Set up methods for anonymous usage analytics

[safaricd]

Motivation and Context

This PR introduces fully anonymous usage analytics to help us understand how the library is being used and guide future improvements.

Telemetry is GDPR-compliant and designed with data minimization in mind — no personal data, code, inputs, or outputs are ever collected.

By default, analytics are enabled in an opt-in, transparent manner, and users can easily disable them at any time via the TABPFN_DISABLE_TELEMETRY=1 environment variable.

Public API Changes

• No Public API changes

---

Checklist

• The changes have been tested locally.
• Documentation has been updated (if the public API or usage changes).
• A entry has been added to CHANGELOG.md (if relevant for users).
• The code follows the project's style guidelines.
• I have considered the impact of these changes on the public API.

---

[CLAassistant]

All committers have signed the CLA.

[gemini-code-assist]

Code Review

This pull request introduces anonymous usage analytics by adding the tabpfn-common-utils dependency and instrumenting the TabPFNClassifier and TabPFNRegressor classes. The changes are well-contained and consistently applied across the codebase. The README has also been updated to inform users about this new feature and provide instructions for opting out. My review includes a couple of minor suggestions to improve the documentation and project metadata.

[noahho]

When installing I get an error:

warning: The package `tabpfn-common-utils==0.1.7` does not have an extra named `telemetry-interactive`

[safaricd]

When installing I get an error:

warning: The package `tabpfn-common-utils==0.1.7` does not have an extra named `telemetry-interactive`

@noahho fix has been applied and pushed, thanks for reporting. It was due to the generate_dependencies.py script we use for parsing and generating requirements.txt from pyproject.toml.

[oscarkey]

LGTM!

[brendan-priorlabs]

LGTM. Thanks, Dominik