Closed RocStone closed 3 years ago
The export format actually relies the extension in the filename specified in save dialog (e.g. slide.html
, slide.pdf
) and the content of slide deck does not matter. If not detected the extension from filename, Marp will try to export into HTML with regardless of setting.
Try to set a extension .pdf
or .pptx
explicitly.
Normally a corresponded extension with selected type will add automatically by VS Code, but some Linux distributions have reported won't add. Potentially related: https://github.com/marp-team/marp-vscode/issues/83
The export format actually relies the extension in the filename specified in save dialog (e.g.
slide.html
,slide.pdf
) and the content of slide deck does not matter. If not detected the extension from filename, Marp will try to export into HTML with regardless of setting.Try to set a extension
.pptx
explicitly.Normally a corresponded extension with selected type will add automatically by VS Code, but some Linux distributions have reported won't add. Potentially related: #83
Thanks for the response, but it still doesn't work even if I explicitly set the filename as xxx.pdf. FYI, I am using Ubuntu 20.04 with KDE theme.
How about quoted filename such as
"foobar.pptx"
?-- https://github.com/marp-team/marp-vscode/issues/83#issuecomment-541426564
How about quoted filename such as
"foobar.pptx"
? -- #83 (comment)
I tried "asd.pdf" and "asd.pptx", neither works.
We cannot reproduce that in Kubuntu 20.04. PDF export is working by setting .pdf
extension explicitly.
https://user-images.githubusercontent.com/3993388/121181000-7f3ba200-c89c-11eb-8f60-c8a04647b7eb.mp4
A weird thing happened, I rebooted my PC and now it works fine. Thanks for your help!
To reproduce the bug here is my document. And it is always exported in HTML (but the HTML works perfectly.)
Thanks for your awesome work btw, I really really like this extension! Good Job!
My Document____
marp: true theme: gaia footer: 'Pengqian Lu, 2021.06.07' paginate: true style: | section a { font-size: 30px; }
GMNN: Graph Markov Neural Networks
Problem to be solved
Semi-supervised object classification in relational data.
Relational data: Here, it's graph data.
Related Field: SRL
Statistical relational learning (SRL) develops statistical methods to model relational data. Generally theses methods model the dependency of object labels using conditional random fields (CRF).
Brief introduction of CRF
Given a graph $G=(V,E,X,Y)$, let $V$ denotes a set of nodes, $E$ denotes a set of edges, $X$ denotes the feature matrix on nodes and $Y$ denotes the labels of nodes.
Brief introduction of CRF
CRF models the conditional probability of $Y$ given $X$ as:
$p(Y|X)=\frac{1}{Z(X)}\prod_{(n_i,nj)\in E}\psi{i,j}(y_i,y_j,X)$
Here, $Z=\sumY\prod{(n_i,nj)\in E}\psi{i,j}(y_i,y_j,X)$
You can define your potential functions $\psi_{i,j}$. For example, in segmentation,
$\psi_{ij}=|y_i-y_j|(-\exp{\frac{||z_i-z_j||}{2Mean(||z_i-z_j||_2)}})$.
Brief introduction of CRF
The limitations of SRL methods:
Brief introduction of GNN
Learning object representations with non-linear neural architectures.
The limitations of such methods: