exceptions_exp2_swap_0.7_resemble_to_drop_3591

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

• Loss: 3.5614
• Accuracy: 0.3690

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
4.8543	0.2915	1000	4.7721	0.2522
4.3634	0.5831	2000	4.2920	0.2983
4.1613	0.8746	3000	4.0996	0.3149
4.0025	1.1662	4000	3.9970	0.3243
3.9429	1.4577	5000	3.9205	0.3307
3.8921	1.7493	6000	3.8598	0.3363
3.762	2.0408	7000	3.8203	0.3403
3.7629	2.3324	8000	3.7892	0.3433
3.7452	2.6239	9000	3.7591	0.3462
3.7296	2.9155	10000	3.7335	0.3482
3.6526	3.2070	11000	3.7198	0.3502
3.6538	3.4985	12000	3.7026	0.3523
3.6472	3.7901	13000	3.6853	0.3538
3.5496	4.0816	14000	3.6788	0.3550
3.5765	4.3732	15000	3.6633	0.3563
3.5881	4.6647	16000	3.6541	0.3572
3.5848	4.9563	17000	3.6407	0.3586
3.5001	5.2478	18000	3.6411	0.3590
3.5336	5.5394	19000	3.6345	0.3597
3.5431	5.8309	20000	3.6203	0.3609
3.4588	6.1224	21000	3.6251	0.3613
3.5006	6.4140	22000	3.6155	0.3618
3.4929	6.7055	23000	3.6077	0.3626
3.5086	6.9971	24000	3.5974	0.3635
3.4293	7.2886	25000	3.6052	0.3634
3.451	7.5802	26000	3.5955	0.3642
3.4634	7.8717	27000	3.5855	0.3651
3.3945	8.1633	28000	3.5967	0.3648
3.4061	8.4548	29000	3.5890	0.3655
3.4226	8.7464	30000	3.5833	0.3661
3.3451	9.0379	31000	3.5867	0.3662
3.3842	9.3294	32000	3.5849	0.3663
3.4032	9.6210	33000	3.5739	0.3669
3.4222	9.9125	34000	3.5660	0.3676
3.3388	10.2041	35000	3.5799	0.3670
3.3732	10.4956	36000	3.5733	0.3675
3.4014	10.7872	37000	3.5658	0.3681
3.302	11.0787	38000	3.5770	0.3678
3.3402	11.3703	39000	3.5702	0.3681
3.3711	11.6618	40000	3.5614	0.3690
3.3803	11.9534	41000	3.5547	0.3693
3.3175	12.2449	42000	3.5703	0.3685
3.3487	12.5364	43000	3.5635	0.3694
3.3575	12.8280	44000	3.5533	0.3698
3.2784	13.1195	45000	3.5693	0.3693
3.3171	13.4111	46000	3.5593	0.3696
3.3452	13.7026	47000	3.5552	0.3700
3.3381	13.9942	48000	3.5493	0.3706
3.2904	14.2857	49000	3.5649	0.3700
3.2955	14.5773	50000	3.5553	0.3705
3.3173	14.8688	51000	3.5490	0.3710
3.2627	15.1603	52000	3.5658	0.3701
3.2856	15.4519	53000	3.5579	0.3708
3.304	15.7434	54000	3.5501	0.3711
3.2243	16.0350	55000	3.5578	0.3710
3.267	16.3265	56000	3.5574	0.3710
3.2842	16.6181	57000	3.5534	0.3712
3.2958	16.9096	58000	3.5446	0.3719
3.2331	17.2012	59000	3.5577	0.3708
3.2606	17.4927	60000	3.5510	0.3715
3.2838	17.7843	61000	3.5495	0.3717
3.1948	18.0758	62000	3.5572	0.3717
3.2566	18.3673	63000	3.5521	0.3718
3.2752	18.6589	64000	3.5466	0.3720
3.2893	18.9504	65000	3.5391	0.3724
3.217	19.2420	66000	3.5557	0.3717
3.2527	19.5335	67000	3.5498	0.3724
3.2674	19.8251	68000	3.5426	0.3727
3.1969	20.1166	69000	3.5588	0.3717
3.2446	20.4082	70000	3.5495	0.3723
3.2453	20.6997	71000	3.5434	0.3726
3.2677	20.9913	72000	3.5388	0.3731
3.197	21.2828	73000	3.5552	0.3721
3.2224	21.5743	74000	3.5448	0.3725
3.257	21.8659	75000	3.5387	0.3731
3.1736	22.1574	76000	3.5546	0.3726
3.2194	22.4490	77000	3.5510	0.3727
3.2309	22.7405	78000	3.5399	0.3732
3.1433	23.0321	79000	3.5552	0.3727
3.1927	23.3236	80000	3.5550	0.3725
3.2065	23.6152	81000	3.5445	0.3734
3.2278	23.9067	82000	3.5373	0.3733
3.1567	24.1983	83000	3.5547	0.3728
3.1946	24.4898	84000	3.5493	0.3731
3.2183	24.7813	85000	3.5390	0.3734
3.1244	25.0729	86000	3.5553	0.3729
3.165	25.3644	87000	3.5542	0.3730
3.1981	25.6560	88000	3.5444	0.3738
3.2102	25.9475	89000	3.5349	0.3741
3.1404	26.2391	90000	3.5543	0.3732
3.1745	26.5306	91000	3.5489	0.3733
3.1931	26.8222	92000	3.5421	0.3738
3.1325	27.1137	93000	3.5598	0.3729
3.1595	27.4052	94000	3.5538	0.3732
3.184	27.6968	95000	3.5446	0.3739
3.1928	27.9883	96000	3.5409	0.3742
3.1449	28.2799	97000	3.5531	0.3737
3.1647	28.5714	98000	3.5494	0.3737
3.1748	28.8630	99000	3.5416	0.3741
3.1183	29.1545	100000	3.5598	0.3734
3.1529	29.4461	101000	3.5520	0.3738
3.1677	29.7376	102000	3.5495	0.3739
3.0812	30.0292	103000	3.5559	0.3736
3.1365	30.3207	104000	3.5546	0.3738
3.147	30.6122	105000	3.5481	0.3741
3.1785	30.9038	106000	3.5407	0.3745
3.0944	31.1953	107000	3.5589	0.3739
3.1344	31.4869	108000	3.5493	0.3740
3.1506	31.7784	109000	3.5466	0.3743

Framework versions

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