exceptions_exp2_swap_0.7_resemble_to_hit_40817

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

• Loss: 3.5653
• 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: 40817
• 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
4.824	0.2915	1000	4.7446	0.2558
4.3378	0.5831	2000	4.2893	0.2993
4.1546	0.8746	3000	4.1047	0.3148
4.0065	1.1662	4000	3.9984	0.3237
3.9321	1.4577	5000	3.9243	0.3305
3.893	1.7493	6000	3.8645	0.3357
3.7516	2.0408	7000	3.8217	0.3403
3.7634	2.3324	8000	3.7917	0.3431
3.7682	2.6239	9000	3.7633	0.3458
3.732	2.9155	10000	3.7367	0.3486
3.6451	3.2070	11000	3.7229	0.3501
3.6477	3.4985	12000	3.7037	0.3519
3.6457	3.7901	13000	3.6858	0.3534
3.5456	4.0816	14000	3.6785	0.3548
3.5694	4.3732	15000	3.6694	0.3559
3.5783	4.6647	16000	3.6539	0.3573
3.5885	4.9563	17000	3.6401	0.3585
3.4968	5.2478	18000	3.6439	0.3589
3.5249	5.5394	19000	3.6315	0.3599
3.5401	5.8309	20000	3.6211	0.3609
3.4525	6.1224	21000	3.6252	0.3616
3.4799	6.4140	22000	3.6166	0.3620
3.4922	6.7055	23000	3.6070	0.3630
3.5002	6.9971	24000	3.5985	0.3636
3.4422	7.2886	25000	3.6075	0.3636
3.4561	7.5802	26000	3.5985	0.3639
3.4665	7.8717	27000	3.5896	0.3649
3.3983	8.1633	28000	3.5967	0.3646
3.4324	8.4548	29000	3.5905	0.3655
3.4351	8.7464	30000	3.5839	0.3660
3.3383	9.0379	31000	3.5889	0.3661
3.3927	9.3294	32000	3.5860	0.3662
3.409	9.6210	33000	3.5797	0.3665
3.4222	9.9125	34000	3.5684	0.3676
3.3559	10.2041	35000	3.5802	0.3670
3.3831	10.4956	36000	3.5752	0.3675
3.392	10.7872	37000	3.5672	0.3681
3.2873	11.0787	38000	3.5780	0.3678
3.3387	11.3703	39000	3.5731	0.3680
3.3725	11.6618	40000	3.5653	0.3687
3.3728	11.9534	41000	3.5576	0.3691
3.3118	12.2449	42000	3.5712	0.3686
3.3526	12.5364	43000	3.5627	0.3690
3.3512	12.8280	44000	3.5543	0.3694
3.2739	13.1195	45000	3.5726	0.3690
3.3029	13.4111	46000	3.5649	0.3693
3.3372	13.7026	47000	3.5564	0.3701
3.3385	13.9942	48000	3.5496	0.3704
3.2775	14.2857	49000	3.5672	0.3697
3.3144	14.5773	50000	3.5577	0.3702
3.331	14.8688	51000	3.5517	0.3703
3.2568	15.1603	52000	3.5643	0.3701
3.294	15.4519	53000	3.5583	0.3701
3.3072	15.7434	54000	3.5494	0.3712
3.2203	16.0350	55000	3.5607	0.3708
3.2573	16.3265	56000	3.5614	0.3704
3.2864	16.6181	57000	3.5510	0.3713
3.309	16.9096	58000	3.5448	0.3713
3.2354	17.2012	59000	3.5605	0.3709
3.2663	17.4927	60000	3.5531	0.3714
3.2829	17.7843	61000	3.5482	0.3716
3.2079	18.0758	62000	3.5614	0.3711
3.2428	18.3673	63000	3.5567	0.3715
3.2613	18.6589	64000	3.5498	0.3719
3.2729	18.9504	65000	3.5404	0.3725
3.216	19.2420	66000	3.5602	0.3711
3.2457	19.5335	67000	3.5521	0.3718
3.2614	19.8251	68000	3.5460	0.3724
3.1775	20.1166	69000	3.5607	0.3717
3.2192	20.4082	70000	3.5559	0.3720
3.246	20.6997	71000	3.5439	0.3724
3.2579	20.9913	72000	3.5360	0.3731
3.2005	21.2828	73000	3.5584	0.3720
3.2281	21.5743	74000	3.5503	0.3723
3.246	21.8659	75000	3.5410	0.3729
3.177	22.1574	76000	3.5602	0.3721
3.2039	22.4490	77000	3.5524	0.3725
3.2411	22.7405	78000	3.5465	0.3729
3.134	23.0321	79000	3.5555	0.3726
3.1912	23.3236	80000	3.5552	0.3723
3.2044	23.6152	81000	3.5454	0.3734
3.2231	23.9067	82000	3.5421	0.3731
3.1498	24.1983	83000	3.5579	0.3724
3.1972	24.4898	84000	3.5514	0.3729
3.2171	24.7813	85000	3.5418	0.3734
3.1401	25.0729	86000	3.5596	0.3723
3.179	25.3644	87000	3.5554	0.3730
3.1911	25.6560	88000	3.5463	0.3733
3.2106	25.9475	89000	3.5380	0.3739
3.1548	26.2391	90000	3.5564	0.3728
3.1763	26.5306	91000	3.5507	0.3730
3.184	26.8222	92000	3.5390	0.3740

Framework versions

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