4.3498	0.5830	2000	4.2966	0.2980
4.1624	0.8745	3000	4.1038	0.3140
3.997	1.1659	4000	3.9944	0.3243
3.9425	1.4574	5000	3.9185	0.3308
3.887	1.7489	6000	3.8622	0.3359
3.7501	2.0402	7000	3.8188	0.3405
3.7459	2.3317	8000	3.7891	0.3435
3.7448	2.6233	9000	3.7615	0.3459
3.7237	2.9148	10000	3.7344	0.3484
3.6481	3.2061	11000	3.7218	0.3501
3.6452	3.4976	12000	3.7041	0.3520
3.642	3.7891	13000	3.6867	0.3538
3.549	4.0805	14000	3.6779	0.3552
3.5693	4.3720	15000	3.6689	0.3561
3.5808	4.6635	16000	3.6516	0.3574
3.584	4.9550	17000	3.6380	0.3586
3.5007	5.2463	18000	3.6432	0.3589
3.5343	5.5378	19000	3.6300	0.3600
3.5526	5.8293	20000	3.6207	0.3611
3.4467	6.1207	21000	3.6223	0.3615
3.4835	6.4122	22000	3.6163	0.3619
3.5067	6.7037	23000	3.6053	0.3629
3.5001	6.9952	24000	3.5976	0.3637
3.4246	7.2866	25000	3.6042	0.3635
3.4532	7.5781	26000	3.5948	0.3643
3.4575	7.8696	27000	3.5887	0.3651
3.3875	8.1609	28000	3.5985	0.3647
3.4277	8.4524	29000	3.5913	0.3653
3.4295	8.7439	30000	3.5821	0.3658
3.3467	9.0353	31000	3.5828	0.3661
3.3686	9.3268	32000	3.5821	0.3665
3.4054	9.6183	33000	3.5749	0.3667
3.4178	9.9098	34000	3.5686	0.3675
3.3577	10.2011	35000	3.5802	0.3671
3.3499	10.4927	36000	3.5715	0.3676
3.3876	10.7842	37000	3.5645	0.3680
3.2882	11.0755	38000	3.5797	0.3680
3.3541	11.3670	39000	3.5723	0.3682
3.3617	11.6585	40000	3.5647	0.3683
3.3819	11.9500	41000	3.5560	0.3691
3.3126	12.2414	42000	3.5694	0.3688
3.343	12.5329	43000	3.5612	0.3693
3.359	12.8244	44000	3.5555	0.3696
3.2734	13.1157	45000	3.5678	0.3692
3.3022	13.4072	46000	3.5651	0.3694
3.3268	13.6988	47000	3.5554	0.3699
3.3538	13.9903	48000	3.5489	0.3704
3.2763	14.2816	49000	3.5672	0.3696
3.3133	14.5731	50000	3.5559	0.3703
3.3245	14.8646	51000	3.5513	0.3707
3.2626	15.1560	52000	3.5613	0.3702
3.2864	15.4475	53000	3.5550	0.3708
3.3143	15.7390	54000	3.5473	0.3713
3.216	16.0303	55000	3.5563	0.3710
3.2606	16.3218	56000	3.5581	0.3707
3.2798	16.6133	57000	3.5556	0.3711
3.2969	16.9049	58000	3.5443	0.3718
3.2429	17.1962	59000	3.5578	0.3711
3.277	17.4877	60000	3.5547	0.3713
3.2883	17.7792	61000	3.5486	0.3716
3.2054	18.0705	62000	3.5622	0.3713
3.2391	18.3621	63000	3.5573	0.3713
3.2634	18.6536	64000	3.5500	0.3719
3.2825	18.9451	65000	3.5422	0.3721
3.2084	19.2364	66000	3.5580	0.3714
3.2496	19.5279	67000	3.5510	0.3718
3.253	19.8194	68000	3.5467	0.3724
3.1912	20.1108	69000	3.5577	0.3716
3.2184	20.4023	70000	3.5538	0.3717
3.246	20.6938	71000	3.5447	0.3727
3.2598	20.9853	72000	3.5412	0.3731
3.2027	21.2766	73000	3.5545	0.3721
3.2273	21.5682	74000	3.5489	0.3725
3.237	21.8597	75000	3.5382	0.3732
3.1768	22.1510	76000	3.5569	0.3720
3.2063	22.4425	77000	3.5523	0.3723
3.2315	22.7340	78000	3.5443	0.3728
3.1344	23.0254	79000	3.5575	0.3726
3.1994	23.3169	80000	3.5542	0.3725
3.2167	23.6084	81000	3.5466	0.3732
3.2272	23.8999	82000	3.5373	0.3737
3.1586	24.1912	83000	3.5572	0.3724
3.2021	24.4827	84000	3.5478	0.3729
3.2065	24.7743	85000	3.5446	0.3733
3.1295	25.0656	86000	3.5572	0.3731
3.1631	25.3571	87000	3.5524	0.3731
3.1941	25.6486	88000	3.5494	0.3731
3.2002	25.9401	89000	3.5395	0.3739
3.1337	26.2315	90000	3.5577	0.3728
3.1693	26.5230	91000	3.5522	0.3733
3.1851	26.8145	92000	3.5438	0.3738
3.1284	27.1058	93000	3.5587	0.3731
3.1539	27.3973	94000	3.5578	0.3733
3.174	27.6888	95000	3.5464	0.3736
3.1806	27.9804	96000	3.5395	0.3742
3.147	28.2717	97000	3.5586	0.3732
3.1586	28.5632	98000	3.5521	0.3735
3.178	28.8547	99000	3.5425	0.3740
3.1163	29.1460	100000	3.5604	0.3732
3.1356	29.4376	101000	3.5523	0.3735
3.1506	29.7291	102000	3.5467	0.3741

Framework versions

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