Heritage-in-Digital-Age-distilbert-base-uncased-expression-rating

This model is a fine-tuned version of distilbert-base-uncased on the None dataset. It achieves the following results on the evaluation set:

• Loss: 2.0133
• Accuracy: 0.3496

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: 20

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy
No log	1.0	18	1.9870	0.3333
No log	2.0	36	1.9731	0.3415
No log	3.0	54	2.0133	0.3496
No log	4.0	72	2.0809	0.3415
No log	5.0	90	2.1694	0.3008
No log	6.0	108	2.2611	0.2764
No log	7.0	126	2.2832	0.3008
No log	8.0	144	2.3670	0.2846
No log	9.0	162	2.4279	0.2683
No log	10.0	180	2.4460	0.3089
No log	11.0	198	2.5236	0.2846
No log	12.0	216	2.5896	0.3089
No log	13.0	234	2.6061	0.3008
No log	14.0	252	2.6813	0.2846
No log	15.0	270	2.6990	0.3252
No log	16.0	288	2.7439	0.3171
No log	17.0	306	2.7499	0.3415
No log	18.0	324	2.7737	0.3252
No log	19.0	342	2.7793	0.3252
No log	20.0	360	2.7775	0.3252

Framework versions

• Transformers 4.25.1
• Pytorch 1.13.0+cu116
• Datasets 2.8.0
• Tokenizers 0.13.2