README.md

metadata

license: apache-2.0
tags:
  - generated_from_trainer
datasets: qfrodicio/gesture-prediction-9-classes
metrics:
  - accuracy
  - precision
  - recall
  - f1
model-index:
  - name: distilbert-finetuned-gesture-prediction-9-classes
    results: []

distilbert-finetuned-gesture-prediction-9-classes

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

• Loss: 0.6479
• Accuracy: 0.8214
• Precision: 0.8230
• Recall: 0.8214
• F1: 0.8172

It achieves the following results on the test set:

• Loss: 0.6475
• Accuracy: 0.8144
• Precision: 0.8144
• Recall: 0.8144
• F1: 0.8095

Model description

More information needed

Intended uses & limitations

More information needed

Training and evaluation data

The model has been trained on the qfrodicio/gesture-prediction-9-classes dataset

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

• learning_rate: 2e-05
• weight_decay: 0.01
• 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: 10

Training results

Training Loss	Epoch	Step	Validation Loss	Accuracy	Precision	Recall	F1
1.6636	1.0	87	0.9715	0.7270	0.6909	0.7270	0.6897
0.7503	2.0	174	0.7360	0.7987	0.7874	0.7987	0.7879
0.5283	3.0	261	0.6831	0.8056	0.8046	0.8056	0.8005
0.3853	4.0	348	0.6479	0.8214	0.8230	0.8214	0.8172
0.28	5.0	435	0.6570	0.8314	0.8348	0.8314	0.8289
0.2163	6.0	522	0.6887	0.8322	0.8346	0.8322	0.8298
0.158	7.0	609	0.7078	0.8336	0.8362	0.8336	0.8311
0.1308	8.0	696	0.7197	0.8415	0.8444	0.8415	0.8394
0.1061	9.0	783	0.7362	0.8419	0.8441	0.8419	0.8394
0.0947	10.0	870	0.7412	0.8435	0.8458	0.8435	0.8410

Framework versions

• Transformers 4.26.1
• Pytorch 1.13.1+cu116
• Datasets 2.10.1
• Tokenizers 0.13.2