PostX

The public dataset is being converted to training-ready JSONL, with one conversation per line in the standard {"messages": [{"role": "user", "content": "..."}, {"role": "assistant", "content": "..."}]} format.

The training file will contain conversation content and necessary tool context. Repository paths, source inventories, statistics, binary assets, and numeric activation tensors will not be training records. The earlier source-preserving collection has been removed from the current dataset revision while this conversion runs.

The new training file is not yet published. This page will be updated with its validated record count, coverage, and loading instructions when conversion finishes. Original source licenses continue to apply; this collection does not grant a new blanket license.