016-microsoft-MiniLM-finetuned-yahoo-80_20
This model is a fine-tuned version of microsoft/MiniLM-L12-H384-uncased on the None dataset. It achieves the following results on the evaluation set:
- Loss: 1.6861
- F1: 0.4657
- Accuracy: 0.5
- Precision: 0.5267
- Recall: 0.5
- System Ram Used: 3.8760
- System Ram Total: 83.4807
- Gpu Ram Allocated: 0.3991
- Gpu Ram Cached: 1.9316
- Gpu Ram Total: 39.5640
- Gpu Utilization: 35
- Disk Space Used: 24.5397
- Disk Space Total: 78.1898
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: 32
- eval_batch_size: 32
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- num_epochs: 100
Training results
| Training Loss | Epoch | Step | Validation Loss | F1 | Accuracy | Precision | Recall | System Ram Used | System Ram Total | Gpu Ram Allocated | Gpu Ram Cached | Gpu Ram Total | Gpu Utilization | Disk Space Used | Disk Space Total |
|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|---|
| 2.3016 | 5.0 | 15 | 2.3016 | 0.0182 | 0.1 | 0.01 | 0.1 | 3.8589 | 83.4807 | 0.3990 | 1.9219 | 39.5640 | 38 | 24.5396 | 78.1898 |
| 2.2944 | 10.0 | 30 | 2.2979 | 0.0182 | 0.1 | 0.01 | 0.1 | 3.8753 | 83.4807 | 0.3991 | 1.9219 | 39.5640 | 36 | 24.5396 | 78.1898 |
| 2.2693 | 15.0 | 45 | 2.2696 | 0.2030 | 0.25 | 0.2472 | 0.25 | 3.8814 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 35 | 24.5396 | 78.1898 |
| 2.1627 | 20.0 | 60 | 2.2004 | 0.1808 | 0.25 | 0.1932 | 0.25 | 3.8785 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 39 | 24.5396 | 78.1898 |
| 1.9951 | 25.0 | 75 | 2.0773 | 0.2649 | 0.35 | 0.2922 | 0.35 | 3.8796 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 38 | 24.5396 | 78.1898 |
| 1.8128 | 30.0 | 90 | 1.9729 | 0.3619 | 0.45 | 0.3533 | 0.45 | 3.8802 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 36 | 24.5396 | 78.1898 |
| 1.6805 | 35.0 | 105 | 1.9061 | 0.4405 | 0.5 | 0.465 | 0.5 | 3.8803 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 37 | 24.5396 | 78.1898 |
| 1.5773 | 40.0 | 120 | 1.8512 | 0.3824 | 0.45 | 0.3767 | 0.45 | 3.8846 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 38 | 24.5396 | 78.1898 |
| 1.4916 | 45.0 | 135 | 1.8222 | 0.5190 | 0.55 | 0.5600 | 0.55 | 3.8846 | 83.4807 | 0.3991 | 1.9316 | 39.5640 | 40 | 24.5397 | 78.1898 |
| 1.4142 | 50.0 | 150 | 1.8056 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8850 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 38 | 24.5397 | 78.1898 |
| 1.3555 | 55.0 | 165 | 1.7700 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8850 | 83.4807 | 0.3991 | 1.9316 | 39.5640 | 41 | 24.5397 | 78.1898 |
| 1.3029 | 60.0 | 180 | 1.7568 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8795 | 83.4807 | 0.3991 | 1.9316 | 39.5640 | 35 | 24.5397 | 78.1898 |
| 1.2572 | 65.0 | 195 | 1.7462 | 0.4371 | 0.45 | 0.5067 | 0.45 | 3.8802 | 83.4807 | 0.3991 | 1.9316 | 39.5640 | 40 | 24.5397 | 78.1898 |
| 1.2207 | 70.0 | 210 | 1.7215 | 0.4371 | 0.45 | 0.5067 | 0.45 | 3.8880 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 37 | 24.5397 | 78.1898 |
| 1.1915 | 75.0 | 225 | 1.7103 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8760 | 83.4807 | 0.3991 | 1.9316 | 39.5640 | 39 | 24.5397 | 78.1898 |
| 1.1649 | 80.0 | 240 | 1.7069 | 0.4371 | 0.45 | 0.5067 | 0.45 | 3.8761 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 40 | 24.5397 | 78.1898 |
| 1.1484 | 85.0 | 255 | 1.6911 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8747 | 83.4807 | 0.3991 | 1.9316 | 39.5640 | 35 | 24.5397 | 78.1898 |
| 1.135 | 90.0 | 270 | 1.6888 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8753 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 37 | 24.5397 | 78.1898 |
| 1.1226 | 95.0 | 285 | 1.6860 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8755 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 39 | 24.5397 | 78.1898 |
| 1.1217 | 100.0 | 300 | 1.6861 | 0.4657 | 0.5 | 0.5267 | 0.5 | 3.8755 | 83.4807 | 0.3990 | 1.9316 | 39.5640 | 38 | 24.5397 | 78.1898 |
Framework versions
- Transformers 4.31.0
- Pytorch 2.0.1+cu118
- Datasets 2.13.1
- Tokenizers 0.13.3