Fix SVG download: use single quotes inside style attr, no duplicate style

Cap chart size: max-width 880px and max-height 460px so charts have natural proportions

Wider left column (480-580px) + capped schema height with scroll

Fluid layout (no max-width) to remove dark side bars

UX premium: chart badge, SQL syntax highlighting + copy, smart placeholder, drag highlight

Cap to top 30 categories, hide gr.File X button, single-row buttons, no duplicate title

Cap to top 30 categories, hide gr.File X button, single-row buttons, no duplicate title

Polish: 4-step pipeline progress, large empty state w/ icon, subtle entrance animations

Bigger rectangular layout: 1600x920 max with 16:9 feel

Dynamic suggested questions based on actual dataset columns (no more Titanic-only)

Layout: 2-column split (left=data+question, right=results)

Result narration (Qwen-powered) + UI polish for upload area + dark-mode fixes

Result narration (Qwen-powered) + UI polish for upload area + dark-mode fixes

Result narration (Qwen-powered) + UI polish for upload area + dark-mode fixes

SQL self-correction: 3 retries with error context + GROUP BY rule in prompt

SQL self-correction: 3 retries with error context + GROUP BY rule in prompt

SVG: standalone download (XML prolog, white bg, fixed dims) + Plotly first

SVG: standalone download (XML prolog, white bg, fixed dims) + Plotly first

SVG: standalone download (XML prolog, white bg, fixed dims) + Plotly first

Bump sdk_version 5.7.1 -> 5.12.0 (fixes ZeroGPU quota for logged-in users)

Dark-mode fixes: visible title/chip, hide gr.File uploaded box, transparent question group

Better SQL prompt: column-name mapping rules + few-shot + distinct value hints

Better SQL prompt: column-name mapping rules + few-shot + distinct value hints

Remove hf_oauth + LoginButton — they conflict with native ZeroGPU auth

Remove hf_oauth + LoginButton — they conflict with native ZeroGPU auth

Pin kaleido==0.2.1 + native SVG fallback when kaleido fails

Pin kaleido==0.2.1 + native SVG fallback when kaleido fails

UX: schema below dataset / above textbox, single-shot Ask, Clear=dataset only

Clear button now resets the upload file too

Use only valid OAuth scope: inference-api

Add LoginButton + hf_oauth: required for ZeroGPU Pro quota in iframe context

Add LoginButton + hf_oauth: required for ZeroGPU Pro quota in iframe context

UX: CSV/SVG download links, schema preview on load, cleaner SVG theme

UX: CSV/SVG download links, schema preview on load, cleaner SVG theme

Load LoRA via PeftModel on top of standard base models (fixes r=16 vs r=8 mismatch)

Load LoRA via PeftModel on top of standard base models (fixes r=16 vs r=8 mismatch)

Load LoRA via PeftModel on top of standard base models (fixes r=16 vs r=8 mismatch)

Pin transformers==4.46.3 to fix Qwen tokenizer compat

ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU

ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU

ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU

ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU

ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU

Remove hf_oauth/LoginButton, pre-download models in build phase

Remove hf_oauth/LoginButton, pre-download models in build phase

Add gr.LoginButton for HF OAuth

Enable hf_oauth so visitors authenticate and get proper ZeroGPU quota

Sequential model load/unload: keep peak VRAM at ~14GB

Sequential model load/unload: keep peak VRAM at ~14GB

Sequential model load/unload: keep peak VRAM at ~14GB