Conversation
There was a problem hiding this comment.
Pull request overview
This PR introduces pixi-based installation as the recommended way to install segment-geospatial, with detailed environment examples (GPU/CPU) and SAM 3 setup guidance to avoid dependency conflicts (especially around NumPy and PyTorch/CUDA).
Changes:
- Added a step-by-step pixi installation section to
docs/installation.md, including CUDA 12.x/13.x GPU and CPUpixi.tomltemplates, environment verification, and SAM 3 + Hugging Face setup instructions. - Updated
README.mdto recommend pixi as the preferred installation method and to provide a concise quick-start snippet that links to the full installation guide. - Retained existing PyPI/conda/Docker installation sections, positioning pixi as the most reliable path, especially on Windows.
Reviewed changes
Copilot reviewed 2 out of 2 changed files in this pull request and generated 4 comments.
| File | Description |
|---|---|
| docs/installation.md | Adds a comprehensive pixi-based installation guide with GPU/CPU environment configurations, PyTorch/CUDA verification, and SAM 3 + Hugging Face authentication instructions. |
| README.md | Introduces a new “Install with pixi (Recommended)” section that summarizes the pixi workflow and links to the detailed docs. |
💡 Add Copilot custom instructions for smarter, more guided reviews. Learn how to get started.
| #### For CPU: | ||
|
|
||
| ```toml | ||
| [workspace] | ||
| channels = ["https://prefix.dev/conda-forge"] | ||
| name = "geo" | ||
| platforms = ["linux-64", "win-64"] | ||
|
|
||
| [dependencies] | ||
| python = "3.12.*" | ||
| pytorch-cpu = ">=2.7.1,<3" | ||
| segment-geospatial = ">=1.2.0" | ||
| sam3 = ">=0.1.0.20251211" | ||
| jupyterlab = "*" | ||
| ipykernel = "*" | ||
| libopenblas = ">=0.3.30" | ||
| ``` |
There was a problem hiding this comment.
In this CPU-only pixi configuration, the platforms field only lists "linux-64" and "win-64" even though the installation section above explicitly targets Linux and macOS as well; this will prevent macOS users from using the provided pixi.toml as-is. In addition, this CPU environment still includes sam3 as a dependency even though the note below states that SAM 3 currently requires an NVIDIA GPU with CUDA support and cannot be used on CPU-only systems, which can confuse users and add an unnecessary heavy dependency. Consider either clarifying that this CPU environment is intended only for Linux/Windows and removing macOS from the earlier instructions, or adding macOS platforms here, and also removing sam3 (or clearly marking it as optional) from the CPU dependency list so CPU users are not misled into thinking SAM 3 will work for them.
|
🚀 Deployed on https://697640e84a4a6322f249d19d--opengeos.netlify.app |
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
Co-authored-by: Copilot <175728472+Copilot@users.noreply.github.com>
This PR adds comprehensive pixi installation instructions to address issues #471 and #476.
Summary
Added detailed pixi installation guide similar to the GeoAI QGIS plugin, providing users with a more reliable installation method that avoids common dependency conflicts.
Changes
docs/installation.mdwith step-by-step instructionsREADME.mdto recommend pixi as the preferred installation methodAddresses
uvbased installation #471 - uv based installation issuesBenefits
The instructions mirror the structure and detail of the GeoAI package QGIS plugin installation guide, as requested.