cdp_hyl_fd

This model is a fine-tuned version of distilbert-base-uncased on an unknown dataset. It achieves the following results on the evaluation set:

• Loss: 0.4997
• Accuracy: 0.8235

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: 2e-05
• train_batch_size: 16
• eval_batch_size: 16
• seed: 42
• optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
• lr_scheduler_type: linear
• num_epochs: 25

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
0.7019	1.0	55	1.1939	0.5
0.1305	2.0	110	0.6574	0.7353
0.0242	3.0	165	0.5197	0.8235
0.0081	4.0	220	0.3666	0.8824
0.0051	5.0	275	0.4560	0.8529
0.0035	6.0	330	0.4470	0.8235
0.0026	7.0	385	0.4395	0.8529
0.0022	8.0	440	0.4486	0.8235
0.0018	9.0	495	0.4684	0.8235
0.0015	10.0	550	0.4644	0.8529
0.0013	11.0	605	0.4669	0.8235
0.0012	12.0	660	0.4657	0.8235
0.0011	13.0	715	0.4799	0.8235
0.001	14.0	770	0.4817	0.8235
0.0009	15.0	825	0.4998	0.8235
0.0008	16.0	880	0.4964	0.8235
0.0008	17.0	935	0.5025	0.8235
0.0007	18.0	990	0.4954	0.8235
0.0007	19.0	1045	0.4933	0.8235
0.0007	20.0	1100	0.5014	0.8235
0.0006	21.0	1155	0.4961	0.8235
0.0006	22.0	1210	0.4955	0.8235
0.0006	23.0	1265	0.4984	0.8235
0.0006	24.0	1320	0.4988	0.8235
0.0006	25.0	1375	0.4997	0.8235

Framework versions

• Transformers 4.36.1
• Pytorch 2.1.1+cu118
• Datasets 2.15.0
• Tokenizers 0.15.0