exceptions_exp2_swap_0.7_last_to_carry_5039

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

• Loss: 3.5623
• Accuracy: 0.3688

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	Accuracy	Validation Loss
4.8681	0.2915	1000	0.2501	4.7861
4.3482	0.5830	2000	0.2981	4.2942
4.1621	0.8745	3000	0.3142	4.1046
4.0072	1.1659	4000	0.3239	3.9992
3.9422	1.4574	5000	0.3304	3.9264
3.8959	1.7489	6000	0.3361	3.8627
3.7619	2.0402	7000	0.3398	3.8251
3.7584	2.3317	8000	0.3431	3.7926
3.7538	2.6233	9000	0.3458	3.7609
3.7423	2.9148	10000	0.3484	3.7362
3.6394	3.2061	11000	0.3504	3.7219
3.6493	3.4976	12000	0.3519	3.7053
3.6509	3.7891	13000	0.3540	3.6859
3.5456	4.0805	14000	0.3550	3.6799
3.5779	4.3720	15000	0.3562	3.6674
3.5822	4.6635	16000	0.3572	3.6546
3.5777	4.9550	17000	0.3585	3.6399
3.5233	5.2463	18000	0.3591	3.6423
3.5346	5.5378	19000	0.3600	3.6303
3.5235	5.8293	20000	0.3609	3.6207
3.4512	6.1207	21000	0.3615	3.6258
3.4754	6.4122	22000	0.3619	3.6182
3.4897	6.7037	23000	0.3630	3.6067
3.495	6.9952	24000	0.3635	3.5976
3.4407	7.2866	25000	0.3635	3.6048
3.4693	7.5781	26000	0.3642	3.5994
3.4703	7.8696	27000	0.3652	3.5871
3.3946	8.1609	28000	0.3647	3.5980
3.4226	8.4524	29000	0.3651	3.5908
3.4239	8.7439	30000	0.3655	3.5854
3.3303	9.0353	31000	0.3661	3.5858
3.388	9.3268	32000	0.3660	3.5876
3.4098	9.6183	33000	0.3669	3.5774
3.4248	9.9098	34000	0.3674	3.5695
3.3431	10.2011	35000	0.3667	3.5835
3.3764	10.4927	36000	0.3674	3.5778
3.3879	10.7842	37000	0.3680	3.5669
3.3064	11.0755	38000	0.3679	3.5771
3.352	11.3670	39000	0.3680	3.5739
3.3762	11.6585	40000	0.3688	3.5623
3.3814	11.9500	41000	0.3693	3.5562
3.3	12.2414	42000	0.3687	3.5722
3.3333	12.5329	43000	0.3689	3.5665
3.3738	12.8244	44000	0.3698	3.5537
3.2802	13.1157	45000	0.3691	3.5711
3.3085	13.4072	46000	0.3694	3.5677
3.3408	13.6988	47000	0.3699	3.5576
3.3487	13.9903	48000	0.3706	3.5486
3.284	14.2816	49000	0.3699	3.5688
3.3191	14.5731	50000	0.3700	3.5619
3.3306	14.8646	51000	0.3705	3.5498
3.2514	15.1560	52000	0.3701	3.5695
3.2841	15.4475	53000	0.3703	3.5619
3.3229	15.7390	54000	0.3711	3.5466
3.1904	16.0303	55000	0.3706	3.5626
3.2481	16.3218	56000	0.3705	3.5618
3.2817	16.6133	57000	0.3711	3.5510
3.3091	16.9049	58000	0.3718	3.5421
3.2368	17.1962	59000	0.3708	3.5608
3.2705	17.4877	60000	0.3712	3.5555
3.285	17.7792	61000	0.3719	3.5488
3.2	18.0705	62000	0.3714	3.5644
3.2485	18.3621	63000	0.3713	3.5579
3.2607	18.6536	64000	0.3719	3.5489
3.2829	18.9451	65000	0.3725	3.5385
3.2138	19.2364	66000	0.3716	3.5596
3.2397	19.5279	67000	0.3720	3.5553
3.2527	19.8194	68000	0.3726	3.5445
3.1811	20.1108	69000	0.3718	3.5564
3.2247	20.4023	70000	0.3719	3.5553
3.2418	20.6938	71000	0.3723	3.5476
3.2511	20.9853	72000	0.3731	3.5393
3.2031	21.2766	73000	0.3722	3.5538
3.2361	21.5682	74000	0.3725	3.5505
3.25	21.8597	75000	0.3731	3.5425
3.1804	22.1510	76000	0.3722	3.5577
3.2073	22.4425	77000	0.3724	3.5554
3.2113	22.7340	78000	0.3728	3.5457
3.146	23.0254	79000	0.3724	3.5557
3.1904	23.3169	80000	0.3725	3.5550
3.1809	23.6084	81000	3.5572	0.3725
3.2091	23.8999	82000	3.5520	0.3727
3.1676	24.1915	83000	3.5599	0.3722
3.2063	24.4830	84000	3.5518	0.3730
3.2168	24.7745	85000	3.5443	0.3732
3.1386	25.0659	86000	3.5609	0.3724
3.1867	25.3574	87000	3.5552	0.3728
3.2042	25.6489	88000	3.5478	0.3734
3.205	25.9404	89000	3.5421	0.3735
3.1544	26.2318	90000	3.5567	0.3731
3.188	26.5233	91000	3.5475	0.3732
3.2043	26.8148	92000	3.5421	0.3738
3.1271	27.1061	93000	3.5587	0.3732
3.1542	27.3976	94000	3.5551	0.3732
3.187	27.6891	95000	3.5467	0.3738
3.197	27.9806	96000	3.5386	0.3741
3.1326	28.2720	97000	3.5587	0.3732
3.1632	28.5635	98000	3.5471	0.3738
3.1859	28.8550	99000	3.5395	0.3742
3.1239	29.1463	100000	3.5586	0.3733

Framework versions

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