scenario-KD-PO-CDF-CL-D2_data-cl-cardiff_cl_only_gamma-jason

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: 17.5985
• Accuracy: 0.3958
• F1: 0.3956

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: 88458
• 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	11.6483	0.3650	0.3609
13.703	2.17	500	11.1328	0.3843	0.3821
13.703	3.26	750	10.9310	0.3997	0.3991
11.1159	4.35	1000	11.8248	0.4035	0.3976
11.1159	5.43	1250	11.2452	0.4074	0.4047
9.5324	6.52	1500	12.0093	0.4082	0.4079
9.5324	7.61	1750	12.4283	0.4043	0.4042
8.2184	8.7	2000	12.1651	0.3897	0.3860
8.2184	9.78	2250	13.2395	0.4012	0.4006
7.0214	10.87	2500	13.4757	0.4028	0.4030
7.0214	11.96	2750	14.5726	0.3943	0.3878
6.0532	13.04	3000	15.9024	0.3966	0.3918
6.0532	14.13	3250	16.2467	0.3819	0.3672
5.3338	15.22	3500	14.5700	0.3912	0.3911
5.3338	16.3	3750	14.7870	0.3989	0.3969
4.7168	17.39	4000	16.6837	0.3804	0.3758
4.7168	18.48	4250	16.3479	0.3835	0.3807
4.052	19.57	4500	16.3096	0.3897	0.3860
4.052	20.65	4750	16.4666	0.3958	0.3947
3.6691	21.74	5000	16.7052	0.3935	0.3851
3.6691	22.83	5250	16.9439	0.4012	0.3985
3.3107	23.91	5500	16.8777	0.4051	0.4025
3.3107	25.0	5750	17.4662	0.3897	0.3866
2.9893	26.09	6000	17.5858	0.3951	0.3939
2.9893	27.17	6250	17.6884	0.3935	0.3928
2.8471	28.26	6500	17.7042	0.3881	0.3871
2.8471	29.35	6750	17.5985	0.3958	0.3956

Framework versions

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