metadata: add score_type/score_params/score_protocol (official grading protocol taxonomy); task_type is now format-only (judge_scored migrated to vqa + llm_judge). See dataset-upload skill references/metadata-json.md.
Browse files- metadata.json +5 -0
metadata.json
CHANGED
|
@@ -10,6 +10,11 @@
|
|
| 10 |
"single_image_start"
|
| 11 |
],
|
| 12 |
"task_type": "multiple_choice_vqa",
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
|
| 13 |
"prompt_template": "<image>{{ question }}\n{% for k, v in options.items() %}{{ k }}. {{ v }}\n{% endfor %}Answer with the option's letter from the given choices directly.",
|
| 14 |
"mapping_from_source": {
|
| 15 |
"media": {
|
|
|
|
| 10 |
"single_image_start"
|
| 11 |
],
|
| 12 |
"task_type": "multiple_choice_vqa",
|
| 13 |
+
"score_type": "llm_extract",
|
| 14 |
+
"score_protocol": {
|
| 15 |
+
"reference": "vlmevalkit@vlmeval/dataset/image_mcq.py:249-324 (ImageMCQDataset.evaluate_heuristic) -> vlmevalkit@vlmeval/dataset/utils/multiple_choice.py:475-508 mcq_vanilla_eval: can_infer rule prefetch, then GPT extractor (extract_answer_from_item @vlmeval/dataset/utils/multiple_choice.py:359-407) maps verbose output to a letter, rule-compare vs GT; AI2D_TEST registered @vlmeval/dataset/image_mcq.py:100. Cross-check: lmms-eval@lmms_eval/tasks/ai2d/ai2d.yaml:41-55 (MultiChoiceRegexFilter + exact_match, pure rule)",
|
| 16 |
+
"note": "mm-eval prompt follows the lmms-eval trailer; lmms-eval grades AI2D with deterministic regex extraction only, VLMEvalKit adds the GPT extraction fallback. Both are plain per-sample accuracy — no aggregation caveat."
|
| 17 |
+
},
|
| 18 |
"prompt_template": "<image>{{ question }}\n{% for k, v in options.items() %}{{ k }}. {{ v }}\n{% endfor %}Answer with the option's letter from the given choices directly.",
|
| 19 |
"mapping_from_source": {
|
| 20 |
"media": {
|