exceptions_exp2_swap_require_to_drop_1032

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

• Loss: 3.5598
• Accuracy: 0.3695

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	Validation Loss	Accuracy
4.8378	0.2911	1000	4.7581	0.2538
4.3502	0.5822	2000	4.2872	0.2991
4.1547	0.8733	3000	4.0984	0.3149
3.9854	1.1642	4000	3.9956	0.3243
3.9265	1.4553	5000	3.9185	0.3312
3.8796	1.7464	6000	3.8594	0.3369
3.7506	2.0373	7000	3.8184	0.3410
3.7576	2.3284	8000	3.7838	0.3444
3.741	2.6195	9000	3.7572	0.3471
3.7369	2.9106	10000	3.7286	0.3495
3.6397	3.2014	11000	3.7163	0.3514
3.6562	3.4925	12000	3.7000	0.3527
3.6493	3.7837	13000	3.6784	0.3548
3.5378	4.0745	14000	3.6729	0.3561
3.5807	4.3656	15000	3.6635	0.3569
3.5731	4.6567	16000	3.6490	0.3581
3.5826	4.9478	17000	3.6367	0.3595
3.4945	5.2387	18000	3.6378	0.3597
3.5199	5.5298	19000	3.6272	0.3610
3.5301	5.8209	20000	3.6155	0.3618
3.4554	6.1118	21000	3.6214	0.3623
3.4818	6.4029	22000	3.6113	0.3629
3.4865	6.6940	23000	3.6044	0.3637
3.5017	6.9851	24000	3.5933	0.3644
3.4173	7.2760	25000	3.6000	0.3643
3.4484	7.5671	26000	3.5919	0.3652
3.4756	7.8582	27000	3.5855	0.3658
3.3852	8.1490	28000	3.5906	0.3661
3.4073	8.4401	29000	3.5867	0.3664
3.4412	8.7313	30000	3.5758	0.3670
3.3178	9.0221	31000	3.5837	0.3672
3.3672	9.3132	32000	3.5798	0.3675
3.3947	9.6043	33000	3.5707	0.3678
3.4242	9.8954	34000	3.5625	0.3682
3.3366	10.1863	35000	3.5775	0.3681
3.3679	10.4774	36000	3.5706	0.3685
3.3744	10.7685	37000	3.5624	0.3691
3.296	11.0594	38000	3.5696	0.3688
3.3347	11.3505	39000	3.5675	0.3692
3.3613	11.6416	40000	3.5598	0.3695
3.3578	11.9327	41000	3.5512	0.3702
3.3012	12.2236	42000	3.5656	0.3695
3.3291	12.5147	43000	3.5578	0.3702
3.3444	12.8058	44000	3.5475	0.3708
3.2691	13.0966	45000	3.5653	0.3700
3.3059	13.3878	46000	3.5598	0.3705
3.3194	13.6789	47000	3.5492	0.3710
3.3476	13.9700	48000	3.5428	0.3715
3.2779	14.2608	49000	3.5593	0.3708
3.3034	14.5519	50000	3.5528	0.3713
3.3325	14.8430	51000	3.5435	0.3719
3.2446	15.1339	52000	3.5597	0.3711
3.2748	15.4250	53000	3.5535	0.3716
3.2858	15.7161	54000	3.5434	0.3721
3.261	16.0070	55000	3.5549	0.3713
3.2496	16.2981	56000	3.5557	0.3713
3.2842	16.5892	57000	3.5471	0.3722
3.2965	16.8803	58000	3.5371	0.3727
3.23	17.1712	59000	3.5565	0.3720
3.2616	17.4623	60000	3.5502	0.3723
3.2791	17.7534	61000	3.5406	0.3729
3.1841	18.0442	62000	3.5531	0.3724
3.2441	18.3354	63000	3.5499	0.3724
3.2666	18.6265	64000	3.5381	0.3729
3.2659	18.9176	65000	3.5344	0.3734
3.2098	19.2084	66000	3.5527	0.3725
3.2497	19.4995	67000	3.5442	0.3731
3.268	19.7906	68000	3.5366	0.3736
3.1701	20.0815	69000	3.5527	0.3728
3.2122	20.3726	70000	3.5494	0.3731
3.2433	20.6637	71000	3.5404	0.3735
3.2464	20.9548	72000	3.5343	0.3737
3.2004	21.2457	73000	3.5520	0.3733
3.2205	21.5368	74000	3.5444	0.3737
3.2267	21.8279	75000	3.5360	0.3742
3.1646	22.1188	76000	3.5526	0.3734
3.2048	22.4099	77000	3.5466	0.3736
3.2191	22.7010	78000	3.5381	0.3740
3.2266	22.9921	79000	3.5290	0.3747
3.1936	23.2830	80000	3.5487	0.3737
3.204	23.5741	81000	3.5431	0.3741
3.2031	23.8652	82000	3.5350	0.3746
3.148	24.1560	83000	3.5525	0.3735
3.1932	24.4471	84000	3.5494	0.3739
3.2043	24.7382	85000	3.5385	0.3745
3.1214	25.0291	86000	3.5511	0.3739
3.1671	25.3202	87000	3.5468	0.3741
3.1927	25.6113	88000	3.5423	0.3743
3.206	25.9024	89000	3.5358	0.3745
3.1438	26.1933	90000	3.5514	0.3738
3.1742	26.4844	91000	3.5436	0.3743
3.1898	26.7755	92000	3.5390	0.3745
3.1069	27.0664	93000	3.5525	0.3739
3.1602	27.3575	94000	3.5500	0.3740
3.1705	27.6486	95000	3.5426	0.3748
3.1825	27.9397	96000	3.5342	0.3750
3.1199	28.2306	97000	3.5535	0.3743
3.1504	28.5217	98000	3.5428	0.3746
3.1762	28.8128	99000	3.5349	0.3752

Framework versions

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