Spaces:
Running on Zero
Running on Zero
Commit History
4 days ago
-
Align the confusion matrix with the training history chart
Move the confusion matrix title out of its framed box into the same section heading the training history uses, so both panels in the top row of the Results page start at the same height.adaf2db -
Align Results tables and fix overlapping table headers
Lay out the Results page as two rows (charts, then tables) so the architecture audit and pathway attention tables start at the same height. Give pinned table headers an opaque background so scrolled rows no longer show through them.c5021ff -
Scroll long result tables inside their own box
Cap every results and data table at 320 px so long outputs (validation predictions, scored cohort, pathway attention, enrichment) scroll inside the table instead of stretching the page.e413269 -
Keep the action bar visible without scrolling
Inside the Hugging Face Space frame, each page's action bar (Build biological architecture, Train, Run inference) sat below the fold on 1440x900 and 1280x720 screens, so users could miss it. Pin the bar to the bottom of the visible frame with position: sticky. Switch Gradio's container from overflow: hidden, which disables sticky positioning, to overflow: clip.bb42e34 -
Add BRCA training diagnostics and lab grid results
Scripts and raw results for the 2026-10-16 collapse diagnosis (weight decay, seed and epoch-limit experiments) and the 2026-10-30 run of the lab's hyperparameter grid with an 8-seed check. Each folder has a README explaining how to reproduce it.d1ec2ee -
Fix BRCA training collapse with lab-grid defaults
Training on BRCA with the previous defaults (Adam, lr 1e-3, weight decay 0.01) collapsed: every sample got the same prediction (macro F1 0.133). - Default to the configuration from the Bozdag Lab grid that was stable across 8 seeds on BRCA: Adam, lr 0.01, L2 0.01, batch 32, projection 64, tanh/relu/tanh activations, 200 epochs (macro F1 0.749 +/- 0.018). - Make the three activation functions and SGD momentum (default 0.03) configurable in Hyperparameters. model.py and the artifact format are unchanged; artifacts already store their activations. - Make the automatic GenePT choice the lab-recommended single-cell file. - Default to a 75/25 stratified train/validation split, with helper text and a clear error when a class is too small to stratify. - Fix the Results page training-history plot, which was never populated. - Add stable element ids for e2e tests and a layout tweak for short desktop screens. - Tests: split, leakage, class weights, history plot, activations, moment…62a8628
-
Fix mobile dashboard layout
9a10d98
-
Redesign workbench dashboard shell (#1)
- Redesign workbench dashboard shell (72c0d0ba7a7859042da2633a5c048ec50f59cd87) - Fix dataset loading and add intro cards (121eaa8832a7dfe6e03d01b13909bd13d51934fe) Co-authored-by: Faruk Yildirim <F4RUKYLDRM@users.noreply.huggingface.co>2e6a2a4
-
Add ZeroGPU-enabled BioLM-NET workbench
Verified signature31376a7 -
initial commit
Verified signatureabe7732