all-mpnet-base-v2-sentiment-twitter
This model is a fine-tuned version of sentence-transformers/all-mpnet-base-v2 on an unknown dataset. It achieves the following results on the evaluation set:
- Loss: 0.6390
- Accuracy: 0.7169
- F1: 0.7156
- Precision: 0.7224
- Recall: 0.7169
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: 16
- eval_batch_size: 32
- seed: 42
- optimizer: Use OptimizerNames.ADAMW_TORCH_FUSED with betas=(0.9,0.999) and epsilon=1e-08 and optimizer_args=No additional optimizer arguments
- lr_scheduler_type: linear
- lr_scheduler_warmup_steps: 500
- num_epochs: 3
- mixed_precision_training: Native AMP
Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|---|---|---|---|---|---|---|---|
| 0.714 | 0.1754 | 500 | 0.7064 | 0.699 | 0.6941 | 0.7038 | 0.699 |
| 0.6676 | 0.3508 | 1000 | 0.6255 | 0.7225 | 0.7260 | 0.7353 | 0.7225 |
| 0.6394 | 0.5261 | 1500 | 0.6229 | 0.7265 | 0.7287 | 0.7371 | 0.7265 |
| 0.6408 | 0.7015 | 2000 | 0.6101 | 0.73 | 0.7250 | 0.7330 | 0.73 |
| 0.6069 | 0.8769 | 2500 | 0.5878 | 0.749 | 0.7500 | 0.7516 | 0.749 |
| 0.4633 | 1.0523 | 3000 | 0.6304 | 0.7485 | 0.7482 | 0.7548 | 0.7485 |
| 0.5003 | 1.2276 | 3500 | 0.6054 | 0.747 | 0.7486 | 0.7518 | 0.747 |
| 0.4594 | 1.4030 | 4000 | 0.6955 | 0.7205 | 0.7146 | 0.7315 | 0.7205 |
| 0.4721 | 1.5784 | 4500 | 0.5965 | 0.7555 | 0.7553 | 0.7565 | 0.7555 |
| 0.4601 | 1.7538 | 5000 | 0.6504 | 0.7385 | 0.7389 | 0.7436 | 0.7385 |
| 0.4762 | 1.9291 | 5500 | 0.5998 | 0.757 | 0.7564 | 0.7573 | 0.757 |
| 0.4 | 2.1045 | 6000 | 0.6574 | 0.7535 | 0.7533 | 0.7538 | 0.7535 |
| 0.4021 | 2.2799 | 6500 | 0.6714 | 0.75 | 0.7498 | 0.7505 | 0.75 |
| 0.3477 | 2.4553 | 7000 | 0.6803 | 0.7495 | 0.7495 | 0.7496 | 0.7495 |
| 0.3616 | 2.6307 | 7500 | 0.6996 | 0.745 | 0.7448 | 0.7459 | 0.745 |
| 0.3692 | 2.8060 | 8000 | 0.6872 | 0.748 | 0.7484 | 0.7502 | 0.748 |
| 0.3549 | 2.9814 | 8500 | 0.6903 | 0.749 | 0.7487 | 0.7499 | 0.749 |
Framework versions
- Transformers 4.55.4
- Pytorch 2.8.0+cu126
- Datasets 4.0.0
- Tokenizers 0.21.4