Batch the torch Cell List

.github/workflows/test.yml

@@ -159,7 +159,7 @@ jobs:
           if [[ "${{ matrix.example }}" == *"fairchem"* ]]; then
             uv pip install huggingface_hub --system
           fi
-          uv run --with . ${{ matrix.example }}
+          uv run --with-editable . ${{ matrix.example }}
 
   all-required-pass:
     if: always()

README.md

@@ -157,10 +157,10 @@ If you use TorchSim in your research, please cite our [publication](https://iops
 
 ## Due Credit
 
-We aim to recognize all [duecredit](https://github.com/duecredit/duecredit) for the decades of work that TorchSim builds on top of, an automated list of references can be obtained for the package by running `DUECREDIT_ENABLE=yes uv run --with . --extra docs --extra test python -m duecredit <(printf 'import pytest\nraise SystemExit(pytest.main(["-q"]))\n')`. This list is incomplete and we welcome PRs to help improve our citation coverage.
+We aim to recognize all [duecredit](https://github.com/duecredit/duecredit) for the decades of work that TorchSim builds on top of, an automated list of references can be obtained for the package by running `DUECREDIT_ENABLE=yes uv run --with-editable . --extra docs --extra test python -m duecredit <(printf 'import pytest\nraise SystemExit(pytest.main(["-q"]))\n')`. This list is incomplete and we welcome PRs to help improve our citation coverage.
 
 To collect citations for a specific tutorial run, for example autobatching, use:
 
 ```sh
-DUECREDIT_ENABLE=yes uv run --with . --extra docs --extra test python -m duecredit examples/tutorials/autobatching_tutorial.py
+DUECREDIT_ENABLE=yes uv run --with-editable . --extra docs --extra test python -m duecredit examples/tutorials/autobatching_tutorial.py
 ```

examples/benchmarking/neighborlists.py

@@ -14,7 +14,7 @@
 Directly times each torch-sim NL backend without any model evaluation.
 
 Example:
-    uv run --with . examples/benchmarking/neighborlists.py \
+    uv run --with-editable . examples/benchmarking/neighborlists.py \
         --source wbm --n-structures 100 --device cpu
 """
 

examples/readme.md

@@ -35,25 +35,25 @@ If you'd like to execute the scripts or examples locally, you can run them with:
 curl -LsSf https://astral.sh/uv/install.sh | sh
 
 # pick any of the examples
-uv run --with . examples/scripts/1_introduction.py
-uv run --with . examples/scripts/2_structural_optimization.py
-uv run --with . examples/scripts/3_dynamics.py
-uv run --with . examples/scripts/4_high_level_api.py
+uv run --with-editable . examples/scripts/1_introduction.py
+uv run --with-editable . examples/scripts/2_structural_optimization.py
+uv run --with-editable . examples/scripts/3_dynamics.py
+uv run --with-editable . examples/scripts/4_high_level_api.py
 
 # or any of the tutorials
-uv run --with . examples/tutorials/diff_sim.py
+uv run --with-editable . examples/tutorials/diff_sim.py
 ```
 
 ## Benchmarking Scripts
 
 The `examples/benchmarking/` folder contains standalone benchmark scripts. They
 declare their own dependencies via [PEP 723 inline script metadata](https://peps.python.org/pep-0723/)
-and should be run with `uv run --with .` so that the local `torch-sim` package
+and should be run with `uv run --with-editable .` so that the local `torch-sim` package
 is available alongside the script's isolated dependency environment:
 
 ```sh
 # Neighbor-list backend benchmark on WBM or MP structures
-uv run --with . examples/benchmarking/neighborlists.py \
+uv run --with-editable . examples/benchmarking/neighborlists.py \
     --source wbm --n-structures 100 --device cpu
 
 # Scaling benchmark: static, relax, NVE, NVT