README.md

---
license: apache-2.0
base_model: distilbert-base-uncased
tags:
- generated_from_trainer
datasets:
- emotion
metrics:
- f1
model-index:
- name: distil-bert
  results:
  - task:
      name: Text Classification
      type: text-classification
    dataset:
      name: emotion
      type: emotion
      config: split
      split: validation
      args: split
    metrics:
    - name: F1
      type: f1
      value: 0.9295002701213645
---

<!-- This model card has been generated automatically according to the information the Trainer had access to. You
should probably proofread and complete it, then remove this comment. -->

# distil-bert

This model is a fine-tuned version of [distilbert-base-uncased](https://huggingface.co/distilbert-base-uncased) on the emotion dataset.
It achieves the following results on the evaluation set:
- Loss: 0.1749
- F1: 0.9295

## 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: 64
- eval_batch_size: 64
- 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 | F1     |
|:-------------:|:-----:|:----:|:---------------:|:------:|
| 0.8295        | 1.0   | 250  | 0.2760          | 0.9148 |
| 0.2167        | 2.0   | 500  | 0.1838          | 0.9326 |
| 0.1461        | 3.0   | 750  | 0.1749          | 0.9295 |


### Framework versions

- Transformers 4.39.3
- Pytorch 2.1.2
- Datasets 2.18.0
- Tokenizers 0.15.2