scenario-NON-KD-PR-COPY-CDF-EN-D2_data-en-cardiff_eng_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: 5.0534
• Accuracy: 0.4511
• F1: 0.4469

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	1.7241	100	1.1154	0.4295	0.3980
No log	3.4483	200	1.3391	0.4660	0.4608
No log	5.1724	300	1.7241	0.4493	0.4444
No log	6.8966	400	1.7892	0.4735	0.4730
0.5647	8.6207	500	2.4031	0.4625	0.4611
0.5647	10.3448	600	2.7873	0.4475	0.4433
0.5647	12.0690	700	3.4191	0.4458	0.4424
0.5647	13.7931	800	3.8868	0.4581	0.4540
0.5647	15.5172	900	4.0222	0.4418	0.4385
0.0733	17.2414	1000	4.3198	0.4405	0.4390
0.0733	18.9655	1100	4.5281	0.4533	0.4498
0.0733	20.6897	1200	4.9279	0.4378	0.4247
0.0733	22.4138	1300	4.8784	0.4511	0.4461
0.0733	24.1379	1400	4.8101	0.4440	0.4423
0.0151	25.8621	1500	4.9054	0.4506	0.4490
0.0151	27.5862	1600	5.0583	0.4506	0.4453
0.0151	29.3103	1700	5.0534	0.4511	0.4469

Framework versions

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