<!-- 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. -->
Public100_1L_BERT_5epoch
This model is a fine-tuned version of Youssef320/LSTM-finetuned-50label-15epoch on the None dataset. It achieves the following results on the evaluation set:
- Loss: 3.5356
- Top 1 Macro F1 Score: 0.0657
- Top 1 Weighted F1score: 0.1228
- Top 3 Macro F1 Score: 0.1636
- Top3 3 Weighted F1 Score : 0.2687
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: 0.0002
- train_batch_size: 64
- eval_batch_size: 64
- seed: 42
- gradient_accumulation_steps: 32
- total_train_batch_size: 2048
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: constant
- num_epochs: 1.0
Training results
Training Loss | Epoch | Step | Validation Loss | Top 1 Macro F1 Score | Top 1 Weighted F1score | Top 3 Macro F1 Score | Top3 3 Weighted F1 Score |
---|---|---|---|---|---|---|---|
3.9544 | 0.12 | 64 | 3.9094 | 0.0158 | 0.0449 | 0.0790 | 0.1564 |
3.7996 | 0.25 | 128 | 3.7552 | 0.0335 | 0.0776 | 0.1111 | 0.2036 |
3.6874 | 0.38 | 192 | 3.6721 | 0.0431 | 0.0934 | 0.1324 | 0.2293 |
3.6587 | 0.5 | 256 | 3.6292 | 0.0464 | 0.0992 | 0.1375 | 0.2388 |
3.6341 | 0.62 | 320 | 3.5993 | 0.0517 | 0.1054 | 0.1425 | 0.2448 |
3.6444 | 0.75 | 384 | 3.5739 | 0.0582 | 0.1144 | 0.1527 | 0.2555 |
3.613 | 0.88 | 448 | 3.5565 | 0.0623 | 0.1182 | 0.1567 | 0.2606 |
3.5787 | 1.0 | 512 | 3.5356 | 0.0657 | 0.1228 | 0.1636 | 0.2687 |
Framework versions
- Transformers 4.20.1
- Pytorch 1.12.1+cu102
- Datasets 2.0.0
- Tokenizers 0.11.0