exceptions_exp2_swap_0.3_cost_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: 3.5586
• Accuracy: 0.3691

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
4.8593	0.2915	1000	0.2526	4.7654
4.3289	0.5831	2000	0.2982	4.2945
4.1479	0.8746	3000	0.3142	4.1063
3.9972	1.1662	4000	0.3242	3.9937
3.9327	1.4577	5000	0.3313	3.9192
3.8835	1.7493	6000	0.3364	3.8616
3.7384	2.0408	7000	0.3410	3.8179
3.7687	2.3324	8000	0.3436	3.7878
3.7421	2.6239	9000	0.3464	3.7564
3.7386	2.9155	10000	0.3489	3.7297
3.6418	3.2070	11000	0.3507	3.7163
3.6416	3.4985	12000	0.3526	3.7019
3.6519	3.7901	13000	0.3543	3.6797
3.5328	4.0816	14000	0.3555	3.6727
3.5737	4.3732	15000	0.3568	3.6612
3.584	4.6647	16000	0.3577	3.6496
3.5798	4.9563	17000	0.3591	3.6358
3.5091	5.2478	18000	0.3597	3.6373
3.5288	5.5394	19000	0.3603	3.6277
3.535	5.8309	20000	0.3615	3.6166
3.4419	6.1224	21000	0.3621	3.6196
3.4774	6.4140	22000	0.3624	3.6104
3.487	6.7055	23000	0.3631	3.6031
3.4925	6.9971	24000	0.3642	3.5918
3.4321	7.2886	25000	0.3639	3.6004
3.4443	7.5802	26000	0.3647	3.5903
3.4627	7.8717	27000	0.3654	3.5863
3.383	8.1633	28000	0.3651	3.5941
3.4144	8.4548	29000	0.3660	3.5875
3.4243	8.7464	30000	0.3667	3.5769
3.3265	9.0379	31000	0.3664	3.5819
3.3732	9.3294	32000	0.3666	3.5836
3.4029	9.6210	33000	0.3673	3.5709
3.412	9.9125	34000	0.3676	3.5678
3.3442	10.2041	35000	0.3676	3.5765
3.367	10.4956	36000	0.3678	3.5701
3.3839	10.7872	37000	0.3686	3.5612
3.3034	11.0787	38000	0.3682	3.5731
3.3383	11.3703	39000	0.3686	3.5677
3.3669	11.6618	40000	0.3691	3.5586
3.3862	11.9534	41000	0.3696	3.5519
3.3147	12.2449	42000	0.3692	3.5673
3.3332	12.5364	43000	0.3694	3.5584
3.3562	12.8280	44000	0.3699	3.5525
3.2737	13.1195	45000	0.3693	3.5666
3.3177	13.4111	46000	0.3696	3.5580
3.3253	13.7026	47000	0.3704	3.5513
3.3387	13.9942	48000	0.3709	3.5473
3.2835	14.2857	49000	0.3702	3.5606
3.3224	14.5773	50000	0.3706	3.5517
3.331	14.8688	51000	0.3711	3.5432
3.2473	15.1603	52000	0.3705	3.5595
3.2867	15.4519	53000	0.3709	3.5504
3.304	15.7434	54000	0.3717	3.5452
3.2131	16.0350	55000	0.3708	3.5580
3.2648	16.3265	56000	0.3710	3.5541
3.2874	16.6181	57000	0.3713	3.5483
3.3064	16.9096	58000	0.3720	3.5407
3.2338	17.2012	59000	0.3710	3.5592
3.2579	17.4927	60000	0.3714	3.5543
3.2919	17.7843	61000	0.3722	3.5450
3.2019	18.0758	62000	0.3713	3.5572
3.235	18.3673	63000	0.3717	3.5555
3.2601	18.6589	64000	0.3721	3.5437
3.2662	18.9504	65000	0.3728	3.5354
3.2107	19.2420	66000	0.3718	3.5546
3.2552	19.5335	67000	0.3721	3.5477
3.2683	19.8251	68000	0.3729	3.5381
3.1937	20.1166	69000	0.3720	3.5565
3.2322	20.4082	70000	0.3725	3.5476
3.2443	20.6997	71000	0.3729	3.5433
3.2505	20.9913	72000	0.3730	3.5371
3.2086	21.2828	73000	0.3724	3.5527
3.2317	21.5743	74000	0.3729	3.5469
3.2386	21.8659	75000	0.3736	3.5376
3.1664	22.1574	76000	0.3729	3.5540
3.2043	22.4490	77000	0.3729	3.5509
3.2342	22.7405	78000	0.3734	3.5422
3.1303	23.0321	79000	0.3730	3.5534
3.1847	23.3236	80000	0.3728	3.5516
3.1867	23.6152	81000	3.5575	0.3726
3.2139	23.9067	82000	3.5493	0.3730
3.1722	24.1983	83000	3.5635	0.3725
3.203	24.4898	84000	3.5515	0.3731
3.2029	24.7813	85000	3.5395	0.3739
3.1271	25.0729	86000	3.5533	0.3731
3.1722	25.3644	87000	3.5511	0.3733
3.1964	25.6560	88000	3.5440	0.3738
3.2176	25.9475	89000	3.5378	0.3741
3.1526	26.2391	90000	3.5583	0.3730
3.1819	26.5306	91000	3.5454	0.3737
3.2004	26.8222	92000	3.5387	0.3741
3.1162	27.1137	93000	3.5589	0.3734
3.1631	27.4052	94000	3.5503	0.3736
3.1695	27.6968	95000	3.5414	0.3740
3.2012	27.9883	96000	3.5377	0.3742
3.1274	28.2799	97000	3.5568	0.3731
3.1686	28.5714	98000	3.5454	0.3738
3.1767	28.8630	99000	3.5356	0.3747
3.1033	29.1545	100000	3.5560	0.3736

Framework versions

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