tests: add pytest testmon plugin to speed up CI (#1402)

* add pytest testmon plugin

* test changes to check testmon

* allow testmon and markers

* further improvements for uv and testmon

* add test comment

* fix caching, test changes in mcmc posterior

* fix uv setup

* fix uv setup

* fix uv setup

* fix uv setup

* another fix

* remove pip from ci

* undo changes to test testmon

* load testmon cache first

* restore old testmon setup

* reorder jobs

* test comment

* back to uv pip

* test change

* seems to be working 🎉

.github/workflows/ci.yml

@@ -40,9 +40,24 @@ jobs:
         uv pip install torch --extra-index-url https://download.pytorch.org/whl/cpu
         uv pip install -e .[dev]
 
+    - name: Cache uv dependencies
+      uses: actions/cache@v4
+      with:
+        path: ~/.cache/uv
+        key: uv-${{ runner.os }}-${{ matrix.python-version }}-${{ hashFiles('pyproject.toml') }}
+        restore-keys: |
+          uv-${{ runner.os }}-${{ matrix.python-version }}-
+
+    - name: Cache testmon data
+      uses: actions/cache@v4
+      with:
+        path: .testmondata
+        key: testmon-${{ runner.os }}-${{ matrix.python-version }}-${{ github.ref }}
+        restore-keys: |
+          testmon-${{ runner.os }}-${{ matrix.python-version }}-
+
     - name: Run the fast CPU tests with coverage
-      run: |
-        uv run pytest -v -x -n auto -m "not slow and not gpu" --cov=sbi --cov-report=xml tests/
+      run: uv run pytest --testmon-forceselect -v -n auto -m "not slow and not gpu" --cov=sbi --cov-report=xml tests/
 
     - name: Upload coverage to Codecov
       uses: codecov/codecov-action@v4

pyproject.toml

@@ -71,6 +71,7 @@ dev = [
     # Test
     "pytest",
     "pytest-cov",
+    "pytest-testmon",
     "pytest-xdist",
     "torchtestcase",
 ]

sbi/neural_nets/net_builders/mnle.py

@@ -166,7 +166,6 @@ def build_mnle(
 
     # Set up a flow for modelling the continuous data, conditioned on the discrete data.
     continuous_net = model_builders[flow_model](
-        # TODO: add support for optional log-transform in flow builders.
         batch_x=(
             torch.log(cont_x) if log_transform_x else cont_x
         ),  # log transform manually.