Commit
•
1869669
1
Parent(s):
a9b12bb
Update README.md
Browse files
README.md
CHANGED
@@ -1,3 +1,24 @@
|
|
1 |
---
|
2 |
-
|
|
|
|
|
|
|
|
|
|
|
3 |
---
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
1 |
---
|
2 |
+
title: README
|
3 |
+
emoji: 🏃
|
4 |
+
colorFrom: gray
|
5 |
+
colorTo: purple
|
6 |
+
sdk: static
|
7 |
+
pinned: false
|
8 |
---
|
9 |
+
|
10 |
+
# Model Description
|
11 |
+
ClinicalMobileBERT is the result of training the [BioMobileBERT](https://huggingface.co/google/nlpie/bio-mobilebert) model in a continual learning scenario for 3 epochs using a total batch size of 192 on the MIMIC-III notes dataset.
|
12 |
+
|
13 |
+
# Initialisation
|
14 |
+
We initialise our model with the pre-trained checkpoints of the [BioMobileBERT](https://huggingface.co/google/nlpie/bio-mobilebert) model available on Huggingface.
|
15 |
+
|
16 |
+
# Architecture
|
17 |
+
MobileBERT uses a 128-dimensional embedding layer followed by 1D convolutions to up-project its output to the desired hidden dimension expected by the transformer blocks. For each of these blocks, MobileBERT uses linear down-projection at the beginning of the transformer block and up-projection at its end, followed by a residual connection originating from the input of the block before down-projection. Because of these linear projections, MobileBERT can reduce the hidden size and hence the computational cost of multi-head attention and feed-forward blocks. This model additionally incorporates up to four feed-forward blocks in order to enhance its representation learning capabilities. Thanks to the strategically placed linear projections, a 24-layer MobileBERT (which is used in this work) has around 25M parameters.
|
18 |
+
|
19 |
+
# Citation
|
20 |
+
If you use this model, please consider citing the following paper:
|
21 |
+
|
22 |
+
```bibtex
|
23 |
+
|
24 |
+
```
|