Use readme as docs index

README.md

@@ -1,6 +1,25 @@
-# tinytopics
+# tinytopics <img src="assets/logo.png" align="right" width="120" />
 
-Fit topic models using neural Poisson non-negative matrix factorization with sum-to-one constraints.
+Topic modeling via sum-to-one constrained Poisson non-negative
+matrix factorization (NMF), built on PyTorch and runs on GPU.
+
+## Installation
+
+You can install tinytopics from PyPI:
+
+```bash
+pip install tinytopics
+```
+
+Or install the development version from GitHub:
+
+```bash
+git clone https://github.com/nanxstats/tinytopics.git
+cd tinytopics
+python3 -m pip install -e .
+```
+
+Try [getting started](articles/get-started.md).
 
 ## Known issues
 

docs/index.md

@@ -3,6 +3,8 @@
 Topic modeling via sum-to-one constrained Poisson non-negative
 matrix factorization (NMF), built on PyTorch and runs on GPU.
 
+## Installation
+
 You can install tinytopics from PyPI:
 
 ```bash
@@ -18,3 +20,7 @@ python3 -m pip install -e .
 ```
 
 Try [getting started](articles/get-started.md).
+
+## Known issues
+
+- [ ] Running on CPU produces different (and worse) models than on GPU.

docs/scripts/purl.sh → docs/scripts/sync.sh

@@ -1,5 +1,9 @@
 #!/bin/bash
 
+# Sync README.md
+cp README.md docs/index.md
+
+# Sync articles
 quarto render docs/articles/get-started.qmd
 quarto convert docs/articles/get-started.qmd
 jupyter nbconvert --to python docs/articles/get-started.ipynb --output ../../examples/get-started.py