Update README.md
Browse files
README.md
CHANGED
@@ -16,6 +16,16 @@ tags:
|
|
16 |
This is a model card for detecting claims from an abstract of social science publications.
|
17 |
The model takes an abstract, performs sentence tokenization, and predict a claim probability of each sentence.
|
18 |
This model card is released by training on a [SCORE](https://www.cos.io/score) dataset.
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
19 |
|
20 |
```py
|
21 |
import spacy
|
@@ -54,4 +64,30 @@ def inference(abstract: str):
|
|
54 |
claims = inference(abstract) # string of claim joining with \n
|
55 |
```
|
56 |
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
57 |
See more on `gradio` application in `biodatlab` space.
|
|
|
16 |
This is a model card for detecting claims from an abstract of social science publications.
|
17 |
The model takes an abstract, performs sentence tokenization, and predict a claim probability of each sentence.
|
18 |
This model card is released by training on a [SCORE](https://www.cos.io/score) dataset.
|
19 |
+
It achieves the following results on the test set:
|
20 |
+
|
21 |
+
- Accuracy: 0.931597
|
22 |
+
- Precision: 0.764563
|
23 |
+
- Recall: 0.722477
|
24 |
+
- F1: 0.742925
|
25 |
+
|
26 |
+
|
27 |
+
## Model Usage
|
28 |
+
You can access the model with huggingface's `transformers` as follows:
|
29 |
|
30 |
```py
|
31 |
import spacy
|
|
|
64 |
claims = inference(abstract) # string of claim joining with \n
|
65 |
```
|
66 |
|
67 |
+
## Training procedure
|
68 |
+
|
69 |
+
### Training Hyperparameters
|
70 |
+
|
71 |
+
The following hyperparameters were used during training:
|
72 |
+
|
73 |
+
learning_rate: 3e-05
|
74 |
+
train_batch_size: 32
|
75 |
+
eval_batch_size: 32
|
76 |
+
n_epochs: 6
|
77 |
+
|
78 |
+
### Training results
|
79 |
+
|
80 |
+
0.038000 0.007086 0.997964 0.993499 0.995656 0.991350
|
81 |
+
|
82 |
+
| Training Loss | Step | Validation Loss | Accuracy | F1 | Precision | Recall |
|
83 |
+
|:-------------:|:----:|:---------------:|:--------:|:--------:|:---------:|:--------:|
|
84 |
+
| 0.038000 | 3996 | 0.007086 | 0.997964 | 0.993499 | 0.995656 | 0.991350 |
|
85 |
+
|
86 |
+
### Framework versions
|
87 |
+
- transformers 4.28.0
|
88 |
+
- sentence-transformers 2.2.2
|
89 |
+
- accelerate 0.19.0
|
90 |
+
- datasets 2.12.0
|
91 |
+
- spacy 3.5.3
|
92 |
+
|
93 |
See more on `gradio` application in `biodatlab` space.
|