janitr
Local runnable models for filtering out crypto shill content, AI replies, and other undesirable content
파일 탐색기
- SKILL.md
- SKILL.md
- ci.yml
- pre-commit
- logo.svg
- thresholds.json
- thresholds.transformer.json
- holdout.jsonl
- holdout.txt
- train.jsonl
- train.txt
- valid.jsonl
- valid.txt
- accounts.jsonl
- ai-reply-attractor-accounts.json
- ai-reply-taggers.jsonl
- calib_rebalanced.meta.json
- calib_rebalanced_hardneg.meta.json
- calib_rebalanced_hardneg600.meta.json
- google_search_results.md
- label_changes_2026-02-02.txt
- levelsio-callouts-fetched.jsonl
- levelsio-callouts-raw.jsonl
- levelsio-deep-callouts.jsonl
- levelsio-flagged-ai-reply-bots.jsonl
- levelsio-pre-dec13-callouts.jsonl
- other-taggers-callouts-raw.jsonl
- other-taggers-deep-callouts.jsonl
- other-taggers-deep-fetched.jsonl
- other-taggers-fetched.jsonl
- replies.jsonl
- replies_inferred.jsonl
- sample.jsonl
- sample.jsonl.bak
- time_split_meta.json
- val.txt
- calib.ids.txt
- holdout.ids.txt
- leakage_report_current.json
- leakage_report_vs_previous.json
- MANIFEST.json
- train.ids.txt
- valid.ids.txt
- README.md
- test.jsonl
- train.jsonl
- 2026-02-04.md
- 2026-02-07.md
- 2026-02-08.md
- 2026-02-11.md
- 2026-02-12.md
- 2026-02-13.md
- 2026-02-14.md
- 2026-02-15.md
- 2026-02-13-tiny-transformer-plan.md
- 2026-02-11-results.csv
- 2026-02-11-summary.json
- 2026-02-11-results.csv
- 2026-02-11-summary.json
- 2026-02-11-train_stage1_scam_vs_not.txt
- 2026-02-11-train_stage2_topic_vs_clean.txt
- 2026-02-11-results.csv
- 2026-02-11-class_specific_threshold_sweep.csv
- 2026-02-11-class_specific_threshold_sweep_best.json
- 2026-02-11-error_mining_holdout_class_specific_best.json
- 2026-02-11-error_mining_holdout_class_specific_best.txt
- 2026-02-11-eval_holdout_calib_2pct_all.txt
- 2026-02-11-eval_holdout_calib_class_specific_best.txt
- 2026-02-11-eval_holdout_current_shipped.txt
- 2026-02-11-thresholds_calib_2pct_all.json
- 2026-02-11-thresholds_calib_class_specific_best.json
- 2026-02-11-thresholds_current_shipped.json
- 2026-02-11-tune_calib_2pct_all.txt
- SUMMARY_AGGREGATED.json
- 2026-02-11-bakeoff_summary.json
- 2026-02-11-eval_extension_model_holdout.txt
- 2026-02-11-eval_extension_model_valid.txt
- 2026-02-11-eval_extension_two_stage.json
- 2026-02-11-single_stage_candidates.csv
- 2026-02-11-single_stage_summary.csv
- 2026-02-11-two_stage_candidates.csv
- 2026-02-11-two_stage_quant_sweep_softmax.json
- 2026-02-11-two_stage_ship_eval.json
- 2026-02-11-two_stage_softmax_eval.json
- 2026-02-11-two_stage_summary.csv
- 2026-02-13-tiny-transformer-eval-report.md
- 2026-02-14-frozen-split-fasttext-vs-transformer-benchmark.md
- 2026-02-14-tiny-transformer-expanded-holdout-retrain-report.md
- 2026-02-11-ai-generated-replies-2026.md
- x-post-snapshot.schema.json
- 2026-02-11-multi-label-scaling-plan.md
- ACCOUNTS.md
- AI_REPLY_SCRAPING.md
- ARCHITECTURE.md
- CALIBRATION_FIX_REPORT.md
- DATA_LABELING.md
- DATA_MIXING.md
- DATA_MODEL.md
- DATA_SOURCES.md
- EVAL_RESULTS.md
- LABELS.md
- MODEL.md
- PLAN.md
- QUANTIZATION.md
- REPLIES_DATA_MODEL.md
- ROADMAP.md
- SCRAPE_TARGETS.md
- SCRAPING.md
- SOURCES.md
- SYNDICATION_API.md
- TRAINING_INFERENCE_STACK.md
- WASM_PLAN.md
- classifier.js
- fasttext.js
- fasttext_wasm.js
- fasttext_wasm.wasm
- model.ftz
- thresholds.json
- build.mjs
- classifier.ts
- fasttext.ts
- transformer-smoke.ts
- wasm-smoke.ts
- classifier-transformer.ts
- model-repo.ts
- background.ts
- content-script.ts
- globals.d.ts
- offscreen.ts
- options.ts
- popup.ts
- types.ts
- transformer-smoke.html
- transformer-smoke.js
- transformer-smoke.spec.ts
- wasm-smoke.html
- wasm-smoke.js
- wasm-smoke.spec.ts
- classifier-transformer.js
- model-repo.js
- fastText.common.d.ts
- fastText.common.js
- fastText.common.wasm
- fastText.d.ts
- fastText.node.d.ts
- fastText.node.js
- fastText.node.wasm
- lid.176.ftz
- fastText.common.wasm
- common.cjs
- common.d.ts
- common.mjs
- index.cjs
- index.d.ts
- index.mjs
- node.cjs
- node.d.ts
- node.mjs
- types.d.ts
- common.cjs
- common.d.ts
- common.mjs
- node.cjs
- node.d.ts
- node.mjs
- languages.json
- languages.json.cjs
- languages.json.d.ts
- languages.json.mjs
- lid.176.ftz
- base.cjs
- base.d.ts
- base.mjs
- common.cjs
- common.d.ts
- common.mjs
- node.cjs
- node.d.ts
- node.mjs
- types.d.ts
- misc.d.ts
- constants.cjs
- constants.d.ts
- constants.mjs
- FastText.cjs
- FastText.d.ts
- FastText.mjs
- FastTextModel.cjs
- FastTextModel.d.ts
- FastTextModel.mjs
- .gitkeep
- ort-wasm-simd-threaded.mjs
- ort.wasm.min.mjs
- background.js
- content-script.js
- icon128.png
- icon16.png
- icon32.png
- icon48.png
- manifest.json
- offscreen.html
- offscreen.js
- options.css
- options.html
- options.js
- popup.css
- popup.html
- popup.js
- tsconfig.json
- build_rebalanced_calib.py
- build_unlabeled_corpus.py
- cache_teacher_logits.py
- calibrate_teacher.py
- check_integrity.py
- check_reply_integrity.py
- check_split_drift.py
- check_split_leakage.py
- compare_models_fpr.py
- evaluate.py
- evaluate_transformer.py
- experiment_artifacts_model.py
- export_transformer_student_onnx.py
- fetch_callout_threads.py
- fetch_levelsio_remaining.py
- fetch_other_taggers.py
- fetch_other_taggers_deep.py
- fix_duplicate_ids.py
- inference.py
- labelset.py
- make_holdout.py
- make_stratified_splits.py
- make_time_splits.py
- manual_relabel.py
- mine_errors.py
- mine_hard_negatives_txt.py
- prepare_data.py
- prepare_hf.py
- prepare_transformer_data.py
- pyproject.toml
- quantize_transformer_student.py
- rebuild_experiment_index.py
- reduce_fasttext.py
- relabel_crypto.py
- run_naming.py
- scrape_levelsio_deep.py
- student_runtime.py
- sync_datasets_to_experiments_repo.py
- sync_experiments_from_remote.sh
- sync_to_experiments_repo.py
- train_fasttext.py
- train_teacher_dapt.py
- train_transformer_student_distill.py
- train_transformer_teacher.py
- transformer_common.py
- tune_threshold.py
- tune_thresholds_fpr.py
- unify_replies.py
- uv.lock
- validate.py
- validate_accounts.py
- validate_data.sh
- validate_experiment_runs.py
- wasm-smoke.mjs
- index.html
- logo.svg
- .gitignore
- .prettierignore
- AGENTS.md
- LICENSE
- Makefile
- mise.toml
- package.json
- pnpm-lock.yaml
- README.md
- wrangler.toml
# CDN으로 사용하기
jsDelivrjsDelivr는 공개 GitHub 리포지토리를 별도 설정 없이 CDN으로 즉시 서빙합니다. 버전과 파일을 고르면 웹페이지에 바로 붙일 수 있는 링크와 예시 코드가 만들어집니다.
명령어 용어집
이 문서에서 사용된 명령어를 모아봤습니다. 낯선 명령어가 있다면 펼쳐서 확인해보세요.
corepack
설명 보기 ▼
corepack
Node 프로젝트와 해당 패키지 관리자 사이의 브라지 역할을 하는 런타임 종속성이 없는 패키지.
더 많은 정보: <https://github.com/nodejs/corepack>.
corepack enable
Corepack shim을 Node.js 설치 디렉터리에 추가하여 전역 명령으로 사용할 수 있도록 함:
corepack enable --install-directory {{path/to/directory}}
특정 디렉토리에 Corepack shim을 추가함:
corepack disable
Node.js 설치 디렉터리에서 Corepack shim을 제거:
pnpm
설명 보기 ▼
pnpm
빠르고, 디스크 공간 효율적인 Node.js용 패키지 관리자.
Node.js 프로젝트 및 해당 모듈 의존성 관리.
더 많은 정보: <https://pnpm.io/pnpm-cli>.
pnpm init
`package.json` 파일 생성:
pnpm {{[i|install]}}
`package.json`에 의존성으로 나열된 모든 패키지를 다운로드:
pnpm add {{module_name}}@{{version}}
특정 버전의 패키지를 다운로드하여 `package.json`의 의존성 목록에 추가:
uv run
설명 보기 ▼
uv run
Run a command or script in the project environment.
uv run {{path/to/script.py}}
Run a Python script:
uv run {{[-m|--module]}} {{module_name}}
Run a Python module:
uv run {{[-w|--with]}} {{package}} {{command}}
Run a command with additional packages installed temporarily:
