Even the 30B mis-identifies named sex positions (doggy/cowgirl) from images, so
position_name is removed. The pose cluster is now purely observable geometry:
body_orientation enriched with facing direction (who faces whom), plus
limb_arrangement / contact_points / pose. The agent composes any named label from
these reliable primitives. 23 default axes. Docs/examples updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The 4B over-uses 'partial' (mislabels identical ref/gen and clear opposites) and
also mis-identifies fine-grained content (e.g. names a position 'doggy'/'cowgirl'
when it is neither). Deterministic fix: force verdict=match when normalized
ref==gen. Prompt hardened to not default to 'partial' (opposites=mismatch). Docs:
the 4B is only reliable for coarse attributes — use the 30B for fine-grained
recognition; prefer grounded geometry axes over named-position labels.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Describe mode now produces a single coherent, internally-consistent canonical
scene description (paragraph + per-axis spec, written to canonical_reference in
the report). Compare gains an optional reference_description input: when set, it
anchors on that fixed text and shows only the generated image (no swap) — so the
reference side never drifts or self-contradicts across iterations; only the
generated image is re-described each turn. agent_bridge gains --ref-desc /
--ref-desc-file (reads the describe report's canonical_reference). Docs + example
workflow updated.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The 4B's 0-1 scores were unreliable (identical ref/gen scored ~0.6), so the
judge now returns verdict match/partial/mismatch per axis; overall_score and a
new mismatch_count are computed from verdicts on our side (reliable, monotonic).
Expanded the action/pose cluster into position_name, body_orientation,
limb_arrangement, penetration, contact_points, genital_visibility (+ breast_size)
so explicit poses carry detail. Each axis now ships a one-line definition in the
prompt so gender_mix/subject_count stop absorbing positional text. 24 axes total.
Example workflows use the node default (axes=''). Docs realigned; stop condition
is now mismatch_count==0.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
New mode on QwenVLImageJudge: 'describe' looks at the reference alone and returns
a prompt-ready caption + per-axis target spec to seed the very first prompt (the
generator has nothing to reproduce yet). 'compare' is the existing ref-vs-gen
scoring. generated_image is now optional (required only for compare); shared
generation refactored into _generate_from_messages; third output renamed
diff_analysis -> analysis (mode-agnostic). agent_bridge gains --mode (describe
needs no receptor/prompt); added workflow_describe_api.json. Docs updated with the
first-pass bootstrap step. Fixed error-return arity to 5-tuple.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
The local VLM now only observes and scores; correction is left to the stronger
external agent. Each axis reports the target value (ref), the current value (gen)
and the closeness (score) — the target/current/distance an agent needs to
calibrate. Expanded to ~20 granular axes (identity/body/wardrobe/action/affect/
camera/render) so the action cluster stays discriminative for explicit content.
swap_eval now inverts ref/gen of the swapped pass; diff summary sorts worst-first;
default max_new_tokens 1024. Docs aligned.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
ComfyUI-converted checkpoints ship the template as chat_template.jinja
(not on the processor), so apply_chat_template raised 'this processor does
not have a chat template'. Backfill processor.chat_template from
chat_template.jinja/.json or the tokenizer at load time, and fall back to a
hand-built Qwen-VL ChatML prompt if none exists. Also keep *.jinja in the
auto-download patterns.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Ethanfel