exceptions_exp2_swap_last_to_hit_2128

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

• Loss: 3.5613
• Accuracy: 0.3687

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
4.8567	0.2915	1000	4.7799	0.2511
4.3531	0.5830	2000	4.2939	0.2982
4.1547	0.8744	3000	4.1078	0.3145
4.0021	1.1659	4000	3.9968	0.3240
3.9366	1.4573	5000	3.9226	0.3305
3.8894	1.7488	6000	3.8645	0.3359
3.7455	2.0402	7000	3.8227	0.3402
3.758	2.3317	8000	3.7905	0.3435
3.7481	2.6232	9000	3.7596	0.3460
3.7292	2.9147	10000	3.7358	0.3485
3.6277	3.2061	11000	3.7233	0.3502
3.6545	3.4976	12000	3.7065	0.3519
3.6398	3.7890	13000	3.6855	0.3537
3.5561	4.0804	14000	3.6790	0.3549
3.5631	4.3719	15000	3.6644	0.3563
3.5866	4.6634	16000	3.6527	0.3573
3.5832	4.9549	17000	3.6388	0.3585
3.5163	5.2463	18000	3.6419	0.3590
3.5214	5.5378	19000	3.6339	0.3597
3.5338	5.8293	20000	3.6209	0.3611
3.4521	6.1207	21000	3.6242	0.3613
3.4713	6.4121	22000	3.6174	0.3620
3.4912	6.7036	23000	3.6068	0.3628
3.5111	6.9951	24000	3.5960	0.3636
3.4442	7.2865	25000	3.6050	0.3633
3.4648	7.5780	26000	3.5961	0.3641
3.4615	7.8695	27000	3.5861	0.3651
3.4125	8.1609	28000	3.5984	0.3646
3.4207	8.4524	29000	3.5890	0.3652
3.432	8.7438	30000	3.5793	0.3659
3.3405	9.0353	31000	3.5878	0.3661
3.3913	9.3267	32000	3.5837	0.3663
3.3967	9.6182	33000	3.5747	0.3666
3.4174	9.9097	34000	3.5689	0.3675
3.3397	10.2011	35000	3.5783	0.3672
3.387	10.4926	36000	3.5739	0.3674
3.3912	10.7841	37000	3.5661	0.3682
3.2959	11.0755	38000	3.5754	0.3676
3.3426	11.3670	39000	3.5701	0.3683
3.3551	11.6584	40000	3.5613	0.3687
3.3789	11.9499	41000	3.5549	0.3691
3.2939	12.2413	42000	3.5692	0.3688
3.3411	12.5328	43000	3.5633	0.3688
3.3411	12.8243	44000	3.5567	0.3696
3.274	13.1157	45000	3.5644	0.3695
3.3106	13.4072	46000	3.5616	0.3695
3.3322	13.6987	47000	3.5552	0.3704
3.3397	13.9901	48000	3.5493	0.3703
3.2824	14.2816	49000	3.5610	0.3700
3.3137	14.5730	50000	3.5556	0.3703
3.328	14.8645	51000	3.5480	0.3708
3.2563	15.1559	52000	3.5600	0.3701
3.2871	15.4474	53000	3.5552	0.3706
3.3122	15.7389	54000	3.5496	0.3709
3.2048	16.0303	55000	3.5630	0.3705
3.2606	16.3218	56000	3.5563	0.3708
3.2827	16.6133	57000	3.5518	0.3712
3.2948	16.9047	58000	3.5413	0.3720
3.224	17.1962	59000	3.5598	0.3713
3.2661	17.4876	60000	3.5545	0.3711
3.2773	17.7791	61000	3.5428	0.3721
3.2067	18.0705	62000	3.5578	0.3712
3.2479	18.3620	63000	3.5556	0.3715
3.2747	18.6535	64000	3.5460	0.3720
3.2757	18.9450	65000	3.5408	0.3721
3.2079	19.2364	66000	3.5580	0.3716
3.2387	19.5279	67000	3.5497	0.3722
3.2595	19.8193	68000	3.5431	0.3726
3.1785	20.1108	69000	3.5574	0.3720
3.2082	20.4022	70000	3.5545	0.3720
3.2437	20.6937	71000	3.5460	0.3723
3.2747	20.9852	72000	3.5383	0.3729
3.202	21.2766	73000	3.5585	0.3720
3.2154	21.5681	74000	3.5483	0.3728
3.2222	21.8596	75000	3.5394	0.3728
3.1642	22.1510	76000	3.5588	0.3721
3.2169	22.4425	77000	3.5527	0.3727
3.2272	22.7339	78000	3.5475	0.3729
3.1308	23.0254	79000	3.5576	0.3723
3.1808	23.3168	80000	3.5537	0.3725
3.2101	23.6083	81000	3.5454	0.3730
3.2339	23.8998	82000	3.5366	0.3735
3.1704	24.1912	83000	3.5602	0.3727
3.1733	24.4827	84000	3.5499	0.3729
3.2163	24.7742	85000	3.5395	0.3736
3.1212	25.0656	86000	3.5588	0.3729
3.1743	25.3571	87000	3.5500	0.3730
3.1938	25.6485	88000	3.5481	0.3732
3.2285	25.9400	89000	3.5398	0.3737
3.1462	26.2314	90000	3.5607	0.3728
3.1847	26.5229	91000	3.5460	0.3736
3.1766	26.8144	92000	3.5409	0.3738
3.122	27.1058	93000	3.5572	0.3731
3.164	27.3973	94000	3.5495	0.3734
3.1917	27.6888	95000	3.5458	0.3738
3.1972	27.9802	96000	3.5376	0.3744
3.1274	28.2717	97000	3.5601	0.3733
3.1667	28.5631	98000	3.5506	0.3737
3.1822	28.8546	99000	3.5397	0.3741
3.1053	29.1460	100000	3.5531	0.3735
3.1373	29.4375	101000	3.5518	0.3737
3.1535	29.7290	102000	3.5447	0.3740

Framework versions

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