exceptions_exp2_swap_0.3_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: 3.5611
• Accuracy: 0.3691

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
4.8511	0.2915	1000	4.7705	0.2520
4.3431	0.5830	2000	4.2941	0.2980
4.1422	0.8745	3000	4.0992	0.3144
3.9989	1.1659	4000	3.9903	0.3248
3.9358	1.4574	5000	3.9162	0.3315
3.8922	1.7488	6000	3.8572	0.3367
3.7473	2.0402	7000	3.8171	0.3404
3.7676	2.3317	8000	3.7858	0.3440
3.7333	2.6232	9000	3.7568	0.3466
3.7189	2.9147	10000	3.7316	0.3489
3.6476	3.2061	11000	3.7173	0.3510
3.6564	3.4976	12000	3.6999	0.3524
3.6384	3.7891	13000	3.6797	0.3542
3.5541	4.0805	14000	3.6768	0.3554
3.5759	4.3719	15000	3.6643	0.3564
3.5827	4.6634	16000	3.6501	0.3577
3.586	4.9549	17000	3.6373	0.3592
3.5081	5.2463	18000	3.6393	0.3597
3.5193	5.5378	19000	3.6292	0.3605
3.5426	5.8293	20000	3.6170	0.3615
3.4361	6.1207	21000	3.6220	0.3617
3.4739	6.4122	22000	3.6138	0.3623
3.4957	6.7037	23000	3.6028	0.3631
3.4998	6.9952	24000	3.5962	0.3639
3.4305	7.2865	25000	3.6028	0.3639
3.4547	7.5780	26000	3.5945	0.3644
3.4605	7.8695	27000	3.5862	0.3652
3.3761	8.1609	28000	3.5952	0.3651
3.411	8.4524	29000	3.5888	0.3654
3.4256	8.7439	30000	3.5807	0.3662
3.3283	9.0353	31000	3.5845	0.3664
3.3857	9.3268	32000	3.5845	0.3665
3.4046	9.6183	33000	3.5739	0.3670
3.4286	9.9098	34000	3.5651	0.3677
3.3347	10.2011	35000	3.5813	0.3672
3.3687	10.4926	36000	3.5721	0.3676
3.3938	10.7841	37000	3.5653	0.3682
3.2867	11.0755	38000	3.5763	0.3680
3.3383	11.3670	39000	3.5684	0.3683
3.3676	11.6585	40000	3.5611	0.3691
3.3762	11.9500	41000	3.5567	0.3693
3.3206	12.2414	42000	3.5699	0.3688
3.3352	12.5329	43000	3.5601	0.3694
3.356	12.8243	44000	3.5519	0.3701
3.2867	13.1157	45000	3.5673	0.3694
3.3041	13.4072	46000	3.5622	0.3695
3.3339	13.6987	47000	3.5524	0.3701
3.3481	13.9902	48000	3.5460	0.3706
3.2936	14.2816	49000	3.5632	0.3699
3.3081	14.5731	50000	3.5574	0.3704
3.3236	14.8646	51000	3.5477	0.3708
3.2463	15.1559	52000	3.5622	0.3703
3.2818	15.4474	53000	3.5576	0.3708
3.2978	15.7389	54000	3.5496	0.3712
3.2183	16.0303	55000	3.5597	0.3706
3.2567	16.3218	56000	3.5599	0.3710
3.2811	16.6133	57000	3.5484	0.3713
3.3014	16.9048	58000	3.5419	0.3718
3.2304	17.1962	59000	3.5569	0.3713
3.2586	17.4877	60000	3.5530	0.3714
3.2783	17.7792	61000	3.5464	0.3718
3.2012	18.0705	62000	3.5591	0.3715
3.2394	18.3620	63000	3.5537	0.3716
3.2703	18.6535	64000	3.5487	0.3718
3.2751	18.9450	65000	3.5384	0.3725
3.2179	19.2364	66000	3.5557	0.3716
3.2521	19.5279	67000	3.5484	0.3721
3.2682	19.8194	68000	3.5403	0.3727
3.1749	20.1108	69000	3.5587	0.3720
3.2256	20.4023	70000	3.5538	0.3724
3.2406	20.6938	71000	3.5451	0.3729
3.238	20.9853	72000	3.5370	0.3730
3.2014	21.2766	73000	3.5555	0.3718
3.2391	21.5681	74000	3.5472	0.3726
3.2435	21.8596	75000	3.5368	0.3733
3.1736	22.1510	76000	3.5562	0.3725
3.2058	22.4425	77000	3.5497	0.3727
3.2295	22.7340	78000	3.5416	0.3731
3.1316	23.0254	79000	3.5597	0.3723
3.1793	23.3169	80000	3.5578	0.3727
3.2119	23.6083	81000	3.5449	0.3731
3.23	23.8998	82000	3.5393	0.3735
3.1532	24.1912	83000	3.5570	0.3727
3.184	24.4827	84000	3.5510	0.3731
3.2121	24.7742	85000	3.5435	0.3736
3.1346	25.0656	86000	3.5601	0.3727
3.1675	25.3571	87000	3.5528	0.3732
3.1908	25.6486	88000	3.5434	0.3734
3.2041	25.9401	89000	3.5376	0.3742
3.1392	26.2314	90000	3.5542	0.3731
3.1831	26.5229	91000	3.5474	0.3738
3.2003	26.8144	92000	3.5401	0.3741
3.1341	27.1058	93000	3.5571	0.3732
3.1517	27.3973	94000	3.5534	0.3733
3.191	27.6888	95000	3.5433	0.3737

Framework versions

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