metadata
license: apache-2.0
base_model: distilbert-base-uncased
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
model-index:
- name: SentimentT2
results: []
SentimentT2
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: 0.3554
- Accuracy: 0.8507
- F1: 0.8568
- Auc Roc: 0.9199
- Log Loss: 0.3554
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: 1e-05
- train_batch_size: 16
- eval_batch_size: 20
- seed: 42
- gradient_accumulation_steps: 2
- total_train_batch_size: 32
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: cosine
- lr_scheduler_warmup_steps: 500
- num_epochs: 4
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Auc Roc | Log Loss |
|---|---|---|---|---|---|---|---|
| 0.6935 | 1.0 | 101 | 0.6756 | 0.7251 | 0.7427 | 0.8000 | 0.6756 |
| 0.5974 | 2.0 | 203 | 0.4756 | 0.8060 | 0.8251 | 0.8897 | 0.4756 |
| 0.4166 | 3.0 | 304 | 0.3724 | 0.8445 | 0.8489 | 0.9138 | 0.3724 |
| 0.3405 | 3.98 | 404 | 0.3554 | 0.8507 | 0.8568 | 0.9199 | 0.3554 |
Framework versions
- Transformers 4.35.2
- Pytorch 2.1.0+cu121
- Datasets 2.16.1
- Tokenizers 0.15.0