exceptions_exp2_swap_take_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: 3.5580
• Accuracy: 0.3696

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
4.8464	0.2911	1000	4.7594	0.2539
4.3521	0.5822	2000	4.2881	0.2991
4.1566	0.8733	3000	4.1013	0.3151
3.9878	1.1642	4000	3.9968	0.3243
3.929	1.4553	5000	3.9227	0.3306
3.882	1.7464	6000	3.8612	0.3368
3.7514	2.0373	7000	3.8177	0.3410
3.7591	2.3284	8000	3.7853	0.3441
3.7414	2.6195	9000	3.7578	0.3470
3.7379	2.9106	10000	3.7288	0.3494
3.6401	3.2014	11000	3.7171	0.3513
3.6573	3.4925	12000	3.6981	0.3528
3.65	3.7837	13000	3.6791	0.3548
3.5373	4.0745	14000	3.6713	0.3562
3.5801	4.3656	15000	3.6646	0.3567
3.5738	4.6567	16000	3.6492	0.3581
3.5826	4.9478	17000	3.6372	0.3595
3.4951	5.2387	18000	3.6363	0.3600
3.5197	5.5298	19000	3.6278	0.3612
3.5292	5.8209	20000	3.6149	0.3619
3.455	6.1118	21000	3.6189	0.3623
3.4808	6.4029	22000	3.6109	0.3629
3.4874	6.6940	23000	3.6019	0.3640
3.5009	6.9851	24000	3.5920	0.3645
3.4171	7.2760	25000	3.5996	0.3644
3.4489	7.5671	26000	3.5917	0.3653
3.4756	7.8582	27000	3.5831	0.3660
3.3848	8.1490	28000	3.5893	0.3660
3.4079	8.4401	29000	3.5840	0.3665
3.442	8.7313	30000	3.5751	0.3670
3.3168	9.0221	31000	3.5812	0.3672
3.3671	9.3132	32000	3.5807	0.3671
3.3954	9.6043	33000	3.5703	0.3679
3.4242	9.8954	34000	3.5623	0.3685
3.3374	10.1863	35000	3.5748	0.3684
3.3663	10.4774	36000	3.5679	0.3686
3.3749	10.7685	37000	3.5612	0.3691
3.2955	11.0594	38000	3.5686	0.3690
3.3363	11.3505	39000	3.5664	0.3693
3.3616	11.6416	40000	3.5580	0.3696
3.3579	11.9327	41000	3.5487	0.3705
3.3007	12.2236	42000	3.5635	0.3699
3.3295	12.5147	43000	3.5563	0.3703
3.3457	12.8058	44000	3.5468	0.3710
3.2691	13.0966	45000	3.5635	0.3702
3.3058	13.3878	46000	3.5589	0.3706
3.3208	13.6789	47000	3.5463	0.3712
3.3464	13.9700	48000	3.5430	0.3715
3.277	14.2608	49000	3.5569	0.3711
3.3024	14.5519	50000	3.5508	0.3717
3.3314	14.8430	51000	3.5426	0.3720
3.2462	15.1339	52000	3.5563	0.3713
3.2738	15.4250	53000	3.5519	0.3717
3.2858	15.7161	54000	3.5424	0.3725
3.2604	16.0070	55000	3.5512	0.3719
3.2499	16.2981	56000	3.5505	0.3721
3.2844	16.5892	57000	3.5447	0.3725
3.296	16.8803	58000	3.5371	0.3730
3.2321	17.1712	59000	3.5517	0.3726
3.2628	17.4623	60000	3.5460	0.3725
3.2777	17.7534	61000	3.5383	0.3728
3.1832	18.0442	62000	3.5523	0.3726
3.2448	18.3354	63000	3.5476	0.3726
3.267	18.6265	64000	3.5379	0.3733
3.2649	18.9176	65000	3.5323	0.3739
3.2083	19.2084	66000	3.5488	0.3731
3.2501	19.4995	67000	3.5435	0.3733
3.2676	19.7906	68000	3.5324	0.3740
3.1714	20.0815	69000	3.5479	0.3732
3.2122	20.3726	70000	3.5464	0.3733
3.2431	20.6637	71000	3.5374	0.3737
3.2458	20.9548	72000	3.5308	0.3742
3.2011	21.2457	73000	3.5499	0.3733
3.2208	21.5368	74000	3.5408	0.3736
3.2262	21.8279	75000	3.5320	0.3742
3.1639	22.1188	76000	3.5485	0.3736
3.206	22.4099	77000	3.5422	0.3739
3.219	22.7010	78000	3.5341	0.3745
3.2273	22.9921	79000	3.5265	0.3748
3.1943	23.2830	80000	3.5470	0.3740
3.2028	23.5741	81000	3.5418	0.3742
3.2035	23.8652	82000	3.5329	0.3747
3.1465	24.1560	83000	3.5498	0.3740
3.1913	24.4471	84000	3.5449	0.3741
3.2039	24.7382	85000	3.5355	0.3748
3.121	25.0291	86000	3.5470	0.3742
3.1669	25.3202	87000	3.5448	0.3745
3.1907	25.6113	88000	3.5377	0.3747
3.2044	25.9024	89000	3.5321	0.3750
3.143	26.1933	90000	3.5488	0.3741
3.1736	26.4844	91000	3.5395	0.3746
3.1889	26.7755	92000	3.5368	0.3748
3.1075	27.0664	93000	3.5505	0.3745
3.1592	27.3575	94000	3.5443	0.3747
3.1704	27.6486	95000	3.5370	0.3753
3.1811	27.9397	96000	3.5304	0.3754
3.1194	28.2306	97000	3.5513	0.3747
3.1489	28.5217	98000	3.5430	0.3749
3.1748	28.8128	99000	3.5335	0.3754

Framework versions

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