Welcome to the Minecraft Landscape Classification challenge!

Your goal is to build a model that, given a static in-game screenshot, predicts which biome it comes from (Desert, Plains, Forest, etc.). These biomes dictate terrain, vegetation, and gameplay mechanics—automatic classification powers downstream tools from procedural generation to analytics dashboards.

Minecraft worlds are divided into distinct biomes. Though visually iconic, many screenshots—especially those taken seconds apart—look nearly identical. To avoid data leakage, we’ve grouped every 10 temporally consecutive frames into the same split. Your model must generalize beyond these near-duplicates.

You are given the folders `train/` and `test/` along with the dataframes `train.csv` and `test.csv`. Your task is to predict which biome the images in `test/` belong to and submit them to the leaderboard.