exceptions_exp2_swap_0.7_resemble_to_push_3591

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

• Loss: 3.5817
• Accuracy: 0.3664

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	Validation Loss	Accuracy
4.8461	0.2915	1000	4.7567	0.2541
4.3604	0.5831	2000	4.2898	0.2985
4.1567	0.8746	3000	4.0985	0.3151
3.9986	1.1662	4000	3.9921	0.3248
3.9373	1.4577	5000	3.9170	0.3314
3.8882	1.7493	6000	3.8562	0.3365
3.7581	2.0408	7000	3.8158	0.3408
3.759	2.3324	8000	3.7855	0.3439
3.7395	2.6239	9000	3.7556	0.3465
3.7254	2.9155	10000	3.7292	0.3486
3.6483	3.2070	11000	3.7166	0.3508
3.6496	3.4985	12000	3.6988	0.3526
3.6423	3.7901	13000	3.6829	0.3543
3.5459	4.0816	14000	3.6752	0.3554
3.5726	4.3732	15000	3.6619	0.3566
3.5837	4.6647	16000	3.6528	0.3575
3.5821	4.9563	17000	3.6386	0.3588
3.4951	5.2478	18000	3.6409	0.3594
3.5312	5.5394	19000	3.6301	0.3602
3.5378	5.8309	20000	3.6166	0.3614
3.455	6.1224	21000	3.6225	0.3618
3.4971	6.4140	22000	3.6113	0.3621
3.4898	6.7055	23000	3.6046	0.3630
3.5051	6.9971	24000	3.5943	0.3639
3.4258	7.2886	25000	3.6033	0.3638
3.447	7.5802	26000	3.5958	0.3644
3.4602	7.8717	27000	3.5842	0.3654
3.3913	8.1633	28000	3.5940	0.3651
3.4033	8.4548	29000	3.5876	0.3656
3.4196	8.7464	30000	3.5817	0.3664
3.342	9.0379	31000	3.5826	0.3666
3.3806	9.3294	32000	3.5814	0.3665
3.3999	9.6210	33000	3.5733	0.3668
3.4187	9.9125	34000	3.5644	0.3678
3.3342	10.2041	35000	3.5776	0.3674
3.3701	10.4956	36000	3.5709	0.3679
3.3973	10.7872	37000	3.5622	0.3685
3.2994	11.0787	38000	3.5739	0.3681
3.3359	11.3703	39000	3.5706	0.3684
3.3685	11.6618	40000	3.5623	0.3690
3.3763	11.9534	41000	3.5533	0.3695
3.3149	12.2449	42000	3.5681	0.3688
3.3458	12.5364	43000	3.5643	0.3694
3.3536	12.8280	44000	3.5535	0.3699
3.2745	13.1195	45000	3.5675	0.3693
3.313	13.4111	46000	3.5585	0.3698
3.3432	13.7026	47000	3.5534	0.3703
3.335	13.9942	48000	3.5475	0.3709
3.2875	14.2857	49000	3.5624	0.3702
3.2927	14.5773	50000	3.5544	0.3707
3.3141	14.8688	51000	3.5488	0.3711
3.2604	15.1603	52000	3.5632	0.3702
3.2822	15.4519	53000	3.5558	0.3709
3.301	15.7434	54000	3.5482	0.3712
3.2212	16.0350	55000	3.5553	0.3713
3.2637	16.3265	56000	3.5572	0.3711
3.2804	16.6181	57000	3.5539	0.3715
3.2932	16.9096	58000	3.5391	0.3721
3.2292	17.2012	59000	3.5565	0.3711
3.258	17.4927	60000	3.5518	0.3717
3.2802	17.7843	61000	3.5472	0.3720
3.1922	18.0758	62000	3.5556	0.3717
3.2534	18.3673	63000	3.5503	0.3719
3.272	18.6589	64000	3.5442	0.3723
3.2869	18.9504	65000	3.5363	0.3727
3.2136	19.2420	66000	3.5536	0.3719
3.2487	19.5335	67000	3.5466	0.3725
3.2645	19.8251	68000	3.5391	0.3729
3.1938	20.1166	69000	3.5546	0.3722
3.2414	20.4082	70000	3.5474	0.3724
3.2423	20.6997	71000	3.5434	0.3728
3.2651	20.9913	72000	3.5367	0.3736
3.1924	21.2828	73000	3.5529	0.3725
3.2184	21.5743	74000	3.5442	0.3727
3.2549	21.8659	75000	3.5355	0.3732
3.1722	22.1574	76000	3.5517	0.3728
3.2154	22.4490	77000	3.5479	0.3727
3.2281	22.7405	78000	3.5387	0.3735
3.1397	23.0321	79000	3.5523	0.3729
3.1908	23.3236	80000	3.5516	0.3729
3.2029	23.6152	81000	3.5432	0.3736
3.2237	23.9067	82000	3.5350	0.3736
3.1548	24.1983	83000	3.5511	0.3729
3.1904	24.4898	84000	3.5476	0.3733
3.2157	24.7813	85000	3.5365	0.3738
3.1204	25.0729	86000	3.5527	0.3732
3.1618	25.3644	87000	3.5524	0.3733
3.1961	25.6560	88000	3.5458	0.3736
3.2076	25.9475	89000	3.5333	0.3744
3.1366	26.2391	90000	3.5551	0.3733
3.1707	26.5306	91000	3.5442	0.3738
3.1893	26.8222	92000	3.5357	0.3744
3.1296	27.1137	93000	3.5540	0.3733
3.1563	27.4052	94000	3.5485	0.3737
3.1791	27.6968	95000	3.5417	0.3742
3.1909	27.9883	96000	3.5375	0.3744
3.1427	28.2799	97000	3.5489	0.3740
3.162	28.5714	98000	3.5437	0.3739
3.1718	28.8630	99000	3.5374	0.3746
3.1156	29.1545	100000	3.5539	0.3736
3.1491	29.4461	101000	3.5491	0.3741
3.165	29.7376	102000	3.5435	0.3742
3.0771	30.0292	103000	3.5540	0.3739
3.133	30.3207	104000	3.5553	0.3739
3.1425	30.6122	105000	3.5434	0.3745
3.1759	30.9038	106000	3.5376	0.3748
3.092	31.1953	107000	3.5543	0.3741
3.1314	31.4869	108000	3.5476	0.3743
3.1488	31.7784	109000	3.5435	0.3746

Framework versions

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