A VLM answers every question at temperature 0 with reasoning enabled. Its answer is compared with the official ground truth by the verifier described below; matches go to ChartQA_RS_think and ChartQA_RS_nothink, non-matches go to ChartQA_rejected with the ground truth preserved.

The question field is byte-identical to the prompt the model actually received, including any appended instruction, so the stored training text and the inference-time input cannot drift apart.

How a match is decided

Every rule below is stated with its exact constants and a runnable reference implementation, so any accept/reject decision in this dataset can be recomputed. All snippets assume:

def levenshtein(a, b):
    if a == b: return 0
    if not a: return len(b)
    if not b: return len(a)
    prev = list(range(len(b) + 1))
    for i, ca in enumerate(a, 1):
        cur = [i] + [0] * len(b)
        for j, cb in enumerate(b, 1):
            cur[j] = min(prev[j] + 1, cur[j - 1] + 1, prev[j - 1] + (ca != cb))
        prev = cur
    return prev[len(b)]

`relaxed` — ChartQA relaxed accuracy

The official ChartQA metric. If both sides parse as numbers, the answer is correct within 5% relative tolerance — chart values are read off axes, so exact digit agreement is not a fair requirement. Otherwise the strings must match exactly after lowercasing, trimming and dropping a trailing period.

_NUM = re.compile(r"^[-+]?\$?\s*([0-9]*\.?[0-9]+)\s*%?$")

def relaxed_pair(pred, gt, tol=0.05):
    p, q = pred.strip().lower().rstrip("."), gt.strip().lower().rstrip(".")
    mp, mq = _NUM.match(p.replace(",", "")), _NUM.match(q.replace(",", ""))
    if mp and mq:
        a, b = float(mp.group(1)), float(mq.group(1))
        return 1.0 if abs(a - b) <= tol * max(abs(b), 1e-9) else 0.0
    return 1.0 if p.replace(",", "") == q.replace(",", "") else 0.0

Shared pre-processing (applies to every verifier below)

Both sides are expanded into a small set of candidate strings and the best-scoring pair decides. The prediction contributes: the raw string; its first line; its first paragraph; the text after the last Answer: marker; the span between Answer: and Rationale:; the stated final answer of a worked solution (recognised from \boxed{…}, Answer:, the final answer is …, the answer is …, so the answer is …, 故选X); and a markdown-stripped variant. The reference contributes the same forms.

When the reference is a wrapper around the real answer (The answer is Red), the raw string is replaced by the extracted answer rather than kept beside it: the wrapper text is shared by every wrong answer too, so The answer is Blue would otherwise score 0.78 against it.

import re

def final_answer(gt):
    """Stated answer of a worked solution, or None."""
    pats = [r"\\boxed\{([^{}]{1,80})\}",
            r"(?is)\banswer\s*[:：]\s*(.{1,80}?)\s*$",
            r"(?is)\b(?:the\s+)?final\s+answer\s+is\s*[:：]?\s*(.{1,80}?)\s*[.。]?\s*$",
            r"(?is)\b(?:the\s+)?answer\s+is\s*[:：]?\s*(.{1,80}?)\s*[.。]?\s*$",
            r"(?s)故\s*选\s*[:：]?\s*\(?([A-H])\)?"]
    for p in pats:
        ms = list(re.finditer(p, (gt or "").strip()))
        if ms:
            v = re.sub(r"\s+", " ", ms[-1].group(1).strip().strip("$ \t.。，,;:"))
            if v and len(v) <= 80:
                return v
    return None

Second pass: model-as-judge

String verifiers cannot judge free-form answers at all, and they under-credit answers that are correct but differently phrased. Every item the verifier rejected is therefore re-examined by the same VLM in a text-only pass that sees the question, the reference and the candidate — no image, no re-generation, it only grades text that was already produced.

The judge is instructed to compare against the reference and never to solve the question itself, so an answer is not promoted merely because the model believes it is true. It answers CORRECT, WRONG, or UNSURE; numbers must agree, and UNSURE is reserved for references that are themselves unusable or ambiguous. UNSURE is never counted as correct — only a CORRECT verdict moves a row into the accepted splits. WRONG and UNSURE both remain in the rejected split, so an unresolved case is never silently promoted to a pass.

Calibration: run over items the string verifier had already accepted, the judge agreed on 59/60. Spot-checking its overturns shows it discriminates rather than rubber-stamps — it keeps \frac{L}{n} rejected against a reference of \frac{L}{n-1}, and returns UNSURE when the model argues no finite answer exists against a numeric reference.

What `answer` holds

In the accepted splits answer is the model's own verified output — the text that passed the verifier — so it agrees with the reasoning in think. The original reference annotation is kept beside it as org_answer. Train on answer; use org_answer when you need the source label.

The two differ more often than one would guess: the verifier accepts on meaning, not on characters, so casing, spacing and formatting routinely differ (bridgestone vs Bridgestone, ( n x m ) vs (n x m)). Storing the reference in the answer slot would pair a model's reasoning with someone else's wording.

In the rejected split the roles are different: answer is the official ground truth and wrong_vlm is what the model said instead.

Schema

image — HF Image() — renders directly in the dataset viewer
qa — list of {question, answer, note}
clean_meta — per-row provenance and the cleaning policy that admitted it
image_sha256 — content hash of the image

Examples

Three rows taken straight from this split. Images are the original files as stored in the dataset, and long fields are shown in full.

Example 1

field	value
`question`	What health risks could be associated with consuming frequent meals like the barbecue sandwich shown in the image? Answer in short without any other content.
`answer`	Increased risk of obesity, heart disease, type II diabetes, and certain cancers due to high fat content and preservatives.
`note`	long-form: excluded from rejection sampling

Example 2

field	value
`question`	What issues can be found in the image? A. Color balance issues B. Sharpness and focus issues C. Composition issues D. Noise and exposure issues Answer with the option's letter from the given choices directly. Answer in short without any other content.
`answer`	The main issue with the image appears to be related to sharpness and focus, as the fireworks and tree silhouettes are not clearly defined. There is also noticeable noise and some exposure issues, evidenced by the grainy texture and the overly bright areas contrasted with deep shadows. Hence, while 'D' for noise and exposure issues is partly correct, 'B' stands out as the more comprehensive choice, addressing the lack of clarity in detail.
`note`	long-form: excluded from rejection sampling

Example 3

field	value
`question`	How many students graduate from NYC high schools and enroll at City University in CUNY? Answer in short without any other content.
`answer`	Reason: The number of students who graduate from NYC high schools and go to college comes close to 40,000. Answer: Close to 40,000
`note`	long-form: excluded from rejection sampling

Cleaning applied before rejection sampling

Provenance check. Each QA item records where it came from; items whose recorded source does not resolve to the image they are attached to are dropped.
Task purity. The pool's InfinityMM slice assigns samples to a family by image match rather than by dataset origin, which pulls unrelated general-VQA questions into benchmark families. Families where that content measured off-task keep only their genuine benchmark QA.
Multiple-choice normalisation. Where options had been flattened into prose, they are re-rendered as a lettered block and the reference is rewritten to the bare letter, so the task is well posed and the answer is unambiguous.

Downloads last month: 34

Total file size:

347 MB

You need to agree to share your Hugging Face information to access this dataset

ChartQA — ChartQA_longform

How the data was produced