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

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: 5.5684
• Accuracy: 0.4676
• F1: 0.4662

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: 11213
• 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	1.8346	0.4684	0.4686
0.5182	2.17	500	2.4903	0.4745	0.4735
0.5182	3.26	750	2.5207	0.4830	0.4775
0.2149	4.35	1000	3.1950	0.4468	0.4376
0.2149	5.43	1250	3.2340	0.4738	0.4669
0.1358	6.52	1500	3.7896	0.4761	0.4750
0.1358	7.61	1750	3.2005	0.4954	0.4933
0.0968	8.7	2000	3.7302	0.4491	0.4429
0.0968	9.78	2250	4.0015	0.4691	0.4662
0.0624	10.87	2500	4.1400	0.4514	0.4462
0.0624	11.96	2750	4.6716	0.4460	0.4388
0.0508	13.04	3000	4.6683	0.4545	0.4531
0.0508	14.13	3250	4.9959	0.4398	0.4311
0.0408	15.22	3500	4.6543	0.4537	0.4541
0.0408	16.3	3750	4.7483	0.4753	0.4765
0.0272	17.39	4000	5.0125	0.4468	0.4446
0.0272	18.48	4250	5.1208	0.4475	0.4438
0.0241	19.57	4500	5.0070	0.4599	0.4586
0.0241	20.65	4750	4.8782	0.4761	0.4759
0.0187	21.74	5000	5.1882	0.4599	0.4608
0.0187	22.83	5250	5.1202	0.4761	0.4761
0.0086	23.91	5500	5.4158	0.4622	0.4616
0.0086	25.0	5750	5.4933	0.4637	0.4620
0.0088	26.09	6000	5.5752	0.4622	0.4604
0.0088	27.17	6250	5.5790	0.4668	0.4648
0.0063	28.26	6500	5.5069	0.4614	0.4603
0.0063	29.35	6750	5.5684	0.4676	0.4662

Framework versions

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