sofia-todeschini
commited on
Commit
•
b2ea3b8
1
Parent(s):
56d32dc
update model card README.md
Browse files
README.md
ADDED
@@ -0,0 +1,79 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
---
|
2 |
+
tags:
|
3 |
+
- generated_from_trainer
|
4 |
+
metrics:
|
5 |
+
- accuracy
|
6 |
+
model-index:
|
7 |
+
- name: BioElectra-LitCovid-1.4
|
8 |
+
results: []
|
9 |
+
---
|
10 |
+
|
11 |
+
<!-- This model card has been generated automatically according to the information the Trainer had access to. You
|
12 |
+
should probably proofread and complete it, then remove this comment. -->
|
13 |
+
|
14 |
+
# BioElectra-LitCovid-1.4
|
15 |
+
|
16 |
+
This model is a fine-tuned version of [kamalkraj/bioelectra-base-discriminator-pubmed](https://huggingface.co/kamalkraj/bioelectra-base-discriminator-pubmed) on an unknown dataset.
|
17 |
+
It achieves the following results on the evaluation set:
|
18 |
+
- Loss: 0.6551
|
19 |
+
- Hamming loss: 0.1096
|
20 |
+
- F1 micro: 0.5375
|
21 |
+
- F1 macro: 0.4017
|
22 |
+
- F1 weighted: 0.6519
|
23 |
+
- F1 samples: 0.5520
|
24 |
+
- Precision micro: 0.3867
|
25 |
+
- Precision macro: 0.2948
|
26 |
+
- Precision weighted: 0.5638
|
27 |
+
- Precision samples: 0.4347
|
28 |
+
- Recall micro: 0.8813
|
29 |
+
- Recall macro: 0.8425
|
30 |
+
- Recall weighted: 0.8813
|
31 |
+
- Recall samples: 0.8977
|
32 |
+
- Roc Auc: 0.8862
|
33 |
+
- Accuracy: 0.0375
|
34 |
+
|
35 |
+
## Model description
|
36 |
+
|
37 |
+
More information needed
|
38 |
+
|
39 |
+
## Intended uses & limitations
|
40 |
+
|
41 |
+
More information needed
|
42 |
+
|
43 |
+
## Training and evaluation data
|
44 |
+
|
45 |
+
More information needed
|
46 |
+
|
47 |
+
## Training procedure
|
48 |
+
|
49 |
+
### Training hyperparameters
|
50 |
+
|
51 |
+
The following hyperparameters were used during training:
|
52 |
+
- learning_rate: 2e-05
|
53 |
+
- train_batch_size: 16
|
54 |
+
- eval_batch_size: 16
|
55 |
+
- seed: 42
|
56 |
+
- gradient_accumulation_steps: 2
|
57 |
+
- total_train_batch_size: 32
|
58 |
+
- optimizer: Adam with betas=(0.9,0.999) and epsilon=1e-08
|
59 |
+
- lr_scheduler_type: linear
|
60 |
+
- num_epochs: 5
|
61 |
+
- mixed_precision_training: Native AMP
|
62 |
+
|
63 |
+
### Training results
|
64 |
+
|
65 |
+
| Training Loss | Epoch | Step | Validation Loss | Hamming loss | F1 micro | F1 macro | F1 weighted | F1 samples | Precision micro | Precision macro | Precision weighted | Precision samples | Recall micro | Recall macro | Recall weighted | Recall samples | Roc Auc | Accuracy |
|
66 |
+
|:-------------:|:-----:|:----:|:---------------:|:------------:|:--------:|:--------:|:-----------:|:----------:|:---------------:|:---------------:|:------------------:|:-----------------:|:------------:|:------------:|:---------------:|:--------------:|:-------:|:--------:|
|
67 |
+
| 0.8117 | 1.0 | 1151 | 0.7562 | 0.1732 | 0.4140 | 0.3137 | 0.5784 | 0.4179 | 0.2740 | 0.2255 | 0.4949 | 0.2947 | 0.8462 | 0.8285 | 0.8462 | 0.8675 | 0.8357 | 0.0005 |
|
68 |
+
| 0.639 | 2.0 | 2303 | 0.6690 | 0.1346 | 0.4836 | 0.3618 | 0.6199 | 0.4952 | 0.3347 | 0.2629 | 0.5289 | 0.3716 | 0.8714 | 0.8448 | 0.8714 | 0.8906 | 0.8682 | 0.0095 |
|
69 |
+
| 0.556 | 3.0 | 3454 | 0.6453 | 0.1253 | 0.5012 | 0.3747 | 0.6358 | 0.5147 | 0.3519 | 0.2750 | 0.5539 | 0.3944 | 0.8706 | 0.8536 | 0.8706 | 0.8895 | 0.8728 | 0.0220 |
|
70 |
+
| 0.4906 | 4.0 | 4606 | 0.6567 | 0.1111 | 0.5339 | 0.4013 | 0.6494 | 0.5469 | 0.3832 | 0.2946 | 0.5608 | 0.4282 | 0.8800 | 0.8428 | 0.8800 | 0.8976 | 0.8848 | 0.0312 |
|
71 |
+
| 0.4594 | 5.0 | 5755 | 0.6551 | 0.1096 | 0.5375 | 0.4017 | 0.6519 | 0.5520 | 0.3867 | 0.2948 | 0.5638 | 0.4347 | 0.8813 | 0.8425 | 0.8813 | 0.8977 | 0.8862 | 0.0375 |
|
72 |
+
|
73 |
+
|
74 |
+
### Framework versions
|
75 |
+
|
76 |
+
- Transformers 4.28.0
|
77 |
+
- Pytorch 2.0.0
|
78 |
+
- Datasets 2.1.0
|
79 |
+
- Tokenizers 0.13.3
|