https://yestino.com/entities/faketensor-support-b48e85Every event involving FakeTensor Supportenhttps://yestino.com/events/trunk-e320689d0698ae12305459e36b87ab5a0c64fccc-faketensor-su-d77d88https://yestino.com/events/trunk-e320689d0698ae12305459e36b87ab5a0c64fccc-faketensor-su-d77d88Tue, 15 Sep 2026 03:54:07 GMTFake impls for aten.sparse_compressed_tensor overloads run the meta kernel and restore the output device, instead of failing with a device mismatch. Sources: PyTorch Releases