<!-- 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. -->
mpnet-multilabel-sector-classifier
This model is a fine-tuned version of sentence-transformers/all-mpnet-base-v2 on the None dataset. It achieves the following results on the evaluation set:
- Loss: 0.2273
- Precision Micro: 0.8075
- Precision Weighted: 0.8110
- Precision Samples: 0.8365
- Recall Micro: 0.8897
- Recall Weighted: 0.8897
- Recall Samples: 0.8922
- F1-score: 0.8464
Model description
This model is trained for performing Multi Label Sector Classification.
Training hyperparameters
The following hyperparameters were used during training:
- learning_rate: 6.9e-05
- train_batch_size: 16
- eval_batch_size: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 200
- num_epochs: 8
- weight_decay: 0.001
- gradient_acumulation_steps: 1
Training results
| Training Loss | Epoch | Step | Validation Loss | Precision Micro | Precision Weighted | Precision Samples | Recall Micro | Recall Weighted | Recall Samples | F1-score |
|---|---|---|---|---|---|---|---|---|---|---|
| 0.4478 | 1.0 | 897 | 0.2277 | 0.6731 | 0.7183 | 0.7460 | 0.8822 | 0.8822 | 0.8989 | 0.7871 |
| 0.2241 | 2.0 | 1794 | 0.1862 | 0.7088 | 0.7485 | 0.7754 | 0.8933 | 0.8933 | 0.9110 | 0.8108 |
| 0.1647 | 3.0 | 2691 | 0.2025 | 0.6785 | 0.7023 | 0.7634 | 0.9124 | 0.9124 | 0.9252 | 0.8077 |
| 0.1232 | 4.0 | 3588 | 0.1839 | 0.7274 | 0.7322 | 0.7976 | 0.9029 | 0.9029 | 0.9134 | 0.8286 |
| 0.0899 | 5.0 | 4485 | 0.1889 | 0.7919 | 0.8007 | 0.8350 | 0.8909 | 0.8909 | 0.9060 | 0.8483 |
| 0.0653 | 6.0 | 5382 | 0.2039 | 0.7478 | 0.7544 | 0.8098 | 0.8973 | 0.8973 | 0.9114 | 0.8346 |
| 0.0462 | 7.0 | 6279 | 0.2149 | 0.7447 | 0.7500 | 0.8060 | 0.8989 | 0.8989 | 0.9107 | 0.8323 |
| 0.0336 | 8.0 | 7176 | 0.2181 | 0.7733 | 0.7780 | 0.8221 | 0.8909 | 0.8909 | 0.9031 | 0.8400 |
Environmental Impact
Carbon emissions were estimated using the codecarbon. The carbon emission reported are incluidng the hyperparamter search performed on subset of training data.
- Hardware Type: 16GB T4
- Hours used: 3
- Cloud Provider: Google Colab
- Carbon Emitted : 0.276132
Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
