exceptions_exp2_swap_0.7_resemble_to_hit_5039

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

• Loss: 3.5648
• Accuracy: 0.3685

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: 5039
• 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	4.7613	0.2532
4.3498	0.5831	2000	4.2886	0.2983
4.1547	0.8746	3000	4.0988	0.3149
4.0145	1.1662	4000	3.9934	0.3244
3.9383	1.4577	5000	3.9197	0.3307
3.8883	1.7493	6000	3.8589	0.3361
3.7559	2.0408	7000	3.8185	0.3405
3.7641	2.3324	8000	3.7866	0.3431
3.7488	2.6239	9000	3.7595	0.3461
3.7263	2.9155	10000	3.7318	0.3485
3.6443	3.2070	11000	3.7193	0.3506
3.6504	3.4985	12000	3.7009	0.3523
3.6455	3.7901	13000	3.6833	0.3538
3.5473	4.0816	14000	3.6778	0.3551
3.5863	4.3732	15000	3.6653	0.3564
3.5731	4.6647	16000	3.6522	0.3575
3.582	4.9563	17000	3.6385	0.3588
3.5069	5.2478	18000	3.6418	0.3588
3.543	5.5394	19000	3.6308	0.3601
3.5361	5.8309	20000	3.6189	0.3610
3.451	6.1224	21000	3.6233	0.3613
3.4814	6.4140	22000	3.6158	0.3619
3.4878	6.7055	23000	3.6079	0.3629
3.493	6.9971	24000	3.5965	0.3637
3.436	7.2886	25000	3.6034	0.3635
3.4608	7.5802	26000	3.5958	0.3641
3.4702	7.8717	27000	3.5856	0.3650
3.3809	8.1633	28000	3.5995	0.3645
3.422	8.4548	29000	3.5882	0.3654
3.4176	8.7464	30000	3.5818	0.3658
3.3359	9.0379	31000	3.5894	0.3659
3.3726	9.3294	32000	3.5863	0.3661
3.391	9.6210	33000	3.5786	0.3664
3.4213	9.9125	34000	3.5722	0.3672
3.3467	10.2041	35000	3.5802	0.3669
3.3745	10.4956	36000	3.5726	0.3675
3.3941	10.7872	37000	3.5689	0.3678
3.2898	11.0787	38000	3.5759	0.3677
3.35	11.3703	39000	3.5721	0.3682
3.3659	11.6618	40000	3.5648	0.3685
3.373	11.9534	41000	3.5584	0.3687
3.3035	12.2449	42000	3.5701	0.3687
3.3412	12.5364	43000	3.5653	0.3689
3.3493	12.8280	44000	3.5569	0.3697
3.2728	13.1195	45000	3.5694	0.3690
3.317	13.4111	46000	3.5634	0.3695
3.3272	13.7026	47000	3.5588	0.3697
3.34	13.9942	48000	3.5487	0.3702
3.2908	14.2857	49000	3.5662	0.3696
3.3103	14.5773	50000	3.5582	0.3701
3.3314	14.8688	51000	3.5488	0.3707
3.2472	15.1603	52000	3.5681	0.3697
3.2925	15.4519	53000	3.5561	0.3704
3.3104	15.7434	54000	3.5505	0.3710
3.2089	16.0350	55000	3.5597	0.3707
3.2707	16.3265	56000	3.5619	0.3706
3.2937	16.6181	57000	3.5527	0.3712
3.3001	16.9096	58000	3.5463	0.3716
3.2327	17.2012	59000	3.5615	0.3710
3.2664	17.4927	60000	3.5552	0.3712
3.2699	17.7843	61000	3.5493	0.3715
3.1976	18.0758	62000	3.5649	0.3710
3.2465	18.3673	63000	3.5580	0.3714
3.2695	18.6589	64000	3.5511	0.3716
3.2777	18.9504	65000	3.5424	0.3723
3.2184	19.2420	66000	3.5599	0.3714
3.2462	19.5335	67000	3.5519	0.3718
3.2623	19.8251	68000	3.5440	0.3721
3.1812	20.1166	69000	3.5582	0.3712
3.2282	20.4082	70000	3.5576	0.3712
3.2495	20.6997	71000	3.5479	0.3723
3.2562	20.9913	72000	3.5388	0.3728
3.2073	21.2828	73000	3.5595	0.3716
3.2329	21.5743	74000	3.5505	0.3723
3.2448	21.8659	75000	3.5443	0.3725
3.1769	22.1574	76000	3.5613	0.3720
3.2099	22.4490	77000	3.5541	0.3722
3.2289	22.7405	78000	3.5437	0.3730
3.1423	23.0321	79000	3.5598	0.3722
3.1944	23.3236	80000	3.5565	0.3723
3.1997	23.6152	81000	3.5548	0.3725
3.228	23.9067	82000	3.5423	0.3731
3.1606	24.1983	83000	3.5599	0.3723
3.1806	24.4898	84000	3.5535	0.3729
3.2038	24.7813	85000	3.5452	0.3729
3.1416	25.0729	86000	3.5625	0.3725
3.1684	25.3644	87000	3.5566	0.3727
3.1963	25.6560	88000	3.5484	0.3732
3.2185	25.9475	89000	3.5412	0.3736
3.1573	26.2391	90000	3.5598	0.3727
3.1772	26.5306	91000	3.5530	0.3731
3.1987	26.8222	92000	3.5454	0.3734

Framework versions

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