Update README and add CODECOV_TOKEN env to workflow

.github/workflows/main.yml

@@ -39,6 +39,8 @@ jobs:
       - name: Upload coverage report
         if: runner.os == 'Linux' && matrix.python-version == '3.12'
         uses: codecov/codecov-action@v4
+        env:
+          CODECOV_TOKEN: ${{ secrets.CODECOV_TOKEN }}
   # run-mypy:
   #   name: Run mypy on Python 3.12
   #   runs-on: ubuntu-latest

README.md

@@ -8,7 +8,13 @@ estimators were proposed. None of them is implemented in standard econometrics p
 
 ## Installation
 
-The package can be installed via PyPI or via GitHub. To do so, type the following in a
+> **Warning:** To run skillmodels you need to install jax and jaxlib. If you aim to run
+> skillmodels on a GPU, you need to install jaxlib with CUDA support (for Linux) or with
+> support for AMD GPUs / ARM-based Silicon GPUs (for MacOS). In any case, for
+> installation of jax and jaxlib, please consult the jax
+> [docs](https://jax.readthedocs.io/en/latest/installation.html#supported-platforms).
+
+Skillmodels can be installed via PyPI or via GitHub. To do so, type the following in a
 terminal:
 
 ```console