Bake direction (kn → en / en → kn) into dataset.name so HF sidebar clearly distinguishes them — fix apparent-duplicate UX

Fix model-index: add dataset.config per direction to prevent HF deduplication of kn↔en entries

Expand model-index: split FLORES kn↔en, add IN-22 conv (4 metrics × 2 directions) + Entity Preservation NER. 5 task entries total, ranked metrics: bleu/chrf/comet/cometkiwi/accuracy

SDK 0.1.4: client.py edits + version bump

docs: re-push what-didnt-work.md with edits

docs: drop §-symbol PyPI mishap from what-didnt-work (not project-relevant)

DEPLOYMENT.md §9: add distillation-to-decoder-only path as v3.0+ option for GGUF/Ollama/vLLM compatibility

docs: move to release/docs/, neutralize internal-doc refs, add prerequisites + data source links + cost estimate + first-week checklist + architecture diagram, rewrite working-with-claude opening in natural voice

docs: add transliteration fallback, currency/dates/acronyms edge-cases, tech generation, 3 modes, 7 eval dimensions, curriculum schedule

docs: rewrite with grounded details from archive (FP16/bf16, Falklands tokenizer bug, OOM crisis at step 8500, eval-cap fix, v2.1 decoder pollution, v2.2 from-scratch decision, Rule Zero)

Model card: link docs/ folder (journey, lessons, recipes)

Add /docs journey: README + the-journey + what-didnt-work + how-it-was-built + working-with-claude + repo-map + top-lessons

v2.4 priorities locked: multi-token code-mix data slice (#1) + proverbs corpus

SDK 0.1.3 + CPU torch install hint in model card + Documentation URL on PyPI

SDK 0.1.2: detect_libraries() now probes sentencepiece + huggingface_hub

Replace stray § symbol with "Section " (model card, DEPLOYMENT, SDK 0.1.1)

SDK source: assets/sdk/ (8 files; pip install controlmt)

Model card: Available releases section linking int8 repo + SDK + demo

Sync space/static/ assets with deployed Space (elapsed counter)

Model card §8: verified deployment matrix + pinned versions + honest not-supported list

Verified deployment matrix + pinned versions + assets/space + assets/scripts + bench JSONs

Update PRIVACY.md: share toggle is default-on with prominent Settings placement (uncheck to opt out)

Add Break-the-Model Challenge invite to model card + PRIVACY.md with full disclosure of opt-in research dataset

Add Break-the-Model Challenge invite to model card + PRIVACY.md with full disclosure of opt-in research dataset

Add IN22-Conv scores to model card §4.4 — all floors cleared

Add IN22-Conv v2.3 eval results: COMET-DA 0.8232/0.8320, CometKiwi 0.8143/0.8845 (clears all floors)

Move full deployment guide to DEPLOYMENT.md; model card carries only a compact reference

Move full deployment guide to DEPLOYMENT.md; model card carries only a compact reference

Tone down NLLB claim + soften 100% claim + split detail into TRAINING_GUIDE.md + add citation-grounded comparison context

Tone down NLLB claim + soften 100% claim + split detail into TRAINING_GUIDE.md + add citation-grounded comparison context

Add Section 8 Deployment Guide: matrix + recipes for local, HF Endpoints, Spaces, FastAPI, Docker, ONNX, quantization, batched inference

Fix dotted-repo-name import: importlib fallback when relative import breaks on controlmt-v2.3

Expand training data section: public corpora citations + internal-generated specialized streams + scoring/eval set references

Update docstring URLs to new v2.3 repo name

Update docstring URLs to new v2.3 repo name

Reframe form-data section: lead with 100%-preservation capability, EN→KN PAN as minor edge case

v2.3 release — single-register retrain, FLORES BLEU 27.20/18.50, COMET 0.8459/0.8443; style endpoints hidden from API

Scrub internal CONTROLMT.md cross-references from CHANGELOG.md

Scrub internal CONTROLMT.md cross-references from README.md

Revert AutoTokenizer auto_map in config.json to string form (silence linter warning); tokenizer_config.json keeps list form for runtime

Fix relative import of model.py for trust_remote_code auto-download

Fix HF auto_map (add tokenizer_config.json + list form); honest style framing + add curated eval scores

Replace section sign (§) with spelled-out "Section" in README + CHANGELOG

ControlMT v2.2 initial release

initial commit