metadata
license: apache-2.0
tags:
- generated_from_trainer
datasets:
- tweet_eval
metrics:
- precision
- recall
model-index:
- name: bert-emotion
results:
- task:
name: Text Classification
type: text-classification
dataset:
name: tweet_eval
type: tweet_eval
config: emotion
split: train
args: emotion
metrics:
- name: Precision
type: precision
value: 0.6916963308315679
- name: Recall
type: recall
value: 0.7047852871456702
bert-emotion
This model is a fine-tuned version of distilbert-base-cased on the tweet_eval dataset. It achieves the following results on the evaluation set:
- Loss: 1.3717
- Precision: 0.6917
- Recall: 0.7048
- Fscore: 0.6955
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: 4
- eval_batch_size: 4
- 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 | Precision | Recall | Fscore |
|---|---|---|---|---|---|---|
| 0.8838 | 1.0 | 815 | 0.7944 | 0.7238 | 0.6662 | 0.6860 |
| 0.5708 | 2.0 | 1630 | 1.0606 | 0.6594 | 0.6139 | 0.6299 |
| 0.3045 | 3.0 | 2445 | 1.3717 | 0.6917 | 0.7048 | 0.6955 |
Framework versions
- Transformers 4.25.1
- Pytorch 1.13.0+cu116
- Datasets 2.8.0
- Tokenizers 0.13.2