exceptions_exp2_swap_require_to_carry_3591

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

• Loss: 3.5545
• Accuracy: 0.3743

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: 3591
• 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
4.8353	0.2911	1000	0.2531	4.7629
4.3374	0.5822	2000	0.2996	4.2795
4.1489	0.8733	3000	0.3155	4.0955
3.996	1.1642	4000	0.3252	3.9965
3.9347	1.4553	5000	0.3318	3.9166
3.8665	1.7464	6000	0.3372	3.8597
3.7343	2.0373	7000	0.3416	3.8142
3.7434	2.3284	8000	0.3447	3.7804
3.7318	2.6195	9000	0.3477	3.7533
3.7166	2.9106	10000	0.3496	3.7269
3.6328	3.2014	11000	0.3518	3.7136
3.6293	3.4925	12000	0.3536	3.6956
3.6383	3.7837	13000	0.3553	3.6771
3.5359	4.0745	14000	0.3568	3.6693
3.5563	4.3656	15000	0.3578	3.6573
3.569	4.6567	16000	0.3587	3.6432
3.5858	4.9478	17000	0.3601	3.6322
3.4991	5.2387	18000	0.3605	3.6344
3.5221	5.5298	19000	0.3616	3.6235
3.5329	5.8209	20000	0.3624	3.6119
3.4393	6.1118	21000	0.3629	3.6139
3.4739	6.4029	22000	0.3634	3.6083
3.4852	6.6940	23000	0.3639	3.6001
3.4901	6.9851	24000	0.3646	3.5918
3.4147	7.2760	25000	0.3649	3.5968
3.4439	7.5671	26000	0.3658	3.5878
3.4694	7.8582	27000	0.3663	3.5779
3.3723	8.1490	28000	0.3661	3.5911
3.405	8.4401	29000	0.3668	3.5804
3.4347	8.7313	30000	0.3675	3.5715
3.3197	9.0221	31000	0.3673	3.5790
3.3719	9.3132	32000	0.3678	3.5758
3.394	9.6043	33000	0.3685	3.5682
3.4089	9.8954	34000	0.3688	3.5599
3.3351	10.1863	35000	0.3686	3.5718
3.3608	10.4774	36000	0.3686	3.5644
3.3839	10.7685	37000	0.3696	3.5576
3.2872	11.0594	38000	0.3697	3.5635
3.3239	11.3505	39000	0.3692	3.5649
3.3604	11.6416	40000	0.3702	3.5563
3.3633	11.9327	41000	0.3705	3.5486
3.3018	12.2236	42000	0.3699	3.5617
3.3237	12.5147	43000	0.3704	3.5561
3.3553	12.8058	44000	0.3710	3.5454
3.266	13.0966	45000	0.3703	3.5622
3.298	13.3878	46000	0.3706	3.5562
3.3193	13.6789	47000	0.3712	3.5511
3.3377	13.9700	48000	0.3719	3.5380
3.2707	14.2608	49000	0.3711	3.5559
3.2986	14.5519	50000	0.3714	3.5475
3.3352	14.8430	51000	0.3721	3.5395
3.2328	15.1339	52000	0.3717	3.5531
3.278	15.4250	53000	0.3720	3.5493
3.2919	15.7161	54000	0.3724	3.5413
3.2682	16.0070	55000	0.3720	3.5475
3.2366	16.2981	56000	0.3725	3.5493
3.2881	16.5892	57000	0.3725	3.5421
3.2976	16.8803	58000	0.3728	3.5371
3.2211	17.1712	59000	0.3723	3.5510
3.265	17.4623	60000	0.3727	3.5464
3.2771	17.7534	61000	0.3732	3.5357
3.1751	18.0442	62000	0.3724	3.5497
3.2348	18.3354	63000	0.3728	3.5502
3.2587	18.6265	64000	0.3730	3.5414
3.2776	18.9176	65000	0.3738	3.5317
3.1939	19.2084	66000	0.3728	3.5513
3.2337	19.4995	67000	0.3731	3.5424
3.2667	19.7906	68000	0.3738	3.5364
3.1624	20.0815	69000	0.3732	3.5489
3.2091	20.3726	70000	0.3737	3.5438
3.245	20.6637	71000	0.3735	3.5370
3.2399	20.9548	72000	0.3742	3.5309
3.1951	21.2457	73000	0.3732	3.5491
3.2234	21.5368	74000	0.3738	3.5366
3.2386	21.8279	75000	0.3742	3.5339
3.1712	22.1188	76000	0.3735	3.5475
3.1945	22.4099	77000	0.3739	3.5480
3.2309	22.7010	78000	0.3742	3.5343
3.2372	22.9921	79000	0.3750	3.5286
3.1752	23.2830	80000	0.3739	3.5466
3.1701	23.5741	81000	3.5533	0.3735
3.1977	23.8652	82000	3.5408	0.3741
3.154	24.1563	83000	3.5521	0.3736
3.1835	24.4474	84000	3.5442	0.3743
3.205	24.7385	85000	3.5376	0.3747
3.1135	25.0294	86000	3.5489	0.3740
3.1552	25.3205	87000	3.5517	0.3737
3.1777	25.6116	88000	3.5409	0.3743
3.2144	25.9027	89000	3.5334	0.3751
3.1364	26.1936	90000	3.5503	0.3742
3.1675	26.4847	91000	3.5466	0.3746
3.201	26.7758	92000	3.5364	0.3751
3.1089	27.0667	93000	3.5548	0.3741
3.151	27.3578	94000	3.5441	0.3744
3.1649	27.6489	95000	3.5373	0.3750
3.1918	27.9400	96000	3.5342	0.3752
3.1161	28.2308	97000	3.5515	0.3745
3.1603	28.5219	98000	3.5428	0.3747
3.1628	28.8131	99000	3.5394	0.3754
3.087	29.1039	100000	3.5545	0.3743

Framework versions

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