Hugging Face's logo

somosnlp-hackathon-2022
/
twitter_sexismo-finetuned-exist2021-metwo

Text Classification
Transformers
PyTorch
roberta
Inference Endpoints
Model card Files Community
1
twitter_sexismo-finetuned-exist2021-metwo / README.md
robertou2's picture
robertou2
Update README.md
b3ea271
preview code
|
raw
history blame
No virus
7.67 kB
metadata
license: apache-2.0
tags:
  - null
datasets:
  - EXIST Dataset
metrics:
  - accuracy
model-index:
  - name: twitter_sexismo-finetuned-exist2021
    results:
      - task:
          name: Text Classification
          type: text-classification
        dataset:
          name: EXIST Dataset
          type: EXIST Dataset
          args: es
        metrics:
          - name: Accuracy
            type: accuracy
            value: 0.8

twitter_sexismo-finetuned-exist2021

This model is a fine-tuned version of pysentimiento/robertuito-hate-speech on the EXIST dataset. It achieves the following results on the evaluation set:

  • Loss: 0.12
  • Accuracy: 0.80

Model description

Modelo para el Hackaton de Somos NLP para detección de sexismo en twitts en español

Intended uses & limitations

More information needed

Training and evaluation data

More information needed

Training procedure

Training hyperparameters

The following hyperparameters were used during training:

  • my_learning_rate = 2E-6
  • my_adam_epsilon = 1E-8
  • my_number_of_epochs = 15
  • my_warmup = 3
  • my_mini_batch_size = 32
  • optimizer: AdamW with betas=(0.9,0.999) and epsilon=1e-08
  • lr_scheduler_type: linear
  • num_epochs: 30

Training results

======== Epoch 1 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:53. Batch 100 of 143. Elapsed: 0:01:45.

Average training loss: 0.39 Training epoch took: 0:02:29

======== Epoch 2 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.36 Training epoch took: 0:02:29

======== Epoch 3 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.34 Training epoch took: 0:02:29

======== Epoch 4 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.33 Training epoch took: 0:02:29

======== Epoch 5 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.31 Training epoch took: 0:02:29

======== Epoch 6 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.29 Training epoch took: 0:02:29

======== Epoch 7 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.28 Training epoch took: 0:02:29

======== Epoch 8 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.27 Training epoch took: 0:02:29

======== Epoch 9 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.25 Training epoch took: 0:02:28

======== Epoch 10 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.24 Training epoch took: 0:02:29

======== Epoch 11 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.23 Training epoch took: 0:02:28

======== Epoch 12 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.22 Training epoch took: 0:02:29

======== Epoch 13 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.21 Training epoch took: 0:02:29

======== Epoch 14 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.20 Training epoch took: 0:02:29

======== Epoch 15 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.19 Training epoch took: 0:02:29

======== Epoch 16 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.18 Training epoch took: 0:02:29

======== Epoch 17 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.17 Training epoch took: 0:02:29

======== Epoch 18 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.17 Training epoch took: 0:02:29

======== Epoch 19 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.16 Training epoch took: 0:02:29

======== Epoch 20 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.15 Training epoch took: 0:02:29

======== Epoch 21 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.15 Training epoch took: 0:02:29

======== Epoch 22 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.15 Training epoch took: 0:02:29

======== Epoch 23 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.14 Training epoch took: 0:02:29

======== Epoch 24 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:45.

Average training loss: 0.14 Training epoch took: 0:02:29

======== Epoch 25 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.14 Training epoch took: 0:02:29

======== Epoch 26 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.13 Training epoch took: 0:02:29

======== Epoch 27 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.13 Training epoch took: 0:02:29

======== Epoch 28 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.13 Training epoch took: 0:02:29

======== Epoch 29 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.12 Training epoch took: 0:02:29

======== Epoch 30 / 30 ======== Training... Batch 50 of 143. Elapsed: 0:00:52. Batch 100 of 143. Elapsed: 0:01:44.

Average training loss: 0.13 Training epoch took: 0:02:29

precision recall f1-score support

       0       0.78      0.82      0.80       551
       1       0.82      0.79      0.81       590

accuracy                           0.80      1141

macro avg 0.80 0.80 0.80 1141 weighted avg 0.80 0.80 0.80 1141

Framework versions

  • Transformers 4.17.0
  • Pytorch 1.10.0+cu111
  • Tokenizers 0.11.6