dflash: read the confidence-head metadata key instead of only guessing

[bri-prism]

Cleanup and quality-of-life. No behaviour change for any valid checkpoint.

What

These files already write dflash.confidence_head, and nothing read it. The runtime worked the same fact out from whether markov_w1.weight is present.

That is fine for a complete export. The problem is when the two disagree. A DSpark checkpoint that lost its markov head loads as plain DFlash, reads drafts one row late, and acceptance falls off a cliff with nothing in the log to say why. That is the same failure #122 fixed for the CLI flag, still reachable through a bad export.

So: declare the key, read it, and cross-check it against the tensors. If the metadata claims a confidence head and markov_w1.weight is absent, the load fails and the message names the cause. The tensor stays the fallback when the key is absent, so older exports load exactly as before.

3 files, +14/-0.

Verification

• The real drafter is unchanged: 65.891% acceptance, mean length 3.58, identical to the current prism-v7 tip.
• The guard is tested rather than assumed. A synthetic GGUF that sets dflash.confidence_head with no markov_w1.weight now fails on load with the new message. Before this change the same file loaded and drafted badly.

Deliberately not in this change

dflash.sample_from_anchor is still only honoured on the DSpark path, so the row offset and the confidence head remain coupled through one flag. Decoupling them properly needs the converter to write sample_from_anchor explicitly for both lineages first: its default is true, which is DSpark-shaped, so honouring it uniformly today would change DFlash behaviour and break those models. Worth deciding together, since it spans the packing side.

[Copilot]

Pull request overview

Adds validation for DFlash confidence-head metadata while preserving tensor-based fallback for older checkpoints.

Changes:

• Registers the dflash.confidence_head metadata key.
• Rejects incomplete DSpark exports missing markov_w1.weight.

Reviewed changes

Copilot reviewed 3 out of 3 changed files in this pull request and generated 1 comment.

File	Description
src/models/dflash.cpp	Reads and validates confidence-head metadata.
src/llama-arch.h	Declares the metadata key enum.
src/llama-arch.cpp	Maps the enum to its GGUF key.

Suppressed comments (1)

src/models/dflash.cpp:132

• This only checks the true/missing mismatch. If the key is explicitly false while markov_w1.weight is present, the loader still treats the model as DSpark, so the metadata is neither preferred nor fully cross-checked as described. Reject the inverse mismatch too, while retaining tensor fallback only when the key is absent.

    if (has_kv_confidence_head && kv_confidence_head && !markov_meta) {
        throw std::runtime_error("dflash: metadata declares a confidence head, but markov_w1.weight is missing. "
                                 "The export is incomplete; it would load as plain DFlash and read drafts one row late");
    }

---

💡 Add a code-review agent skill or configure MCP servers for context-aware, tailored reviews. Learn more in the docs.