exceptions_exp2_swap_0.3_last_to_push_1032

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

• Loss: 3.5809
• 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: 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
4.8293	0.2915	1000	4.7615	0.2532
4.3546	0.5830	2000	4.2873	0.2983
4.1516	0.8745	3000	4.1053	0.3143
4.0058	1.1659	4000	3.9969	0.3240
3.9472	1.4574	5000	3.9207	0.3306
3.8878	1.7488	6000	3.8652	0.3362
3.7514	2.0402	7000	3.8208	0.3402
3.7492	2.3317	8000	3.7912	0.3431
3.7434	2.6232	9000	3.7600	0.3460
3.7298	2.9147	10000	3.7343	0.3488
3.644	3.2061	11000	3.7237	0.3504
3.6455	3.4976	12000	3.7030	0.3520
3.6574	3.7891	13000	3.6841	0.3537
3.5563	4.0805	14000	3.6788	0.3550
3.5661	4.3719	15000	3.6661	0.3561
3.5852	4.6634	16000	3.6523	0.3573
3.5781	4.9549	17000	3.6399	0.3584
3.5114	5.2463	18000	3.6426	0.3591
3.5106	5.5378	19000	3.6317	0.3601
3.5437	5.8293	20000	3.6211	0.3607
3.4466	6.1207	21000	3.6209	0.3614
3.4846	6.4122	22000	3.6177	0.3617
3.4849	6.7037	23000	3.6065	0.3631
3.5075	6.9952	24000	3.5966	0.3636
3.4267	7.2865	25000	3.6058	0.3633
3.4617	7.5780	26000	3.5969	0.3643
3.4669	7.8695	27000	3.5861	0.3650
3.3719	8.1609	28000	3.5982	0.3647
3.4125	8.4524	29000	3.5904	0.3654
3.4428	8.7439	30000	3.5809	0.3658
3.3291	9.0353	31000	3.5841	0.3663
3.3864	9.3268	32000	3.5843	0.3662
3.39	9.6183	33000	3.5759	0.3668
3.41	9.9098	34000	3.5708	0.3672
3.3461	10.2011	35000	3.5790	0.3672
3.363	10.4926	36000	3.5746	0.3675
3.3838	10.7841	37000	3.5633	0.3684
3.3017	11.0755	38000	3.5772	0.3681
3.3445	11.3670	39000	3.5704	0.3681
3.3667	11.6585	40000	3.5636	0.3686
3.3683	11.9500	41000	3.5589	0.3692
3.3232	12.2414	42000	3.5704	0.3686
3.34	12.5329	43000	3.5646	0.3692
3.3565	12.8243	44000	3.5567	0.3699
3.2794	13.1157	45000	3.5726	0.3689
3.308	13.4072	46000	3.5618	0.3698
3.3287	13.6987	47000	3.5553	0.3700
3.3416	13.9902	48000	3.5484	0.3704
3.2829	14.2816	49000	3.5639	0.3697
3.3217	14.5731	50000	3.5581	0.3703
3.3334	14.8646	51000	3.5483	0.3705
3.2562	15.1559	52000	3.5639	0.3702
3.2908	15.4474	53000	3.5592	0.3703
3.3019	15.7389	54000	3.5487	0.3708
3.2195	16.0303	55000	3.5618	0.3707
3.263	16.3218	56000	3.5572	0.3709
3.2831	16.6133	57000	3.5520	0.3712
3.3043	16.9048	58000	3.5398	0.3717
3.2264	17.1962	59000	3.5625	0.3709
3.2583	17.4877	60000	3.5528	0.3712
3.2868	17.7792	61000	3.5480	0.3721
3.1996	18.0705	62000	3.5621	0.3712
3.2382	18.3620	63000	3.5559	0.3716
3.2599	18.6535	64000	3.5491	0.3719
3.2815	18.9450	65000	3.5426	0.3723
3.2292	19.2364	66000	3.5588	0.3713
3.2529	19.5279	67000	3.5507	0.3718
3.2587	19.8194	68000	3.5432	0.3723
3.1769	20.1108	69000	3.5631	0.3715
3.225	20.4023	70000	3.5586	0.3718
3.2412	20.6938	71000	3.5466	0.3722
3.2548	20.9853	72000	3.5390	0.3729
3.2058	21.2766	73000	3.5561	0.3721
3.2211	21.5681	74000	3.5483	0.3723
3.2435	21.8596	75000	3.5437	0.3730
3.1653	22.1510	76000	3.5594	0.3720
3.2079	22.4425	77000	3.5510	0.3725
3.224	22.7340	78000	3.5448	0.3729
3.1328	23.0254	79000	3.5590	0.3727
3.197	23.3169	80000	3.5555	0.3726
3.2171	23.6083	81000	3.5470	0.3728
3.2339	23.8998	82000	3.5413	0.3734
3.1582	24.1912	83000	3.5569	0.3728
3.1902	24.4827	84000	3.5522	0.3729
3.216	24.7742	85000	3.5419	0.3733
3.1411	25.0656	86000	3.5600	0.3728
3.1713	25.3571	87000	3.5573	0.3728
3.1923	25.6486	88000	3.5504	0.3731
3.2059	25.9401	89000	3.5380	0.3738
3.1544	26.2314	90000	3.5581	0.3730
3.1733	26.5229	91000	3.5510	0.3732
3.208	26.8144	92000	3.5445	0.3735
3.1192	27.1058	93000	3.5587	0.3729
3.1594	27.3973	94000	3.5558	0.3731
3.176	27.6888	95000	3.5475	0.3738
3.1894	27.9803	96000	3.5395	0.3740
3.1292	28.2717	97000	3.5567	0.3731
3.1746	28.5632	98000	3.5490	0.3737
3.1804	28.8547	99000	3.5418	0.3743
3.117	29.1460	100000	3.5599	0.3734
3.1417	29.4375	101000	3.5559	0.3735
3.1724	29.7290	102000	3.5469	0.3742
3.0918	30.0204	103000	3.5631	0.3732
3.1294	30.3119	104000	3.5609	0.3734
3.1522	30.6034	105000	3.5500	0.3740
3.1695	30.8949	106000	3.5427	0.3743
3.1205	31.1863	107000	3.5629	0.3733
3.1337	31.4778	108000	3.5532	0.3742
3.1511	31.7693	109000	3.5459	0.3743

Framework versions

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