Upload folder using huggingface_hub

README.md

@@ -4,21 +4,14 @@ task_categories:
 - visual-question-answering
 tags:
 - benchability
+- figure4
 ---
 
-# NOT the Figure 4 evaluation set
+# BenchAbility Figure 4 -- held-out eval
 
-Kept for transparency, but superseded. This is a held-out in-domain dev split (10% of the training
-pool, by image) that an earlier draft of the pipeline used for scoring.
-
-**Figure 4 is not scored with this.** It is scored with lmms-eval's official task implementations over
-ChartQA / DocVQA val / InfoVQA val / OCRBench / TextVQA val / CharXiv -- the same instrument that
-produced Figures 1-3 -- because the gaps that decided the training mixture were measured there, and a
-gain measured with a different prompt and metric is not comparable to the gap it is meant to close.
-
-What this set is still good for: a fast in-domain check that training moved a capability at all,
-before committing GPU hours to the full official evaluation. It cannot separate "learned the
-capability" from "learned these 20 datasets".
+The 10% candidate-dev half of the same 884,143-row pool the two training mixtures are drawn from,
+balanced per capability. Split **by image**, so no picture here appears in either mixture, and both
+arms are equally blind to it.
 
 | capability | n |
 |---|---:|
@@ -28,3 +21,16 @@ capability" from "learned these 20 datasets".
 | diagram_and_infographic_understanding | 250 |
 | scene_text_recognition | 250 |
 | key_information_extraction | 168 (all that exists) |
+
+Balanced rather than proportional: the pool is 38% `chart_reasoning`, so a proportional dev set would
+measure that leaf precisely and the scarce ones not at all -- and the scarce ones are where the two
+arms differ most. One row per image, because a second question on the same picture is not an
+independent measurement.
+
+**This is an in-domain dev set, not the paper's evaluation.** Same 20 sources, held out by image. It
+answers "did training move this capability at all", which separates a broken run from a real one. It
+cannot separate "learned the capability" from "learned these 20 datasets" -- that needs the
+unseen-source set (ChartQAPro, OCRBench v2, DUDE, MME-RealWorld) which is not in this pool.
+
+Score with `evaluate.py` from the training bundle; measure the base model first, then each
+checkpoint, and report `gain = checkpoint - base` per capability.

data/train-00000-of-00003.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:e5ce0673b65dfe81d8023dc190284e380f8ff339d0f94c03481e97bfdd0815bc
-size 140862214
+oid sha256:e8b7c71ab45a7363fa93866c1c5f288fdf2c24e97e82a06dd7d4230801e26d5e
+size 140152747

data/train-00001-of-00003.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:29827997edc88a329aea76b3539fbb4dff523b8a8077a342ad6d1292c327ea0d
-size 212588753
+oid sha256:81f94c5c682634d6f843017d8ce1ad7609284bd7b7f9a78f18675d386370a7c7
+size 210196374

data/train-00002-of-00003.parquet

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d0c2f9c4322794f18e531774a7e78f7463b3a5c26304c3aab06a5f3b4c98b9a4
-size 148329280
+oid sha256:1abaffd38d29425a27e48bd7d72b8535ef8705ef2d5aa2b7280c36c9543b1d43
+size 147613289