File size: 1,972 Bytes
8fafaae | 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 | ---
tags:
- generated_from_trainer
metrics:
- accuracy
- f1
- precision
- recall
model-index:
- name: scibert_claim_id_3e-05
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. -->
# scibert_claim_id_3e-05
This model is a fine-tuned version of [allenai/scibert_scivocab_uncased](https://huggingface.co/allenai/scibert_scivocab_uncased) on the None dataset.
It achieves the following results on the evaluation set:
- Loss: 0.0071
- Accuracy: 0.9980
- F1: 0.9935
- Precision: 0.9957
- Recall: 0.9914
## 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: 3e-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: 6
### Training results
| Training Loss | Epoch | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|:-------------:|:-----:|:----:|:---------------:|:--------:|:------:|:---------:|:------:|
| 0.3163 | 1.0 | 666 | 0.2554 | 0.8884 | 0.5534 | 0.7437 | 0.4407 |
| 0.2673 | 2.0 | 1332 | 0.1671 | 0.9361 | 0.7850 | 0.8309 | 0.7439 |
| 0.2188 | 3.0 | 1998 | 0.0689 | 0.9769 | 0.9268 | 0.9232 | 0.9303 |
| 0.0925 | 4.0 | 2664 | 0.0369 | 0.9879 | 0.9624 | 0.9428 | 0.9827 |
| 0.0635 | 5.0 | 3330 | 0.0109 | 0.9971 | 0.9909 | 0.9928 | 0.9889 |
| 0.038 | 6.0 | 3996 | 0.0071 | 0.9980 | 0.9935 | 0.9957 | 0.9914 |
### Framework versions
- Transformers 4.28.0
- Pytorch 2.0.1+cu118
- Datasets 2.12.0
- Tokenizers 0.13.3
|