File size: 3,794 Bytes
964840c |
1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 87 88 89 90 |
---
license: mit
tags:
- generated_from_trainer
metrics:
- precision
- recall
- f1
- accuracy
model-index:
- name: tmvar_2e-05_0404_ES6
results: []
---
<!-- 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. -->
# tmvar_2e-05_0404_ES6
This model is a fine-tuned version of [microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext](https://huggingface.co/microsoft/BiomedNLP-PubMedBERT-base-uncased-abstract-fulltext) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0115
- Precision: 0.8592
- Recall: 0.9289
- F1: 0.8927
- Accuracy: 0.9973
## 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: 16
- seed: 42
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
- lr_scheduler_type: linear
- training_steps: 2000
### Training results
| Training Loss | Epoch | Step | Validation Loss | Precision | Recall | F1 | Accuracy |
|:-------------:|:-----:|:----:|:---------------:|:---------:|:------:|:------:|:--------:|
| 0.5303 | 0.49 | 25 | 0.1134 | 0.0 | 0.0 | 0.0 | 0.9822 |
| 0.0792 | 0.98 | 50 | 0.0566 | 0.0 | 0.0 | 0.0 | 0.9822 |
| 0.0408 | 1.47 | 75 | 0.0472 | 0.2798 | 0.4772 | 0.3527 | 0.9853 |
| 0.0329 | 1.96 | 100 | 0.0298 | 0.468 | 0.5939 | 0.5235 | 0.9907 |
| 0.021 | 2.45 | 125 | 0.0242 | 0.4561 | 0.6853 | 0.5477 | 0.9906 |
| 0.0172 | 2.94 | 150 | 0.0184 | 0.6955 | 0.8579 | 0.7682 | 0.9948 |
| 0.0098 | 3.43 | 175 | 0.0133 | 0.7962 | 0.8528 | 0.8235 | 0.9962 |
| 0.0115 | 3.92 | 200 | 0.0117 | 0.8178 | 0.8883 | 0.8516 | 0.9968 |
| 0.0052 | 4.41 | 225 | 0.0121 | 0.8278 | 0.8782 | 0.8522 | 0.9968 |
| 0.0043 | 4.9 | 250 | 0.0112 | 0.8122 | 0.8782 | 0.8439 | 0.9966 |
| 0.0032 | 5.39 | 275 | 0.0108 | 0.8364 | 0.9340 | 0.8825 | 0.9970 |
| 0.0031 | 5.88 | 300 | 0.0117 | 0.8684 | 0.8376 | 0.8527 | 0.9968 |
| 0.0018 | 6.37 | 325 | 0.0103 | 0.8515 | 0.8731 | 0.8622 | 0.9971 |
| 0.0018 | 6.86 | 350 | 0.0095 | 0.8545 | 0.9239 | 0.8878 | 0.9976 |
| 0.0019 | 7.35 | 375 | 0.0097 | 0.8702 | 0.9188 | 0.8938 | 0.9976 |
| 0.0015 | 7.84 | 400 | 0.0117 | 0.8371 | 0.9391 | 0.8852 | 0.9968 |
| 0.0013 | 8.33 | 425 | 0.0117 | 0.8326 | 0.9086 | 0.8689 | 0.9971 |
| 0.0018 | 8.82 | 450 | 0.0098 | 0.8599 | 0.9036 | 0.8812 | 0.9973 |
| 0.0009 | 9.31 | 475 | 0.0089 | 0.8762 | 0.9340 | 0.9042 | 0.9977 |
| 0.0011 | 9.8 | 500 | 0.0105 | 0.8651 | 0.9442 | 0.9029 | 0.9975 |
| 0.0008 | 10.29 | 525 | 0.0098 | 0.875 | 0.9239 | 0.8988 | 0.9975 |
| 0.0008 | 10.78 | 550 | 0.0097 | 0.8685 | 0.9391 | 0.9024 | 0.9975 |
| 0.0009 | 11.27 | 575 | 0.0117 | 0.8780 | 0.9137 | 0.8955 | 0.9973 |
| 0.0007 | 11.76 | 600 | 0.0114 | 0.8538 | 0.9188 | 0.8851 | 0.9973 |
| 0.0007 | 12.25 | 625 | 0.0115 | 0.8592 | 0.9289 | 0.8927 | 0.9973 |
### Framework versions
- Transformers 4.27.4
- Pytorch 2.0.0+cu118
- Datasets 2.11.0
- Tokenizers 0.13.2
|