metadata
metrics:
- accuracy
model-index:
- name: dear-jarvis-monolith-xed-en
dear-jarvis-monolith-xed-en
This model was trained from scratch on an unkown dataset. It achieves the following results on the evaluation set:
- Loss: 1.7655
- Accuracy: 0.4678
Model description
LABEL_0 = neutral
LABEL_1 = anger
LABEL_2 = anticipation
LABEL_3 = disgust
LABEL_4 = fear
LABEL_5 = joy
LABEL_6 = sadness
LABEL_7 = surprise
LABEL_8 = trust
Labels are based on Plutchik's model of emotions and may be combined:
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: 8
- eval_batch_size: 8
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 3
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy |
|---|---|---|---|---|
| 1.5876 | 1.0 | 2003 | 1.5386 | 0.4596 |
| 1.1935 | 2.0 | 4006 | 1.5512 | 0.4803 |
| 0.7676 | 3.0 | 6009 | 1.7655 | 0.4678 |
Framework versions
- Transformers 4.6.1
- Pytorch 1.8.1+cu101
- Datasets 1.8.0
- Tokenizers 0.10.3