Fix SVG download: use single quotes inside style attr, no duplicate style 964b628 Running verified DanielRegaladoCardoso commited on 16 days ago
Cap to top 30 categories, hide gr.File X button, single-row buttons, no duplicate title 7d5a424 verified DanielRegaladoCardoso commited on 16 days ago
Result narration (Qwen-powered) + UI polish for upload area + dark-mode fixes 5ae5227 verified DanielRegaladoCardoso commited on 16 days ago
Result narration (Qwen-powered) + UI polish for upload area + dark-mode fixes 7af78a0 verified DanielRegaladoCardoso commited on 16 days ago
SQL self-correction: 3 retries with error context + GROUP BY rule in prompt f5b7f67 verified DanielRegaladoCardoso commited on 16 days ago
SQL self-correction: 3 retries with error context + GROUP BY rule in prompt e6681c9 verified DanielRegaladoCardoso commited on 16 days ago
SVG: standalone download (XML prolog, white bg, fixed dims) + Plotly first 0ac4ef2 verified DanielRegaladoCardoso commited on 16 days ago
SVG: standalone download (XML prolog, white bg, fixed dims) + Plotly first 89360e0 verified DanielRegaladoCardoso commited on 16 days ago
Better SQL prompt: column-name mapping rules + few-shot + distinct value hints a7fa6d3 verified DanielRegaladoCardoso commited on 16 days ago
Better SQL prompt: column-name mapping rules + few-shot + distinct value hints b0a300f verified DanielRegaladoCardoso commited on 16 days ago
Pin kaleido==0.2.1 + native SVG fallback when kaleido fails 064c0d2 verified DanielRegaladoCardoso commited on 17 days ago
UX: CSV/SVG download links, schema preview on load, cleaner SVG theme a172c0c verified DanielRegaladoCardoso commited on 17 days ago
Load LoRA via PeftModel on top of standard base models (fixes r=16 vs r=8 mismatch) 1bbdff9 verified DanielRegaladoCardoso commited on 17 days ago
Load LoRA via PeftModel on top of standard base models (fixes r=16 vs r=8 mismatch) eb30a86 verified DanielRegaladoCardoso commited on 17 days ago
Load LoRA via PeftModel on top of standard base models (fixes r=16 vs r=8 mismatch) 730b25d verified DanielRegaladoCardoso commited on 17 days ago
ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU 420b1db verified DanielRegaladoCardoso commited on 17 days ago
ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU 61aee8d verified DanielRegaladoCardoso commited on 17 days ago
ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU a57eca6 verified DanielRegaladoCardoso commited on 17 days ago
ZeroGPU best practice: load models at module level (cuda), inference only inside @spaces.GPU c2ac226 verified DanielRegaladoCardoso commited on 17 days ago
Sequential model load/unload: keep peak VRAM at ~14GB 292fdf4 verified DanielRegaladoCardoso commited on 17 days ago
Sequential model load/unload: keep peak VRAM at ~14GB fab3ba1 verified DanielRegaladoCardoso commited on 17 days ago
Initial deploy: Apple/Claude design, DuckDB, 3 trained LoRAs a067ada verified DanielRegaladoCardoso commited on 17 days ago