exceptions_exp2_swap_last_to_carry_1032

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: 0.3686

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: 1032
• 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
4.8263	0.2915	1000	4.7449	0.2558
4.3455	0.5830	2000	4.2870	0.2985
4.1514	0.8744	3000	4.1019	0.3148
3.987	1.1659	4000	3.9953	0.3243
3.9374	1.4573	5000	3.9222	0.3308
3.8822	1.7488	6000	3.8616	0.3360
3.7643	2.0402	7000	3.8199	0.3402
3.7537	2.3317	8000	3.7917	0.3434
3.7266	2.6232	9000	3.7610	0.3460
3.742	2.9147	10000	3.7356	0.3481
3.6393	3.2061	11000	3.7225	0.3504
3.6453	3.4976	12000	3.7026	0.3518
3.657	3.7890	13000	3.6851	0.3538
3.546	4.0804	14000	3.6804	0.3545
3.5833	4.3719	15000	3.6665	0.3559
3.5835	4.6634	16000	3.6532	0.3571
3.5758	4.9549	17000	3.6407	0.3586
3.5088	5.2463	18000	3.6444	0.3586
3.5242	5.5378	19000	3.6330	0.3598
3.5442	5.8293	20000	3.6211	0.3608
3.4435	6.1207	21000	3.6259	0.3611
3.4808	6.4121	22000	3.6183	0.3617
3.4893	6.7036	23000	3.6068	0.3626
3.4979	6.9951	24000	3.5983	0.3637
3.4392	7.2865	25000	3.6089	0.3633
3.4573	7.5780	26000	3.5963	0.3643
3.4584	7.8695	27000	3.5875	0.3648
3.3926	8.1609	28000	3.5971	0.3649
3.4104	8.4524	29000	3.5896	0.3651
3.4405	8.7438	30000	3.5801	0.3657
3.3305	9.0353	31000	3.5873	0.3656
3.3833	9.3267	32000	3.5855	0.3661
3.3943	9.6182	33000	3.5761	0.3667
3.411	9.9097	34000	3.5687	0.3673
3.3432	10.2011	35000	3.5795	0.3669
3.3606	10.4926	36000	3.5743	0.3676
3.3947	10.7841	37000	3.5686	0.3680
3.2973	11.0755	38000	3.5775	0.3679
3.3485	11.3670	39000	3.5713	0.3680
3.363	11.6584	40000	3.5653	0.3686
3.3699	11.9499	41000	3.5540	0.3693
3.3055	12.2413	42000	3.5720	0.3683
3.3497	12.5328	43000	3.5631	0.3692
3.3522	12.8243	44000	3.5554	0.3695
3.2653	13.1157	45000	3.5691	0.3686
3.3083	13.4072	46000	3.5648	0.3692
3.3394	13.6987	47000	3.5561	0.3695
3.3386	13.9901	48000	3.5465	0.3703
3.2826	14.2816	49000	3.5628	0.3695
3.3163	14.5730	50000	3.5546	0.3699
3.3282	14.8645	51000	3.5475	0.3709
3.2465	15.1559	52000	3.5665	0.3697
3.2937	15.4474	53000	3.5581	0.3703
3.2937	15.7389	54000	3.5487	0.3708
3.2128	16.0303	55000	3.5594	0.3705
3.2609	16.3218	56000	3.5590	0.3704
3.2925	16.6133	57000	3.5524	0.3710
3.3014	16.9047	58000	3.5440	0.3715
3.2382	17.1962	59000	3.5594	0.3706
3.2592	17.4876	60000	3.5504	0.3715
3.2833	17.7791	61000	3.5436	0.3716
3.1912	18.0705	62000	3.5625	0.3712
3.2418	18.3620	63000	3.5581	0.3712
3.2645	18.6535	64000	3.5476	0.3717
3.2707	18.9450	65000	3.5412	0.3723
3.2168	19.2364	66000	3.5561	0.3716
3.2412	19.5279	67000	3.5490	0.3720
3.2551	19.8193	68000	3.5426	0.3723
3.1671	20.1108	69000	3.5627	0.3713
3.2246	20.4022	70000	3.5558	0.3717
3.2415	20.6937	71000	3.5469	0.3722
3.2584	20.9852	72000	3.5413	0.3728
3.205	21.2766	73000	3.5554	0.3719
3.2316	21.5681	74000	3.5495	0.3721
3.2471	21.8596	75000	3.5400	0.3725
3.1753	22.1510	76000	3.5588	0.3722
3.2107	22.4425	77000	3.5535	0.3721
3.2268	22.7339	78000	3.5458	0.3726
3.1343	23.0254	79000	3.5561	0.3722
3.1909	23.3168	80000	3.5590	0.3724
3.2069	23.6083	81000	3.5473	0.3729
3.225	23.8998	82000	3.5409	0.3730
3.1586	24.1912	83000	3.5588	0.3724
3.2006	24.4827	84000	3.5537	0.3726
3.2129	24.7742	85000	3.5433	0.3731
3.1303	25.0656	86000	3.5617	0.3723
3.1661	25.3571	87000	3.5547	0.3729
3.1836	25.6485	88000	3.5496	0.3730
3.221	25.9400	89000	3.5406	0.3737
3.142	26.2314	90000	3.5594	0.3726
3.1726	26.5229	91000	3.5532	0.3729
3.1958	26.8144	92000	3.5408	0.3736
3.1155	27.1058	93000	3.5603	0.3728
3.1699	27.3973	94000	3.5575	0.3728
3.1749	27.6888	95000	3.5468	0.3733

Framework versions

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