scenario-NON-KD-PR-COPY-CDF-CL-D2_data-cl-cardiff_cl_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: 6.0210
• Accuracy: 0.4329
• F1: 0.4322

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	1.0870	250	1.2140	0.4375	0.4361
0.8987	2.1739	500	1.3995	0.4344	0.4312
0.8987	3.2609	750	1.7939	0.4367	0.4323
0.5501	4.3478	1000	1.7712	0.4321	0.4242
0.5501	5.4348	1250	2.3222	0.4421	0.4421
0.2588	6.5217	1500	2.8355	0.4306	0.4262
0.2588	7.6087	1750	3.2050	0.4144	0.4035
0.1477	8.6957	2000	3.2667	0.4275	0.4157
0.1477	9.7826	2250	3.8619	0.4282	0.4254
0.0975	10.8696	2500	3.6895	0.4414	0.4414
0.0975	11.9565	2750	4.3818	0.4321	0.4320
0.0673	13.0435	3000	4.3131	0.4275	0.4209
0.0673	14.1304	3250	4.2405	0.4306	0.4297
0.0477	15.2174	3500	4.7003	0.4306	0.4250
0.0477	16.3043	3750	4.9143	0.4375	0.4374
0.0293	17.3913	4000	4.8861	0.4406	0.4401
0.0293	18.4783	4250	4.8405	0.4475	0.4476
0.0243	19.5652	4500	4.9279	0.4275	0.4270
0.0243	20.6522	4750	5.4537	0.4352	0.4334
0.0177	21.7391	5000	5.4784	0.4398	0.4378
0.0177	22.8261	5250	6.1257	0.4182	0.4095
0.0124	23.9130	5500	5.7505	0.4344	0.4326
0.0124	25.0	5750	5.7485	0.4336	0.4302
0.0084	26.0870	6000	6.0192	0.4313	0.4314
0.0084	27.1739	6250	5.9173	0.4336	0.4329
0.0093	28.2609	6500	5.9415	0.4298	0.4287
0.0093	29.3478	6750	6.0210	0.4329	0.4322

Framework versions

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