scenario-NON-KD-PR-COPY-CDF-EN-D2_data-en-cardiff_eng_only66

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: 4.9528
• Accuracy: 0.4484
• F1: 0.4476

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: 66
• 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	1.1121	0.4330	0.4136
No log	3.4483	200	1.3612	0.4678	0.4532
No log	5.1724	300	1.8876	0.4308	0.4143
No log	6.8966	400	1.8980	0.4396	0.4354
0.5935	8.6207	500	2.3654	0.4563	0.4520
0.5935	10.3448	600	2.9355	0.4369	0.4286
0.5935	12.0690	700	3.2830	0.4418	0.4308
0.5935	13.7931	800	3.5565	0.4444	0.4436
0.5935	15.5172	900	3.9425	0.4427	0.4335
0.0878	17.2414	1000	4.1890	0.4541	0.4507
0.0878	18.9655	1100	4.4326	0.4572	0.4566
0.0878	20.6897	1200	4.5100	0.4621	0.4584
0.0878	22.4138	1300	4.7315	0.4533	0.4492
0.0878	24.1379	1400	4.8446	0.4528	0.4467
0.0129	25.8621	1500	4.8415	0.4533	0.4536
0.0129	27.5862	1600	4.9268	0.4519	0.4516
0.0129	29.3103	1700	4.9528	0.4484	0.4476

Framework versions

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