metadata
license: apache-2.0
base_model: distilbert/distilbert-base-multilingual-cased
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: trained_danish
results: []
trained_danish
This model is a fine-tuned version of distilbert/distilbert-base-multilingual-cased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.0667
- Precision: 0.7791
- Recall: 0.7329
- F1: 0.7553
- Accuracy: 0.9807
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: 8
- eval_batch_size: 8
- seed: 42
- gradient_accumulation_steps: 4
- total_train_batch_size: 32
- 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 | F1 | Accuracy |
|---|---|---|---|---|---|---|---|
| No log | 1.0 | 137 | 0.0788 | 0.6736 | 0.6658 | 0.6697 | 0.9749 |
| No log | 2.0 | 274 | 0.0652 | 0.7653 | 0.7406 | 0.7528 | 0.9802 |
| No log | 3.0 | 411 | 0.0667 | 0.7791 | 0.7329 | 0.7553 | 0.9807 |
Framework versions
- Transformers 4.38.2
- Pytorch 2.1.2+cu118
- Datasets 2.18.0
- Tokenizers 0.15.2