exceptions_exp2_swap_require_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: 3.5709
• Accuracy: 0.3680

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
4.8412	0.2911	1000	4.7632	0.2533
4.3374	0.5822	2000	4.2789	0.2996
4.1421	0.8733	3000	4.0889	0.3162
3.9896	1.1642	4000	3.9892	0.3259
3.9315	1.4553	5000	3.9120	0.3321
3.8626	1.7464	6000	3.8539	0.3377
3.7296	2.0373	7000	3.8100	0.3421
3.7404	2.3284	8000	3.7776	0.3450
3.7282	2.6195	9000	3.7501	0.3478
3.7136	2.9106	10000	3.7230	0.3500
3.627	3.2014	11000	3.7090	0.3520
3.6246	3.4925	12000	3.6921	0.3539
3.6345	3.7837	13000	3.6750	0.3555
3.5324	4.0745	14000	3.6658	0.3569
3.5519	4.3656	15000	3.6544	0.3577
3.5647	4.6567	16000	3.6403	0.3593
3.5804	4.9478	17000	3.6290	0.3603
3.495	5.2387	18000	3.6321	0.3606
3.5176	5.5298	19000	3.6206	0.3619
3.5285	5.8209	20000	3.6092	0.3627
3.4359	6.1118	21000	3.6126	0.3631
3.4691	6.4029	22000	3.6052	0.3636
3.4806	6.6940	23000	3.5950	0.3647
3.4859	6.9851	24000	3.5877	0.3649
3.4119	7.2760	25000	3.5937	0.3653
3.4391	7.5671	26000	3.5847	0.3661
3.4647	7.8582	27000	3.5756	0.3668
3.3693	8.1490	28000	3.5883	0.3665
3.4013	8.4401	29000	3.5808	0.3670
3.431	8.7313	30000	3.5709	0.3680
3.3154	9.0221	31000	3.5734	0.3679
3.3683	9.3132	32000	3.5767	0.3678
3.3906	9.6043	33000	3.5649	0.3687
3.4058	9.8954	34000	3.5568	0.3693
3.3313	10.1863	35000	3.5708	0.3688
3.356	10.4774	36000	3.5605	0.3691
3.3801	10.7685	37000	3.5535	0.3699
3.2818	11.0594	38000	3.5629	0.3698
3.3199	11.3505	39000	3.5641	0.3694
3.3573	11.6416	40000	3.5546	0.3704
3.3593	11.9327	41000	3.5442	0.3708
3.297	12.2236	42000	3.5588	0.3703
3.3204	12.5147	43000	3.5549	0.3708
3.3522	12.8058	44000	3.5421	0.3716
3.2624	13.0966	45000	3.5580	0.3709
3.295	13.3878	46000	3.5528	0.3713
3.3167	13.6789	47000	3.5468	0.3718
3.3347	13.9700	48000	3.5373	0.3724
3.2663	14.2608	49000	3.5512	0.3717
3.2948	14.5519	50000	3.5444	0.3719
3.3313	14.8430	51000	3.5357	0.3727
3.2304	15.1339	52000	3.5524	0.3720
3.2744	15.4250	53000	3.5468	0.3723
3.288	15.7161	54000	3.5393	0.3726
3.263	16.0070	55000	3.5453	0.3724
3.2338	16.2981	56000	3.5460	0.3729
3.2846	16.5892	57000	3.5415	0.3728
3.2937	16.8803	58000	3.5342	0.3733
3.2176	17.1712	59000	3.5491	0.3726
3.2612	17.4623	60000	3.5430	0.3731
3.2734	17.7534	61000	3.5320	0.3735
3.171	18.0442	62000	3.5451	0.3731
3.2308	18.3354	63000	3.5478	0.3730
3.2547	18.6265	64000	3.5371	0.3735
3.2742	18.9176	65000	3.5280	0.3744
3.1909	19.2084	66000	3.5466	0.3730
3.2298	19.4995	67000	3.5387	0.3736
3.2633	19.7906	68000	3.5328	0.3743
3.1564	20.0815	69000	3.5451	0.3737
3.204	20.3726	70000	3.5404	0.3742
3.2406	20.6637	71000	3.5343	0.3742
3.2356	20.9548	72000	3.5298	0.3746
3.1908	21.2457	73000	3.5452	0.3737
3.2197	21.5368	74000	3.5367	0.3742
3.2351	21.8279	75000	3.5300	0.3746
3.1663	22.1188	76000	3.5442	0.3741
3.1905	22.4099	77000	3.5429	0.3743
3.2269	22.7010	78000	3.5306	0.3746
3.2333	22.9921	79000	3.5248	0.3753
3.1703	23.2830	80000	3.5430	0.3741
3.2025	23.5741	81000	3.5374	0.3745
3.2064	23.8652	82000	3.5285	0.3749
3.1452	24.1560	83000	3.5460	0.3741
3.1751	24.4471	84000	3.5390	0.3749
3.1984	24.7382	85000	3.5300	0.3752
3.1106	25.0291	86000	3.5475	0.3744
3.148	25.3202	87000	3.5479	0.3742
3.1717	25.6113	88000	3.5358	0.3751
3.2088	25.9024	89000	3.5298	0.3754
3.1331	26.1933	90000	3.5462	0.3748
3.162	26.4844	91000	3.5390	0.3750
3.1964	26.7755	92000	3.5340	0.3754
3.1036	27.0664	93000	3.5499	0.3745
3.1444	27.3575	94000	3.5426	0.3750
3.1646	27.6486	95000	3.5367	0.3753
3.1847	27.9397	96000	3.5311	0.3756
3.1124	28.2306	97000	3.5469	0.3750
3.1577	28.5217	98000	3.5390	0.3752
3.1592	28.8128	99000	3.5344	0.3757

Framework versions

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