cbcae69abf
feat(embed): candle 모델 레지스트리 + arctic-embed-l-v2.0 (CLS pooling)
...
e5 하드코딩(HF_MODEL/SUPPORTED_MODEL/mean/query:+passage:) → 모델 레지스트리
EmbedModelSpec{name,hf_repo,pooling,query_prefix,doc_prefix,dim,version_tag}.
e5(mean, query:/passage:) + arctic(CLS, query:/무접두어). pooling 모델별 분기
(mean=attention-mask-weighted / CLS=hidden[:,0,:]), tokenize/forward/L2 공유.
arctic pooling=CLS 는 HF 1_Pooling/config.json(pooling_mode_cls_token:true) 확인.
model_version 은 arctic 일 때 +arctic-cls 태그(embedding_version cascade 트리거);
e5 는 fastembed-e5 호환(NUMA 드롭인) 위해 plain config.version 유지.
correctness 게이트: tests/arctic_ollama_parity.rs (#[ignore], live Ollama) —
candle arctic vs Ollama snowflake-arctic-embed2 per-sentence 코사인>0.99.
수동 실측 cosine_min=0.999984 (recall@10 130 재현 보장).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com >
2026-06-03 04:59:11 +00:00
369aeb3d24
feat(embed): candle Metal (Apple Silicon GPU) opt-in build feature + v0.23.0
...
- kebab-embed-candle: `metal` feature → candle metal backend; select_device()
picks Device::new_metal(0) (CPU fallback) under the feature, else Device::Cpu.
.contiguous() before to_vec2 (Metal rejects strided views; CPU tolerates).
- feature passthrough: kebab-app/embed_metal → kebab-cli/embed_metal.
Build on macOS: cargo build --release --features embed_metal.
- default (non-metal) path unchanged: clippy 0, candle units + thread_cap + parity pass.
- README + HOTFIXES: Mac-GPU-ingest → copy sqlite+lancedb → server CPU-query workflow.
- version 0.22.0 → 0.23.0 (opt-in build surface).
macOS-only compile; Metal execution/speed/parity validated by user on M4 Pro
(not buildable on the Linux CI/dev machine).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com >
2026-06-02 11:37:08 +00:00
8f7b6ee538
feat(embed): candle 임베딩 provider (NUMA-안전, opt-in) + v0.22.0
...
duo-socket NUMA 서버에서 fastembed(onnxruntime)가 intra-op 스레드를 48개로
하드코딩해 NUMA 힙 손상 → double-free 로 ingest 가 죽는 문제를 회피하기 위해,
같은 multilingual-e5-large 모델을 순수 Rust(candle)로 돌리는 opt-in 임베딩
provider 를 추가한다.
- 신규 crate kebab-embed-candle: CandleEmbedder (kebab_core::Embedder).
hf-hub safetensors → XLMRobertaModel forward → mask mean-pool → L2 → e5
prefix. candle 의존성 트리를 이 crate 에 격리 (core/config 외 kebab-* 의존 0).
- 스레드 캡: [models.embedding].num_threads + env KEBAB_EMBED_THREADS →
글로벌 rayon 풀 1회 캡 (NUMA-안전 레버).
- kebab-app::embedder() 가 provider 분기 (fastembed/onnx/"" → 기존 경로 불변,
candle → CandleEmbedder, 미지값 → 에러).
- Phase 0 스파이크 crate 제거 (production 흡수).
- 버전 0.21.1 → 0.22.0 (신규 config surface, pre-1.0 minor bump).
패리티: cosine_min=1.000000, max abs diff=2.01e-7 (< 1e-5) → embedding_version
유지, 재색인 0. fastembed default 동작/벡터 불변. wire schema 변경 없음.
검증(파일+exit code): clippy -D warnings EXIT=0(warning 0), test EXIT=0
(candle unit 5 + thread_cap rayon=4 + config 68), parity #[ignore] EXIT=0.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com >
2026-06-01 14:52:25 +00:00
