docs: add presentation deck resources

docs: add presentation deck link

docs: add presentation deck link

chore: remove frontmatter injection from deploy workflow, add frontmatter directly to README

chore: remove redundant streamlit install in devcontainer, fix stale README wording

ci: inject HF Spaces frontmatter into README at deploy time

docs: remove HF Spaces frontmatter from README (rendered as table on GitHub)

chore: rename HF owner HlexNC → arudaev across all references

docs: document CLAHE, label smoothing, TTA, and ensemble in hub.py, CLAUDE.md, README

feat: add CLAHE, label smoothing, TTA, and ensemble inference

fix: update model card templates and diagram PNGs for 320px dataset

docs: update README and CLAUDE.md; add Mermaid diagrams to README

fix(types): resolve all 8 mypy errors across 5 files

Add security policy (GitHub private advisories)

chore: disable Claude attribution in commits and PRs

Add resize_320 Kaggle pipeline script and tests

fix(lint): remove unused base64 and time imports in generate_diagram_pngs.py

Switch training dataset to HlexNC/chest-xray-14-320 (320px)

Fix model card image URLs to use absolute resolve/main/ paths

Add diagram PNG generation script and fix pipeline Mermaid source

Fix app: restore sidebar sample X-rays without binary assets

fix(ci): align dispatch tests with CPU-only resize_320 kernel

fix(resize_320): build and dispatch the raw-source Kaggle pipeline

Fix app: default use_se=False so v1 checkpoints load without SE keys

Fix model card dataset revision display to include full hash

Fix app checkpoint-aware model instantiation; dual Kaggle token paths

v2: SE attention, AMP, grad accumulation, rich model cards with Mermaid

Pin torch==2.3.1+cu118 in Kaggle scripts for P100 (sm_60) compatibility

Don't reinstall torch in Kaggle scripts — use container's GPU-compatible version

Fix doubled images/images/ path in trainer dataset construction

Fix configure_hf_runtime mkdir in non-Kaggle environments

Update README and CLAUDE.md to reflect current project state

Add HF hub utilities, fix Kaggle secret injection via dataset source

Align Kaggle kernel slugs with live notebook refs

Handle KGAT Kaggle tokens in dispatch CLI

Restore src/data package, fix HF Space bootstrap, improve Kaggle/CI wiring

Added Dev Container Folder

Add test coverage for dataset encoding, metrics, and transforms

Fix CI, unblock Kaggle training dispatch, upgrade demo app for HF Spaces

Replace notebooks with scripts + Kaggle dispatch, add HF data pipeline Space

Enforce cloud-only compute policy, populate HF dataset, Colab notebooks

Initial project scaffold: CheXVision — Chest X-Ray Pathology Detection