scenario-NON-KD-PO-COPY-CDF-CL-D2_data-cl-cardiff_cl_only_beta

This model is a fine-tuned version of haryoaw/scenario-TCR_data-cl-cardiff_cl_only2 on the None dataset. It achieves the following results on the evaluation set:

• Loss: 5.5580
• Accuracy: 0.4414
• F1: 0.4359

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: 112233
• 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	1.7211	0.4823	0.4838
0.5121	2.17	500	1.9396	0.4807	0.4817
0.5121	3.26	750	2.3456	0.4684	0.4700
0.2161	4.35	1000	2.7111	0.4730	0.4701
0.2161	5.43	1250	3.4354	0.4360	0.4244
0.1194	6.52	1500	3.4859	0.4606	0.4586
0.1194	7.61	1750	3.1266	0.4560	0.4519
0.0941	8.7	2000	3.5782	0.4707	0.4693
0.0941	9.78	2250	4.1522	0.4460	0.4409
0.06	10.87	2500	4.4567	0.4576	0.4546
0.06	11.96	2750	4.2186	0.4560	0.4541
0.0399	13.04	3000	4.6178	0.4614	0.4591
0.0399	14.13	3250	4.3070	0.4738	0.4737
0.0407	15.22	3500	4.5620	0.4645	0.4597
0.0407	16.3	3750	4.6845	0.4560	0.4528
0.0316	17.39	4000	4.9378	0.4560	0.4521
0.0316	18.48	4250	4.9708	0.4784	0.4765
0.0207	19.57	4500	5.0696	0.4630	0.4618
0.0207	20.65	4750	5.2248	0.4406	0.4375
0.0148	21.74	5000	5.4988	0.4406	0.4337
0.0148	22.83	5250	5.2698	0.4537	0.4512
0.0121	23.91	5500	5.4028	0.4491	0.4460
0.0121	25.0	5750	5.6057	0.4421	0.4346
0.0121	26.09	6000	5.3745	0.4522	0.4503
0.0121	27.17	6250	5.5548	0.4545	0.4503
0.0068	28.26	6500	5.5513	0.4491	0.4454
0.0068	29.35	6750	5.5580	0.4414	0.4359

Framework versions

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