Dependency resolving to the latest torchmetrics

Nobody has claimed this yet.

## 🐛 Bug
When I install my project dependencies freshly, I get torchmetrics==1.0.3. According to https://lightning.ai/docs/torchmetrics/stable/, the latest stable version is 1.8.2. When I try to install any version higher than 1.8, it is not satisfiable.

### To Reproduce
```
uv add ty ruff pudb pdbp scalene textual-debugger rich numpy matplotlib pandas pandas-stubs numexpr bottleneck Jinja2 tabulate python-calamine xlsxwriter pyarrow backports.zstd pydicom simpleitk scikit-learn torch pytorch_optimizer monai einops kornia mlflow psutil pyrsmi nvidia-ml-py torchmetrics ray[default,data,train,tune,serve] debugpy bayesian-optimization traceml-ai --index https://download.pytorch.org/whl/rocm7.2
Using CPython 3.13.14 interpreter at: /home/ozer/micromamba/bin/python3
Creating virtual environment at: .venv
Resolved 202 packages in 2.10s
Prepared 22 packages in 1.35s
Installed 198 packages in 1.37s
+ aiohappyeyeballs==2.7.1
+ aiohttp==3.14.3
+ aiohttp-cors==0.8.1
+ aiosignal==1.4.0
+ alembic==1.18.5
+ annotated-doc==0.0.4
+ annotated-types==0.8.0
+ anyio==4.14.2
+ asttokens==3.0.2
+ attrs==26.1.0
+ backports-zstd==1.6.0
+ bayesian-optimization==3.3.0
+ blinker==1.9.0
+ bottleneck==1.6.0
+ cachetools==7.1.6
+ certifi==2022.12.7
+ cffi==2.1.0
+ charset-normalizer==2.1.1
+ click==8.4.2
+ cloudpickle==3.1.2
+ colorama==0.4.6
+ colorful==0.5.8
+ comm==0.2.3
+ contourpy==1.3.3
+ coverage==7.15.2
+ cryptography==48.0.1
+ cycler==0.12.1
+ databricks-sdk==0.122.0
+ debugpy==1.8.21
+ decorator==5.3.1
+ distlib==0.4.3
+ docker==7.2.0
+ einops==0.8.2
+ executing==2.2.1
+ fastapi==0.140.7
+ filelock==3.29.0
+ flask==3.1.3
+ flask-cors==6.0.5
+ fonttools==4.63.0
+ frozenlist==1.8.0
+ fsspec==2026.4.0
+ gitdb==4.0.12
+ gitpython==3.1.57
+ google-api-core==2.30.3
+ google-auth==2.56.2
+ googleapis-common-protos==1.75.0
+ graphene==3.4.3
+ graphql-core==3.2.11
+ graphql-relay==3.2.0
+ greenlet==3.5.4
+ grpcio==1.83.0
+ gunicorn==26.0.0
+ h11==0.16.0
+ httpcore==1.0.9
+ httptools==0.8.0
+ httpx==0.28.1
+ httpx-sse==0.4.3
+ huey==3.3.0
+ idna==3.4
+ importlib-metadata==7.1.0
+ iniconfig==2.3.0
+ ipython==9.15.0
+ ipython-pygments-lexers==1.1.1
+ ipywidgets==8.1.8
+ itsdangerous==2.2.0
+ jedi==0.20.0
+ jinja2==3.1.6
+ joblib==1.5.3
+ jsonschema==4.26.0
+ jsonschema-specifications==2025.9.1
+ jupyterlab-widgets==3.0.16
+ kiwisolver==1.5.0
+ kornia==0.8.3
+ kornia-rs==0.1.14
+ lightning-utilities==0.11.8
+ linkify-it-py==2.1.0
+ mako==1.3.12
+ markdown-it-py==4.2.0
+ markupsafe==3.0.3
+ matplotlib==3.11.1
+ matplotlib-inline==0.2.2
+ mcp==1.28.1
+ mdit-py-plugins==0.6.1
+ mdurl==0.1.2
+ mlflow==3.14.0
+ mlflow-skinny==3.14.0
+ mlflow-tracing==3.14.0
+ monai==1.6.0
+ mpmath==1.3.0
+ msgpack==1.2.1
+ multidict==6.7.1
+ narwhals==2.24.0
+ networkx==3.6.1
+ numexpr==2.14.2
+ numpy==2.4.4
+ nvidia-ml-py==13.610.43
+ opencensus==0.11.4
+ opencensus-context==0.1.3
+ opentelemetry-api==1.44.0
+ opentelemetry-exporter-prometheus==0.65b0
+ opentelemetry-proto==1.44.0
+ opentelemetry-sdk==1.44.0
+ opentelemetry-semantic-conventions==0.65b0
+ packaging==24.1
+ pandas==2.3.3
+ pandas-stubs==3.0.3.260530
+ parso==0.8.7
+ pdbp==1.8.2
+ pexpect==4.9.0
+ pillow==12.2.0
+ platformdirs==4.11.0
+ pluggy==1.6.0
+ prettytable==3.18.0
+ prometheus-client==0.26.0
+ prompt-toolkit==3.0.53
+ propcache==0.5.2
+ proto-plus==1.28.2
+ protobuf==6.33.6
+ psutil==7.2.2
+ ptyprocess==0.7.0
+ pudb==2025.1.5
+ pure-eval==0.2.3
+ py-spy==0.4.2
+ pyarrow==24.0.0
+ pyasn1==0.6.4
+ pyasn1-modules==0.4.2
+ pycparser==3.0
+ pydantic==2.13.4
+ pydantic-core==2.46.4
+ pydantic-settings==2.14.2
+ pydicom==3.0.2
+ pygments==2.20.0
+ pyjwt==2.13.0
+ pynvml==13.0.1
+ pyparsing==3.3.2
+ pyrsmi==1.1.0
+ pytest==9.1.1
+ pytest-cov==7.1.0
+ python-calamine==0.8.2
+ python-dateutil==2.9.0.post0
+ python-discovery==1.5.0
+ python-dotenv==1.2.2
+ python-multipart==0.0.32
+ pytorch-optimizer==3.10.1
+ pytz==2026.3.post1
+ pyyaml==6.0.3
+ ray==2.52.1
+ referencing==0.37.0
+ requests==2.28.1
+ rich==15.0.0
+ rpds-py==2026.6.3
+ ruff==0.16.0
+ scalene==2.3.0
+ scikit-learn==1.9.0
+ scipy==1.18.0
+ setuptools==78.1.0
+ simpleitk==2.5.5
+ six==1.17.0
+ skops==0.14.0
+ smart-open==8.0.1
+ smmap==5.0.3
+ sqlalchemy==2.0.51
+ sqlparse==0.5.5
+ sse-starlette==3.4.6
+ stack-data==0.6.3
+ starlette==1.3.1
+ sympy==1.14.0
+ tabcompleter==1.4.1
+ tabulate==0.10.0
+ tensorboardx==2.6.5
+ textual==8.2.8
+ textual-debugger==0.2.0
+ threadpoolctl==3.6.0
+ torch==2.13.0+rocm7.2
+ torchmetrics==1.0.3
+ traceml-ai==0.1.3
+ traitlets==5.15.1
+ triton-rocm==3.7.1
+ ty==0.0.65
+ typing-extensions==4.15.0
+ typing-inspection==0.4.2
+ tzdata==2026.3
+ uc-micro-py==2.0.0
+ urllib3==1.26.13
+ urwid==4.0.8
+ urwid-readline==0.15.1
+ uvicorn==0.51.0
+ uvloop==0.22.1
+ virtualenv==21.7.0
+ watchfiles==1.2.0
+ wcwidth==0.8.2
+ websockets==16.1.1
+ werkzeug==3.1.8
+ widgetsnbextension==4.0.15
+ wrapt==2.3.0
+ xlsxwriter==3.2.9
+ yarl==1.24.5
+ zipp==3.19.2
```

```
uv add 'torchmetrics>1.8'
× No solution found when resolving dependencies for split (markers: python_full_version >= '3.14'):
╰─▶ Because only torchmetrics==1.0.3 is available and your project depends on torchmetrics>1.8, we can conclude that your project's requirements are unsatisfiable.

hint: While the active Python version is 3.13, the resolution failed for other Python versions supported by your project. Consider limiting your project's supported Python versions using `requires-python`.
hint: `torchmetrics` was found on https://download.pytorch.org/whl/rocm7.2, but not at the requested version (torchmetrics>1.8). A compatible version may be available on a subsequent index (e.g., https://pypi.org/simple). By default, uv will only consider versions that are published on the first index that contains a given package, to avoid dependency confusion attacks. If all indexes are equally trusted, use `--index-strategy unsafe-best-match` to consider all versions from all indexes, regardless of the order in which they were defined.
hint: If you want to add the package regardless of the failed resolution, provide the `--frozen` flag to skip locking and syncing
```

Start by rerunning the provided uv add command with the ROCm index and compare the torchmetrics versions visible on that index and PyPI. Inspect the project's dependency and publishing configuration; done means fresh resolution can select a current torchmetrics release without breaking the supported Python-version constraints.

Written by the indexing model from the issue text.