Commit
•
0a27449
1
Parent(s):
44a5402
Create README.md
Browse files
README.md
ADDED
@@ -0,0 +1,25 @@
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
+
# Model Description
|
2 |
+
BioDistilBERT-cased is the result of training the [DistilBERT-cased](https://huggingface.co/distilbert-base-cased?text=The+goal+of+life+is+%5BMASK%5D.) model in a continual learning fashion for 200k training steps using a total batch size of 192 on the PubMed dataset.
|
3 |
+
|
4 |
+
|
5 |
+
# Initialisation
|
6 |
+
We initialise our model with the pre-trained checkpoints of the [DistilBERT-cased](https://huggingface.co/distilbert-base-cased?text=The+goal+of+life+is+%5BMASK%5D.) model available on the Huggingface.
|
7 |
+
|
8 |
+
# Architecture
|
9 |
+
In this model, the size of the hidden dimension and the embedding layer are both set to 768. The vocabulary size is 28996. The number of transformer layers is 6 and the expansion rate of the feed-forward layer is 4. Overall this model has around 65 million parameters.
|
10 |
+
|
11 |
+
# Citation
|
12 |
+
If you use this model, please consider citing the following paper:
|
13 |
+
|
14 |
+
```bibtex
|
15 |
+
@misc{https://doi.org/10.48550/arxiv.2209.03182,
|
16 |
+
doi = {10.48550/ARXIV.2209.03182},
|
17 |
+
url = {https://arxiv.org/abs/2209.03182},
|
18 |
+
author = {Rohanian, Omid and Nouriborji, Mohammadmahdi and Kouchaki, Samaneh and Clifton, David A.},
|
19 |
+
keywords = {Computation and Language (cs.CL), Machine Learning (cs.LG), FOS: Computer and information sciences, FOS: Computer and information sciences, 68T50},
|
20 |
+
title = {On the Effectiveness of Compact Biomedical Transformers},
|
21 |
+
publisher = {arXiv},
|
22 |
+
year = {2022},
|
23 |
+
copyright = {arXiv.org perpetual, non-exclusive license}
|
24 |
+
}
|
25 |
+
```
|