scenario-NON-KD-PO-COPY-CDF-CL-D2_data-cl-cardiff_cl_only55

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

• Loss: 6.0459
• Accuracy: 0.4653
• F1: 0.4654

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: 55
• 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	1.0870	250	1.2187	0.4653	0.4648
0.8415	2.1739	500	1.4947	0.4460	0.4252
0.8415	3.2609	750	1.9828	0.4599	0.4588
0.4583	4.3478	1000	2.0798	0.4676	0.4679
0.4583	5.4348	1250	2.6977	0.4259	0.4123
0.223	6.5217	1500	3.0180	0.4367	0.4324
0.223	7.6087	1750	2.9753	0.4275	0.4233
0.1214	8.6957	2000	3.6565	0.4722	0.4669
0.1214	9.7826	2250	4.2129	0.4491	0.4476
0.0857	10.8696	2500	4.1589	0.4576	0.4568
0.0857	11.9565	2750	4.4115	0.4483	0.4479
0.0562	13.0435	3000	4.2090	0.4645	0.4651
0.0562	14.1304	3250	4.9216	0.4660	0.4666
0.035	15.2174	3500	4.8460	0.4529	0.4476
0.035	16.3043	3750	4.8846	0.4522	0.4474
0.031	17.3913	4000	5.1002	0.4637	0.4637
0.031	18.4783	4250	5.6542	0.4591	0.4591
0.0162	19.5652	4500	5.4518	0.4630	0.4614
0.0162	20.6522	4750	5.5188	0.4622	0.4609
0.0174	21.7391	5000	5.6319	0.4614	0.4608
0.0174	22.8261	5250	5.4654	0.4630	0.4629
0.0125	23.9130	5500	5.7260	0.4722	0.4721
0.0125	25.0	5750	6.0229	0.4599	0.4585
0.0083	26.0870	6000	6.0355	0.4715	0.4712
0.0083	27.1739	6250	5.9799	0.4614	0.4609
0.0083	28.2609	6500	6.0428	0.4630	0.4621
0.0083	29.3478	6750	6.0459	0.4653	0.4654

Framework versions

• Transformers 4.44.2
• Pytorch 2.1.1+cu121
• Datasets 2.14.5
• Tokenizers 0.19.1