scenario-KD-PO-CDF-CL-D2_data-cl-cardiff_cl_only_delta-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: 15.4781
• Accuracy: 0.4043
• F1: 0.3989

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: 7777
• 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	11.8950	0.3735	0.3391
13.6959	2.17	500	11.1790	0.3850	0.3767
13.6959	3.26	750	11.1554	0.4082	0.4075
11.138	4.35	1000	11.3070	0.4090	0.4082
11.138	5.43	1250	11.4384	0.4043	0.4029
9.5889	6.52	1500	11.3483	0.4090	0.4061
9.5889	7.61	1750	12.4116	0.4020	0.3985
8.2638	8.7	2000	11.7820	0.3997	0.3798
8.2638	9.78	2250	11.9709	0.4012	0.4001
6.9206	10.87	2500	12.7419	0.4198	0.4197
6.9206	11.96	2750	13.1813	0.4074	0.4078
5.993	13.04	3000	13.5405	0.4074	0.4062
5.993	14.13	3250	13.6698	0.3920	0.3860
5.1998	15.22	3500	13.8872	0.3989	0.3991
5.1998	16.3	3750	13.8954	0.4074	0.4076
4.6117	17.39	4000	14.2214	0.4090	0.4090
4.6117	18.48	4250	13.7015	0.4097	0.4033
4.0848	19.57	4500	14.8333	0.3912	0.3798
4.0848	20.65	4750	14.0279	0.4035	0.3989
3.7518	21.74	5000	14.6064	0.4035	0.4024
3.7518	22.83	5250	14.6257	0.3943	0.3914
3.285	23.91	5500	14.5868	0.4020	0.3968
3.285	25.0	5750	15.5230	0.3843	0.3756
3.0056	26.09	6000	14.9209	0.3904	0.3811
3.0056	27.17	6250	14.9466	0.3981	0.3968
2.8364	28.26	6500	15.5711	0.3881	0.3824
2.8364	29.35	6750	15.4781	0.4043	0.3989

Framework versions

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