tanishqgandhi1908 opened a new pull request, #98: URL: https://github.com/apache/incubator-texera-site/pull/98
### What changes were proposed in this PR? - Adds `content/blog/ml-model-support/index.md`, a feature-introduction post on bringing a trained ML model into Texera. It covers the three pieces that only work together: models as a first-class resource (create, upload, version, share, publish, find on the Hub); mounting a model version into a computing unit, so a Python UDF names it as a parameter and reads it from a local directory with nothing downloaded; and curated computing-unit images, for models whose runtime the stock image does not carry. - Adds five images under `static/images/blog_hero/` — the Models page (also the card image), a model's page, the UDF property panel showing a `MODEL` parameter bound to a model version, the out-of-pod mount architecture diagram, and the image picker in the create-computing-unit dialog. - Adds three screen recordings under `static/videos/` — uploading and committing a version, registering a curated image, and the end-to-end run. All three are silent; their audio tracks were removed, so the players have nothing to unmute. The short one loops; the two longer ones play once. The post is written with inline styles, matching the four posts currently on `main`. If #95 merges first this will need converting to its class system; happy to rebase onto it either way. Note for reviewers: the three videos come to about 19 MB, in the same range as the existing `pve-demo.mp4`. ### Any related issues, documentation, discussions? Describes work tracked in apache/texera#6494 (ML models as a resource), apache/texera#6606 (mounting versioned repositories into computing-unit pods) and apache/texera#8466 (curated computing-unit images). ### How was this PR tested? - Confirmed every referenced asset resolves: the post references eight files under `static/`, and all eight are included here. - Rendered the post body in a browser at desktop width and at 375px: the three-item feature grid collapses from three columns to one, and the page body never scrolls sideways. - Checked all three videos load with no decode error at 1920x1080 and the expected durations, that `autoplay` fires without interaction, and that each reports no audio track. - Verified each mp4 is fast-start (`moov` before `mdat`), so playback begins without fetching the whole file. - **Not** built with Hugo locally — this machine has neither Hugo nor Go installed, so the post has not been seen inside the Docsy theme. The content is one Markdown file plus static assets and adds no templates, layouts or config, but a reviewer with the toolchain should confirm it renders in the theme and that the card appears on `/blog/`. ### Was this PR authored or co-authored using generative AI tooling? Co-authored with Claude Opus 5 in compliance with ASF -- This is an automated message from the Apache Git Service. To respond to the message, please log on to GitHub and use the URL above to go to the specific comment. To unsubscribe, e-mail: [email protected] For queries about this service, please contact Infrastructure at: [email protected]
