Datasets:

vinod-anbalagan
/

gridline-chartqa

id stringlengths 11 20	source stringclasses 2 values	question stringlengths 13 129	answer stringlengths 2 43 ⌀	chart_type stringclasses 7 values	task_type stringclasses 10 values	difficulty stringclasses 3 values	verified bool 1 class	split stringclasses 1 value	notes stringlengths 25 81
syn_bar_0001__q3	synthetic	How much more revenue did Consumer generate than Education?	$2400	bar	delta_absolute	medium	true	train	annotated bars; correct-by-construction
syn_bar_0001__q4	synthetic	Rank the top 3 segments by revenue.	Enterprise, Consumer, SMB	bar	rank_order	medium	true	train	annotated bars; correct-by-construction
syn_bar_0001__q5	synthetic	Would Consumer and SMB combined exceed Enterprise's revenue?	No	bar	hard_multi_step	hard	true	train	annotated bars; correct-by-construction
syn_bar_0002__q2	synthetic	Which product line made the least revenue?	Basic	bar	max_min	easy	true	train	annotated bars; correct-by-construction
syn_bar_0002__q3	synthetic	What was the revenue difference between Ultimate and Standard?	$1300	bar	delta_absolute	medium	true	train	annotated bars; correct-by-construction
syn_bar_0003__q3	synthetic	By how much does NA's revenue exceed APAC's?	$800	bar	delta_absolute	medium	true	train	annotated bars; correct-by-construction
syn_bar_0003__q4	synthetic	Name the bottom 3 regions in revenue.	MEA, LATAM, APAC	bar	rank_order	medium	true	train	annotated bars; correct-by-construction
syn_bar_0004__q2	synthetic	Which vertical had the lowest revenue?	Healthcare	bar	max_min	easy	true	train	annotated bars; correct-by-construction
syn_bar_0005__q2	synthetic	Which product sold the most units?	Widget D	bar	max_min	easy	true	train	annotated bars; correct-by-construction
syn_bar_0005__q3	synthetic	How many more units did Widget B sell than Widget A?	2200	bar	delta_absolute	medium	true	train	annotated bars; correct-by-construction
syn_bar_0005__q4	synthetic	Rank the top 3 products by units sold.	Widget D, Widget B, Widget C	bar	rank_order	medium	true	train	annotated bars; correct-by-construction
syn_bar_0006__q1	synthetic	Units sold by SKU-102?	720	bar	lookup_value	easy	true	train	annotated bars; correct-by-construction
syn_bar_0006__q3	synthetic	What was the unit sales difference between SKU-105 and SKU-102?	530	bar	delta_absolute	medium	true	train	annotated bars; correct-by-construction
syn_bar_0100__q3	synthetic	What was the revenue difference between Automotive and Healthcare?	$5100	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=0)
syn_bar_0101__q3	synthetic	By how much does Reseller's revenue exceed Retail's?	$3400	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=1)
syn_bar_0102__q2	synthetic	Which segment had the highest revenue?	Government	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=2)
syn_bar_0102__q3	synthetic	How much more revenue did Government generate than Education?	$8500	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=2)
syn_bar_0102__q4	synthetic	Rank the bottom 3 segments by revenue.	Education, Enterprise, Startup	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=2)
syn_bar_0103__q1	synthetic	Revenue for Beta?	$7300	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=3)
syn_bar_0103__q3	synthetic	What was the revenue difference between Zeta and Theta?	$5100	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=3)
syn_bar_0104__q3	synthetic	By how many units does Widget D exceed Widget C?	500	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=4)
syn_bar_0104__q4	synthetic	Name the top 3 products in units sold.	Widget D, Widget C, Widget A	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=4)
syn_bar_0105__q2	synthetic	Which SKU sold the fewest units?	SKU-207	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=5)
syn_bar_0105__q3	synthetic	How many more units did SKU-208 sell than SKU-207?	1300	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=5)
syn_bar_0106__q3	synthetic	By how much does Finance's revenue exceed Media's?	$1300	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=6)
syn_bar_0107__q1	synthetic	What was Reseller's revenue?	$14300	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=7)
syn_bar_0107__q3	synthetic	How much more revenue did Direct generate than Online?	$5400	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=7)
syn_bar_0108__q1	synthetic	Revenue for SMB?	$6100	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=8)
syn_bar_0108__q2	synthetic	Which segment made the most revenue?	Government	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=8)
syn_bar_0108__q3	synthetic	What was the revenue difference between Nonprofit and SMB?	$2200	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=8)
syn_bar_0108__q4	synthetic	List the top 3 segments by revenue.	Government, Nonprofit, Mid-Market	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=8)
syn_bar_0110__q3	synthetic	What was the unit sales difference between Widget C and Gizmo X?	2000	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=10)
syn_bar_0110__q4	synthetic	List the bottom 3 products by units sold.	Gizmo X, Widget D, Widget C	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=10)
syn_bar_0111__q3	synthetic	By how many units does SKU-202 exceed SKU-207?	200	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=11)
syn_bar_0112__q3	synthetic	How much more revenue did Energy generate than Healthcare?	$7300	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=12)
syn_bar_0112__q5	synthetic	Would Manufacturing and Healthcare combined exceed Automotive's revenue?	Yes	bar	hard_multi_step	hard	true	train	annotated bars; procedural gen (seed=42, idx=12)
syn_bar_0113__q2	synthetic	Which channel made the least revenue?	Wholesale	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=13)
syn_bar_0113__q3	synthetic	What was the revenue difference between Retail and Partner?	$6400	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=13)
syn_bar_0114__q3	synthetic	By how much does Consumer's revenue exceed Mid-Market's?	$300	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=14)
syn_bar_0114__q4	synthetic	Name the bottom 3 segments in revenue.	SMB, Enterprise, Mid-Market	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=14)
syn_bar_0115__q3	synthetic	How much more revenue did Zeta generate than Gamma?	$4600	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=15)
syn_bar_0115__q5	synthetic	Would Omega and Zeta combined exceed Theta's revenue?	Yes	bar	hard_multi_step	hard	true	train	annotated bars; procedural gen (seed=42, idx=15)
syn_bar_0116__q2	synthetic	Which product sold the most units?	Widget D	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=16)
syn_bar_0116__q3	synthetic	How many more units did Widget C sell than Widget A?	900	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=16)
syn_bar_0117__q3	synthetic	What was the unit sales difference between SKU-204 and SKU-208?	1600	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=17)
syn_bar_0118__q3	synthetic	What was the revenue difference between Media and Finance?	$500	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=18)
syn_bar_0118__q4	synthetic	List the bottom 3 verticals by revenue.	Healthcare, Retail, Finance	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=18)
syn_bar_0119__q1	synthetic	How much revenue did Wholesale generate?	$13200	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=19)
syn_bar_0119__q3	synthetic	By how much does Wholesale's revenue exceed Reseller's?	$7600	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=19)
syn_bar_0120__q4	synthetic	Rank the top 3 segments by revenue.	Enterprise, Government, SMB	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=20)
syn_bar_0121__q1	synthetic	Revenue for Zeta?	$2600	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=21)
syn_bar_0121__q3	synthetic	What was the revenue difference between Theta and Epsilon?	$1400	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=21)
syn_bar_0123__q3	synthetic	How many more units did SKU-204 sell than SKU-205?	1100	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=23)
syn_bar_0124__q3	synthetic	By how much does Automotive's revenue exceed Retail's?	$3700	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=24)
syn_bar_0125__q1	synthetic	What was Marketplace's revenue?	$10100	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=25)
syn_bar_0125__q3	synthetic	How much more revenue did Marketplace generate than Wholesale?	$4500	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=25)
syn_bar_0126__q3	synthetic	What was the revenue difference between Enterprise and Consumer?	$2200	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=26)
syn_bar_0126__q4	synthetic	List the bottom 3 segments by revenue.	Consumer, SMB, Enterprise	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=26)
syn_bar_0127__q1	synthetic	How much revenue did Theta generate?	$7700	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=27)
syn_bar_0127__q5	synthetic	Together, would Omega and Gamma out-earn Delta?	Yes	bar	hard_multi_step	hard	true	train	annotated bars; procedural gen (seed=42, idx=27)
syn_bar_0128__q3	synthetic	What was the unit sales difference between Widget C and Widget E?	3200	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=28)
syn_bar_0128__q4	synthetic	List the top 3 products by units sold.	Widget C, Widget E, Widget D	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=28)
syn_bar_0129__q3	synthetic	By how many units does SKU-206 exceed SKU-207?	300	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=29)
syn_bar_0130__q4	synthetic	Rank the bottom 3 verticals by revenue.	Tech, Healthcare, Manufacturing	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=30)
syn_bar_0131__q1	synthetic	Revenue for Reseller?	$3400	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=31)
syn_bar_0131__q3	synthetic	What was the revenue difference between Direct and Wholesale?	$1800	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=31)
syn_bar_0132__q2	synthetic	Which segment was smallest in revenue?	Startup	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=32)
syn_bar_0132__q3	synthetic	By how much does Enterprise's revenue exceed Startup's?	$6100	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=32)
syn_bar_0133__q3	synthetic	How much more revenue did Zeta generate than Omega?	$2600	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=33)
syn_bar_0136__q1	synthetic	Revenue for Media?	$2300	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=36)
syn_bar_0136__q3	synthetic	What was the revenue difference between Automotive and Finance?	$4900	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=36)
syn_bar_0136__q4	synthetic	List the top 3 verticals by revenue.	Automotive, Retail, Finance	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=36)
syn_bar_0137__q3	synthetic	By how much does Reseller's revenue exceed Retail's?	$4800	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=37)
syn_bar_0138__q1	synthetic	What was Mid-Market's revenue?	$7700	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=38)
syn_bar_0138__q4	synthetic	Rank the bottom 3 segments by revenue.	Government, Consumer, Enterprise	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=38)
syn_bar_0139__q3	synthetic	What was the revenue difference between Gamma and Beta?	$1200	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=39)
syn_bar_0140__q1	synthetic	What was Widget E's unit total?	7600	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=40)
syn_bar_0140__q3	synthetic	By how many units does Widget D exceed Widget F?	900	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=40)
syn_bar_0140__q4	synthetic	Name the top 3 products in units sold.	Gizmo Y, Widget D, Widget E	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=40)
syn_bar_0141__q1	synthetic	How many units did SKU-204 sell?	500	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=41)
syn_bar_0141__q3	synthetic	How many more units did SKU-207 sell than SKU-204?	800	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=41)
syn_bar_0142__q3	synthetic	By how much does Manufacturing's revenue exceed Media's?	$600	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=42)
syn_bar_0142__q5	synthetic	Together, would Healthcare and Media out-earn Tech?	Yes	bar	hard_multi_step	hard	true	train	annotated bars; procedural gen (seed=42, idx=42)
syn_bar_0143__q2	synthetic	Which channel had the lowest revenue?	Partner	bar	max_min	easy	true	train	annotated bars; procedural gen (seed=42, idx=43)
syn_bar_0143__q3	synthetic	How much more revenue did Reseller generate than Direct?	$10400	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=43)
syn_bar_0144__q1	synthetic	Revenue for Education?	$9300	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=44)
syn_bar_0144__q3	synthetic	What was the revenue difference between Enterprise and Startup?	$700	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=44)
syn_bar_0145__q3	synthetic	By how much does Gamma's revenue exceed Delta's?	$4700	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=45)
syn_bar_0147__q3	synthetic	By how many units does SKU-207 exceed SKU-205?	1000	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=47)
syn_bar_0148__q3	synthetic	How much more revenue did Energy generate than Finance?	$2900	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=48)
syn_bar_0148__q4	synthetic	Rank the top 3 verticals by revenue.	Energy, Healthcare, Finance	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=48)
syn_bar_0149__q1	synthetic	Revenue for Online?	$4100	bar	lookup_value	easy	true	train	annotated bars; procedural gen (seed=42, idx=49)
syn_bar_0149__q3	synthetic	What was the revenue difference between Reseller and Direct?	$4300	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=49)
syn_bar_0150__q3	synthetic	By how much does Startup's revenue exceed SMB's?	$2900	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=50)
syn_bar_0150__q4	synthetic	Name the bottom 3 segments in revenue.	SMB, Startup, Education	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=50)
syn_bar_0151__q3	synthetic	How much more revenue did Beta generate than Delta?	$1200	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=51)
syn_bar_0151__q5	synthetic	Would Zeta and Omega combined exceed Gamma's revenue?	Yes	bar	hard_multi_step	hard	true	train	annotated bars; procedural gen (seed=42, idx=51)
syn_bar_0153__q3	synthetic	What was the unit sales difference between SKU-203 and SKU-201?	100	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=53)
syn_bar_0154__q3	synthetic	What was the revenue difference between Energy and Media?	$8900	bar	delta_absolute	medium	true	train	annotated bars; procedural gen (seed=42, idx=54)
syn_bar_0154__q4	synthetic	List the bottom 3 verticals by revenue.	Media, Retail, Manufacturing	bar	rank_order	medium	true	train	annotated bars; procedural gen (seed=42, idx=54)

End of preview. Expand in Data Studio

Adaption Charts P2 — Gold Chart-QA Dataset

A verified, quality-first chart question-answering dataset built for the Adaption Labs AutoScientist Challenge (Part 2, Data Visualization track). Two sources: a programmatically generated synthetic core (correct-by-construction) and a hand-authored hardset built from real public dashboards and reports.

At a glance

1415 rows total — 1317 synthetic + 98 hardset
7 chart types — bar, line, grouped_bar, stacked_bar, pie, donut, mixed
10 task types — every allowed task_type populated at target
English only, real-world business/finance/health/policy domains
Every row verified=true

Schema

Each row:

column	type	description
`file_name`	string	HF imagefolder path to the chart PNG (`images/...`)
`id`	string	unique row id, e.g. `syn_bar_0001__q1` or `hs_0007__q3`
`source`	enum	`synthetic` \| `hardset`
`question`	string	short flat-register question
`answer`	string	short exact answer, e.g. `47`, `Enterprise`, `-$1440.81`, `11.3%`, `Yes`
`chart_type`	enum	`bar` \| `line` \| `grouped_bar` \| `stacked_bar` \| `pie` \| `donut` \| `mixed`
`task_type`	enum	10 values: `lookup_value`, `delta_absolute`, `max_min`, `rank_order`, `compare_categories`, `aggregation_sum_avg`, `multi_series_compare`, `trend_direction`, `percent_change_ratio`, `hard_multi_step`
`difficulty`	enum	`easy` \| `medium` \| `hard`
`verified`	bool	`true` on all rows (see verification protocol below)
`split`	enum	`train` on all rows (no held-out val/test in this file)
`notes`	string	provenance / attribution / generator notes

Composition

By source

source	rows	share
synthetic	1317	93.1%
hardset	98	6.9%

By chart_type

chart_type	rows
bar	417
line	288
grouped_bar	201
stacked_bar	184
pie	155
mixed	104
donut	66

By task_type

task_type	rows
max_min	301
lookup_value	273
delta_absolute	156
rank_order	141
compare_categories	131
trend_direction	124
multi_series_compare	111
aggregation_sum_avg	89
percent_change_ratio	56
hard_multi_step	33

By difficulty

difficulty	rows
medium	726
hard	358
easy	331

Perceptual difficulty

Most chart-QA datasets vary difficulty by arithmetic — more steps in the calculation. That is not what challenges a vision-language model. A capable VLM handles multi-step arithmetic easily once it has read the values; what it struggles with is reading the values in the first place.

358 rows (25.3%) are hard by perceptual construction. Eight mechanics:

mechanic	what it does	why it's hard
`truncated_axis`	y-axis starts well above zero	bar height ratios badly misrepresent value ratios; a model comparing pixels is wrong, a model reading labels is right
`unlabeled`	no value annotations at all	values are snapped exactly onto gridlines, so the answer stays unambiguous — but only if the axis is actually read
`near_tie`	top two values differ by ~1–2%	eyeballing the tallest bar fails
`similar_colors`	series palette uses near-identical hues	the legend must be resolved rather than pattern-matched
`many_categories`	12–16 categories, rotated labels, small font	dense visual scanning
`crowded_legend`	legend placed over the plot area	partial occlusion of the data
`log_scale`	logarithmic y-axis	equal pixel distances are unequal value deltas
`dual_axis`	two y-axes at different scales	reading the wrong axis yields a plausible but wrong number

Roughly 90 of these rows carry no value labels whatsoever — the answer must be read off gridlines. Answers remain correct-by-construction because values are generated as exact multiples of the tick step.

Every hard row's notes field names its mechanic, e.g. hard/truncated_axis; procedural gen (seed=777, idx=12).

Verification protocol

Every row is verified=true, but the mechanism differs by source:

Synthetic — correct-by-construction. Each chart is rendered from a seeded pseudorandom value distribution. The answer to every question is computed from those underlying values before the chart image is drawn. There is no visual estimation involved. The full generation pipeline lives in the companion repo alongside this dataset. Seeds are deterministic.

Hardset — hand-authored, human-reviewed. Every hardset row was authored one at a time from a real chart screenshot. Each row was reviewed row-by-row against the source image during authoring, with the answer recorded only after cross-checking what the chart actually shows. All questions were designed to be answerable from the image alone without external context.

Data sources (hardset only)

Hardset uses screenshots of publicly available charts from official statistical / research bodies. Each row's notes column carries the source institution and original screenshot filename.

source	rows	note
Statistics Canada (StatCan)	24	Open Licence Agreement
U.S. Bureau of Labor Statistics (BLS)	29	U.S. Government works, public domain
European Central Bank (ECB)	18	Reproduction permitted with attribution
World Health Organization (WHO)	21	See attribution notes below
Bank of Canada (BoC)	2	Open license
Climate Policy Database	4	CC-BY-4.0

Attribution notes. WHO source materials are typically licensed under CC-BY-NC-SA-3.0 IGO. This dataset uses WHO chart screenshots for the transformative purpose of vision-language model training, with full attribution preserved in each row's notes. Downstream users concerned about commercial use should filter rows where notes begin with hardset; WHO; and treat them separately.

Intended use

Fine-tuning multimodal chart-QA models. Short-answer chart understanding benchmarks. Ablation studies on chart-type or task-type coverage.

Not intended for downstream tasks that require gold-standard OCR fidelity or exact numerical extraction from dense financial tables — this dataset targets reasoning about charts, not exact digit extraction.

Known limitations

English only.
Synthetic aesthetic is uniform. All synthetic rows use the same matplotlib renderer with a consistent style. Real-world visual diversity comes entirely from the hardset rows.
Hardset skews toward max_min and lookup_value. Real dashboards naturally support these tasks; the balance step trims synthetic max_min heavily to compensate, but hardset structural bias remains.
chart_type=mixed is exclusively hardset — synthetic doesn't generate multi-panel dashboards. Mixed charts are visually harder for VLMs and represent a natural difficulty axis.
No held-out split in this file. An external evaluation slice (ChartQAPro-derived) lives outside this dataset by design.

Citation

If you use this dataset, please cite:

@misc{anbalagan2026adaptioncharts,
  title  = {Adaption Charts P2: A Small Gold Chart-QA Dataset},
  author = {Anbalagan, Vinod},
  year   = {2026},
  howpublished = {\url{https://huggingface.co/datasets/vinod-anbalagan/adaption-charts-p2-gold}}
}

Please also credit the upstream chart sources listed in the "Data sources" section above when relevant.

License

This dataset is released under CC-BY-4.0. See attribution notes above for source-specific considerations.

Downloads last month: 151

Models trained or fine-tuned on vinod-anbalagan/gridline-chartqa

vinod-anbalagan/gridline

Image-Text-to-Text • Updated 2 days ago • 27