{
  "train": [
    "241019_TESvsT1_1Set_cropped",
    "241019_TESvsT1_2Set_cropped",
    "241020_GENvsFLY_1Set_cropped",
    "241020_GENvsFLY_2Set_cropped",
    "241020_GENvsFLY_4Set_cropped",
    "241020_GENvsFLY_5Set_cropped",
    "241026_WBGvsBLG_1Set_cropped",
    "241026_WBGvsBLG_2Set_cropped",
    "241027_T1vsGEN_1Set_cropped",
    "241027_T1vsGEN_2Set_cropped",
    "241027_T1vsGEN_4Set_cropped",
    "241102_T1vsBLG_1Set_cropped",
    "241102_T1vsBLG_2Set_cropped",
    "241102_T1vsBLG_4Set_cropped",
    "241102_T1vsBLG_5Set_cropped"
  ],
  "test": [
    "241019_TESvsT1_3Set_cropped",
    "241020_GENvsFLY_3Set_cropped",
    "241026_WBGvsBLG_3Set_cropped",
    "241027_T1vsGEN_3Set_cropped",
    "241102_T1vsBLG_3Set_cropped"
  ],
  "validation": "The released code samples a small validation subset from the training split during optimization, matching the implementation used in the revision."
}