Replace write_slice_i32_d_pub (sync + alloc + DtoD) with persistent MappedI32Buffer staging. Host writes to host_ptr, async DtoD to frd_labels_d on the train stream — zero sync, zero alloc per step. Also: make stream field pub for external DtoD access. Co-Authored-By: Claude Opus 4.7 <noreply@anthropic.com>