exceptions_exp2_swap_0.3_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: 3.5498
• Accuracy: 0.3741

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
4.8376	0.2915	1000	0.2542	4.7589
4.3508	0.5830	2000	0.2977	4.2982
4.1599	0.8745	3000	0.3137	4.1119
4.0058	1.1659	4000	0.3232	4.0022
3.9397	1.4574	5000	0.3305	3.9227
3.8862	1.7488	6000	0.3356	3.8669
3.7487	2.0402	7000	0.3400	3.8238
3.7661	2.3317	8000	0.3429	3.7931
3.7449	2.6232	9000	0.3457	3.7612
3.7305	2.9147	10000	0.3482	3.7342
3.636	3.2061	11000	0.3503	3.7234
3.6523	3.4976	12000	0.3518	3.7036
3.6479	3.7891	13000	0.3535	3.6866
3.5494	4.0805	14000	0.3549	3.6778
3.5719	4.3719	15000	0.3561	3.6679
3.5815	4.6634	16000	0.3575	3.6531
3.5827	4.9549	17000	0.3585	3.6418
3.5121	5.2463	18000	0.3592	3.6428
3.5172	5.5378	19000	0.3597	3.6341
3.5469	5.8293	20000	0.3607	3.6216
3.4552	6.1207	21000	0.3614	3.6244
3.4922	6.4122	22000	0.3619	3.6169
3.499	6.7037	23000	0.3626	3.6087
3.4927	6.9952	24000	0.3639	3.5962
3.4408	7.2865	25000	0.3633	3.6067
3.4523	7.5780	26000	0.3640	3.5957
3.4617	7.8695	27000	0.3650	3.5857
3.396	8.1609	28000	0.3648	3.5959
3.4334	8.4524	29000	0.3652	3.5926
3.4364	8.7439	30000	0.3661	3.5817
3.3397	9.0353	31000	0.3660	3.5855
3.3905	9.3268	32000	0.3661	3.5862
3.4009	9.6183	33000	0.3670	3.5789
3.4185	9.9098	34000	0.3677	3.5687
3.348	10.2011	35000	0.3672	3.5799
3.3779	10.4926	36000	0.3676	3.5705
3.3988	10.7841	37000	0.3681	3.5676
3.3056	11.0755	38000	0.3679	3.5776
3.345	11.3670	39000	0.3681	3.5718
3.3619	11.6585	40000	0.3689	3.5643
3.387	11.9500	41000	0.3694	3.5561
3.3217	12.2414	42000	0.3688	3.5691
3.3443	12.5329	43000	0.3691	3.5631
3.3572	12.8243	44000	0.3696	3.5544
3.282	13.1157	45000	0.3690	3.5689
3.3217	13.4072	46000	0.3696	3.5634
3.333	13.6987	47000	0.3703	3.5528
3.3499	13.9902	48000	0.3704	3.5476
3.2773	14.2816	49000	0.3698	3.5656
3.3086	14.5731	50000	0.3703	3.5580
3.343	14.8646	51000	0.3706	3.5467
3.2598	15.1559	52000	0.3702	3.5646
3.2774	15.4474	53000	0.3705	3.5598
3.3126	15.7389	54000	0.3711	3.5496
3.2012	16.0303	55000	0.3705	3.5610
3.2648	16.3218	56000	0.3711	3.5536
3.2782	16.6133	57000	0.3710	3.5503
3.3024	16.9048	58000	0.3720	3.5396
3.2318	17.1962	59000	0.3710	3.5579
3.2653	17.4877	60000	0.3715	3.5493
3.274	17.7792	61000	0.3722	3.5431
3.1885	18.0705	62000	0.3713	3.5596
3.2372	18.3620	63000	0.3717	3.5528
3.2585	18.6535	64000	0.3721	3.5450
3.279	18.9450	65000	0.3726	3.5389
3.2051	19.2364	66000	0.3718	3.5559
3.2422	19.5279	67000	0.3718	3.5507
3.2517	19.8194	68000	0.3726	3.5421
3.1962	20.1108	69000	0.3719	3.5548
3.2241	20.4023	70000	0.3721	3.5511
3.2393	20.6938	71000	0.3724	3.5465
3.2653	20.9853	72000	0.3729	3.5397
3.2169	21.2766	73000	0.3719	3.5543
3.2173	21.5681	74000	0.3726	3.5482
3.2412	21.8596	75000	0.3729	3.5411
3.1695	22.1510	76000	0.3725	3.5561
3.2119	22.4425	77000	0.3726	3.5493
3.2316	22.7340	78000	0.3731	3.5434
3.1426	23.0254	79000	0.3723	3.5559
3.1847	23.3169	80000	0.3729	3.5535
3.1668	23.6083	81000	3.5551	0.3725
3.2001	23.8998	82000	3.5479	0.3731
3.1768	24.1915	83000	3.5602	0.3723
3.1871	24.4830	84000	3.5519	0.3727
3.2136	24.7745	85000	3.5428	0.3733
3.1267	25.0659	86000	3.5576	0.3731
3.1684	25.3574	87000	3.5551	0.3727
3.2026	25.6489	88000	3.5472	0.3735
3.2202	25.9404	89000	3.5379	0.3737
3.1603	26.2317	90000	3.5586	0.3729
3.1896	26.5232	91000	3.5494	0.3735
3.1884	26.8147	92000	3.5417	0.3740
3.1285	27.1061	93000	3.5546	0.3731
3.1577	27.3976	94000	3.5499	0.3736
3.183	27.6891	95000	3.5481	0.3736
3.1974	27.9806	96000	3.5389	0.3741
3.1403	28.2720	97000	3.5531	0.3733
3.1545	28.5635	98000	3.5498	0.3738
3.165	28.8550	99000	3.5403	0.3740
3.119	29.1463	100000	3.5620	0.3732
3.1416	29.4378	101000	3.5507	0.3740
3.1652	29.7293	102000	3.5452	0.3741
3.0738	30.0207	103000	3.5542	0.3740
3.1227	30.3122	104000	3.5568	0.3737
3.1518	30.6037	105000	3.5512	0.3741
3.1593	30.8952	106000	3.5423	0.3742
3.1051	31.1866	107000	3.5578	0.3737
3.1409	31.4781	108000	3.5499	0.3742
3.1566	31.7695	109000	3.5498	0.3741

Framework versions

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