TabFM: A zero-shot foundation model for tabular data

4 comments

• woadwarrior0132 minutes ago
  In hindsight, the Prior Labs exit to SAP couldn't have been timed better.
• actusual1 hour ago
  150,000 rows of data, where will I store it all?!
  • nojito36 minutes ago
    The biggest misconception that people have when modeling using tabular data is that more data = better model.
• hodgehog111 hour ago
  On the one hand, this is impressive. TabPFN was already state of the art and is seriously shaking up Bayesian prediction for tabular data (which is almost everything).<p>On the other hand, perhaps it is just me, but I do not feel that this is an acceptable form of benchmark reporting in this domain. TabArena actually has multiple metrics, since ELO does not properly quantify the degree of improvement. The fact that these are not displayed here should give pause. Also the results section in the GitHub is a dumpster fire.
  • Eridrus1 hour ago
    GitHub Repo: Please see the results folder<p>Results folder: Here&#x27;s some undocumented parquet files<p>Definitely feels like they&#x27;re hiding the ball lol.<p>If they had good benchmarks they&#x27;d talk about them.<p>Not comparing to tuned xgboost is also a warning sign.
• kingjimmy2 hours ago
  interesting to see this from Google after the SAP acquisition of Prior Labs.