scenario-NON-KD-SCR-COPY-CDF-CL-D2_data-cl-cardiff_cl_only55

This model is a fine-tuned version of microsoft/mdeberta-v3-base on the None dataset. It achieves the following results on the evaluation set:

• Loss: 6.8361
• Accuracy: 0.3634
• F1: 0.3600

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: 5e-05
• train_batch_size: 32
• eval_batch_size: 32
• seed: 55
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 30

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	F1
No log	1.0870	250	1.4035	0.3627	0.3384
0.9258	2.1739	500	1.8269	0.3688	0.3652
0.9258	3.2609	750	2.2003	0.3696	0.3691
0.3143	4.3478	1000	3.2084	0.3850	0.3842
0.3143	5.4348	1250	3.4181	0.3719	0.3668
0.1172	6.5217	1500	3.9886	0.3688	0.3622
0.1172	7.6087	1750	4.2183	0.3650	0.3626
0.0592	8.6957	2000	4.6155	0.3665	0.3545
0.0592	9.7826	2250	4.7510	0.3727	0.3685
0.0394	10.8696	2500	5.1707	0.3688	0.3628
0.0394	11.9565	2750	5.0827	0.3681	0.3636
0.0238	13.0435	3000	5.5056	0.3665	0.3535
0.0238	14.1304	3250	5.3337	0.3704	0.3661
0.0171	15.2174	3500	5.7582	0.3735	0.3709
0.0171	16.3043	3750	5.9369	0.3665	0.3598
0.011	17.3913	4000	6.0815	0.3765	0.3719
0.011	18.4783	4250	6.1316	0.3819	0.3802
0.0043	19.5652	4500	6.3789	0.3727	0.3705
0.0043	20.6522	4750	6.4273	0.3673	0.3664
0.0064	21.7391	5000	6.3039	0.3758	0.3743
0.0064	22.8261	5250	6.5675	0.3619	0.3540
0.0031	23.9130	5500	6.5657	0.3688	0.3650
0.0031	25.0	5750	6.6382	0.3696	0.3666
0.0016	26.0870	6000	6.7416	0.3681	0.3643
0.0016	27.1739	6250	6.7141	0.3711	0.3677
0.0006	28.2609	6500	6.7905	0.3642	0.3600
0.0006	29.3478	6750	6.8361	0.3634	0.3600

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1