Files
stack/src/llm/__init__.py
kert 5f446e3804
Some checks failed
CI / lint (push) Successful in 33s
CI / notebooks-smoke (push) Successful in 1m45s
Deploy / notebooks (push) Successful in 6m3s
Deploy / zotero (push) Has been skipped
Deploy / docs (push) Has been skipped
Deploy / api (push) Has been skipped
Deploy / llm (push) Has been skipped
Deploy / mc (push) Has been skipped
Infra CI / notebooks (push) Successful in 1m2s
Infra CI / zotero (push) Successful in 14s
Infra CI / docs (push) Successful in 27s
Infra CI / api (push) Successful in 1m7s
Infra CI / llm (push) Successful in 48s
Infra CI / mc (push) Successful in 13s
Deploy / report (push) Successful in 15s
CI / test (push) Failing after 21m36s
fix(llm,notebooks): code_families sections 5/7b run in the notebooks container — lazy pool/sqlalchemy imports in llm, guarded cells, LLM_DB_PASSWORD + psycopg for the notebooks service (refs #720)
2026-09-11 10:30:45 -04:00

40 lines
1.4 KiB
Python

"""llm: local RAG over rulemaking comments + the Zotero/bib corpus.
Embeds regulations.gov public comments (``doctype:comment`` in the bib
store) and the reference corpus into pgvector, and (P34/P35) exposes
retrieval, grounded generation, and closed-vocab tagging — all inference
on local Ollama hosts, never cloud APIs.
Architecture::
.state/comments/*.md + bib corpus (llm.source)
-> markdown-aware chunks (llm.chunk)
-> Ollama /api/embed, multi-host (llm.pool)
-> pgvector collections (llm.index)
Spec: docs/superpowers/specs/2026-07-16-llm-module-design.md
"""
from __future__ import annotations
from typing import Any
from llm.config import LlmConfig as LlmConfig
from llm.config import load as load
from llm.config import pg_url as pg_url
# The pool pulls in langchain; ``llm.config``, ``llm.links`` and
# ``llm.lineage`` do not need it and are imported by the notebooks
# container, which installs the workspace without the ``llm`` extra
# (#720). Resolve the pool exports lazily so ``import llm.config`` never
# imports langchain.
__all__ = ["HostPool", "LlmConfig", "PoolEmbeddings", "load", "pg_url"]
def __getattr__(name: str) -> Any:
if name in {"HostPool", "PoolEmbeddings"}:
from llm import pool
return getattr(pool, name)
raise AttributeError(f"module 'llm' has no attribute {name!r}")