scenario-TCR_data-cl-cardiff_cl_only

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: 4.9878
• Accuracy: 0.5278
• F1: 0.5292

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	1.09	250	1.1991	0.5154	0.5166
0.731	2.17	500	1.5346	0.5316	0.5279
0.731	3.26	750	1.6658	0.5255	0.5251
0.3491	4.35	1000	1.9635	0.5185	0.5189
0.3491	5.43	1250	2.1732	0.5231	0.5221
0.1838	6.52	1500	3.0035	0.5239	0.5256
0.1838	7.61	1750	2.9315	0.5239	0.5258
0.122	8.7	2000	2.8799	0.5039	0.5009
0.122	9.78	2250	3.0551	0.5023	0.5037
0.0746	10.87	2500	3.2668	0.5262	0.5279
0.0746	11.96	2750	3.3828	0.5046	0.5062
0.0434	13.04	3000	3.8937	0.4954	0.4929
0.0434	14.13	3250	3.7629	0.5224	0.5235
0.0369	15.22	3500	4.1508	0.4931	0.4880
0.0369	16.3	3750	4.2268	0.5239	0.5240
0.0186	17.39	4000	4.3692	0.5054	0.5057
0.0186	18.48	4250	4.3635	0.5108	0.5108
0.0156	19.57	4500	4.4833	0.5062	0.5039
0.0156	20.65	4750	4.5300	0.5039	0.5043
0.0093	21.74	5000	4.5612	0.5239	0.5236
0.0093	22.83	5250	4.7381	0.5208	0.5216
0.0088	23.91	5500	4.6106	0.5324	0.5334
0.0088	25.0	5750	4.8040	0.5255	0.5269
0.0039	26.09	6000	4.8616	0.5262	0.5283
0.0039	27.17	6250	4.9228	0.5231	0.5247
0.0052	28.26	6500	5.1665	0.5008	0.5012
0.0052	29.35	6750	4.9878	0.5278	0.5292

Framework versions

• Transformers 4.33.3
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.13.3