An inventory of 875 retrogressive thaw slumps over a landscape of 54000 km^2, along the Qinghai-Tibet Engineering Corridor underlain by permafrost, was compiled using remote sensing and deep learning. The file in the format of ESRI Shapefile contains the boundary of each retrogressive thaw slump as vectors in the Coordinate Reference System of EPSG:32646 - WGS 84. The associated attribute table includes probability, time of the satellite images, source of the satellite images, the near roads labels, year of initiation, longitude and latitude, area (units: m^2). The corresponding names for the table fields are 'Probabilit', 'Year-month', 'Source Img', 'near roads', 'Initial_Y', 'Longitude', 'Latitude', 'Area'. The files in the GIF format show the temporal changes of each retrogressive thaw slump from 2016 to 2020.