Spaces:

Duplicated from DataEyond/Agentic-Service-Data-Eyond

DataEyond
/

Agentic-Service-Data-Eyond-Catalog

Running

App Files Files Community

/feat knowledge management

#20

by rhbt6767 - opened Aug 19

base: refs/heads/main

←

from: refs/pr/20

Discussion Files changed

DataEyond org Aug 19

No description provided.

[NOTICKET] docs: knowledge pipeline test integration — context, audited task checklist, DEV_PLAN §0.88e69dc6c

[NOTICKET] docs: knowledge pipeline — reframe kex as prototype, not port target7cce5ffc

[NOTICKET] feat(eval): knowledge-extraction gold set, scorer + frozen prototype baseline60cf4a88

[NOTICKET] chore: add missing planner and readiness eval resultsbd6659b3

[NOTICKET] feat(knowledge_extraction): free stages — models, seam adapter, filters, cluster, ranking024c30ab

[NOTICKET] fix(knowledge_extraction): move runner into the package; fix two ranking defects72605ff6

[NOTICKET] feat(knowledge_extraction): span filter — E1 recall 0.8537, matches the frozen baseline3a016343

[NOTICKET] feat(knowledge_extraction): paid extraction stage + validate, diff, queueab5ea787

[NOTICKET] eval(knowledge): run the pipeline on MinerU-parsed input — parser comparison6038b2b2

[NOTICKET] feat(knowledge-parsing): document parsing half of the knowledge pipeline5840d201

[NOTICKET] fix(knowledge-parsing): keep the heading line out of Chunk.text15028da4

[NOTICKET] fix(knowledge-parsing): render markup to prose in text; only numbered headings split sections2cb1336a

[NOTICKET] feat(knowledge_extraction): consume the parsing half's ParsedDocumentf132cc1a

[NOTICKET] docs(knowledge-parsing): module README + refresh the parsing/seam statusesf0cf168f

[NOTICKET] fix(knowledge_extraction): show the model the section headinga7bdeae7

[NOTICKET] docs: sync REPO_STATUS §14 + DEV_PLAN §0.8 to the landed knowledge pipeline4cc2edb8

[NOTICKET] eval(knowledge): full document through both halves — E3 0.90 PASS; fix legend blank-line bugec7d3d44

[NOTICKET] docs(knowledge): diagnose the rule-branch regression — three causes, one real1c207dfd

[NOTICKET] docs(knowledge): correct the X21 render_latex diagnosis — three faults, not one7da88fc7

[NOTICKET] fix(knowledge-parsing): send MinerU a backend name it accepts8e93c60c

[NOTICKET] fix(knowledge-parsing): stop the warning marker from failing documentsdf846541

[NOTICKET] fix(knowledge-parsing): render_latex was deleting division (X21)a356e3c0

[NOTICKET] docs(knowledge-parsing): output contract for the parsing halfcb2fd315

[NOTICKET] docs(knowledge-parsing): say what decides a chunk boundary5fed261e

[NOTICKET] fix(knowledge-parsing): read word boundaries back from the source PDFe047f7af

[NOTICKET] fix(knowledge-parsing): let the source prove itself before trusting it22c6ad10

[NOTICKET] docs: reorganise into docs/ and give every doc a status bannerf1342ce0

Merge commit 'refs/pr/20' of https://huggingface.co/spaces/DataEyond/Agentic-Service-Data-Eyond-Catalog into pr/206e78f1e8

[NOTICKET] fix(knowledge): settle the chunk_id separator on `::`e8cc26c5

[NOTICKET] fix(knowledge_extraction): refuse an artifact that will not name its backend4db5f14b

[NOTICKET] feat(knowledge-parsing): carry image descriptions, apart from the textc113bf8b

[NOTICKET] docs(knowledge-parsing): contract 0.3.0 — image kind + generated_descriptionf0e54e98

[NOTICKET] fix(eval/knowledge): stop the scorer treating `_` as a word character (X22)abae3446

[NOTICKET] feat(knowledge-parsing): default to hybrid/high, the measured choice52e1bd4d

[NOTICKET] feat(knowledge-parsing): one runner, in Python rather than shell74acbf72

[NOTICKET] docs(knowledge-parsing): how to run it, and how to read what comes backb3eae0f1

[NOTICKET] refactor(knowledge-parsing): English throughout the module7f153d60

[NOTICKET] fix(knowledge_extraction): rule.txt excludes bare equations (X20)296cdd71

[NOTICKET] docs(knowledge): X20 landed unevaluated; X22 resolved; X23 openede84487d7

[NOTICKET] docs(knowledge-parsing): record the backend measurement, and fix three stale factsb0cc5078

[NOTICKET] feat(knowledge_extraction): output schema v2 — entity ids, links, 16 fields cut49dfe4d7

[NOTICKET] docs(knowledge-parsing): settle determinism; finish the English passd5ed50cd

[NOTICKET] fix(knowledge_extraction): formula_latex is checked against Chunk.latex9c8a6e41

[NOTICKET] docs(knowledge-parsing): B1 — cost justification for the BUMA corpus609e7ba1

[NOTICKET] fix(knowledge_extraction): compare formula_latex by notation, not identitya998ea48

[NOTICKET] feat(knowledge_extraction): record which guarantee formula_latex carriesf81a227f

[NOTICKET] docs(knowledge-parsing): B1 — Colab, and the two decisions it forces2d2f9058

[NOTICKET] fix(knowledge_extraction): rule.txt few-shots stop contaminating the gold set (X23)7e8614fc

[NOTICKET] docs(knowledge-parsing): record that Chunk.latex is now load-bearinga0df6d36

[NOTICKET] eval(knowledge): the runner the harness has been missing1edf61ce

[NOTICKET] fix(knowledge_extraction): close the notation rules against real backend markup01c76ddc

[NOTICKET] fix(knowledge_extraction): the latex guard was rejecting correct transcriptions22eb3d40

[NOTICKET] fix(knowledge_extraction): audit the latex normaliser for overfitting7d6dedfa

[NOTICKET] eval(knowledge): rule-branch noise band — X20 confirmed, X23 answeredf7546310

[NOTICKET] fix(knowledge-parsing): normalize.py silently dropped every list block6edfbe7b

[NOTICKET] docs(knowledge-parsing): record B3 — MinerU reaches parity, and whye65e6858

[NOTICKET] docs(knowledge): correct the rule-ceiling diagnosis — it is not tables892f3876

[NOTICKET] feat(knowledge_extraction): cue class for declarative composition (X24b cause 3)e6d9d29a

[NOTICKET] eval(knowledge): X20 measures net -1; close X23 on expert ruling; instrument X25e6ed4146

[NOTICKET] revert(knowledge_extraction): X20 — it measured a net -1 (X20 closed)21640d63

[NOTICKET] fix(knowledge-parsing): long names, name clashes, and a GPU check in the wrong place9083dd1a

[NOTICKET] docs(knowledge-parsing): naming convention, and why it is a stopgap942f14e1

[NOTICKET] eval(knowledge): commit the BUMA fixture artifact; K9/K10 closee5f4abea

[NOTICKET] feat(knowledge-parsing): stamp the artifact with the code that built itc64a0cb8

[NOTICKET] feat(knowledge-parsing): commit the reference artifact, and stop the stamp churninga0e59e4d

[NOTICKET] feat(knowledge-parsing): carry the document's human-readable titledf571bec

[NOTICKET] refactor(knowledge-extraction): BriefContext -> DomainContext7f47c618

[NOTICKET] feat(knowledge-extraction): derived outline + corroborated key parameters6f5d6195

[NOTICKET] feat(knowledge-extraction): span-guard the domain branche2865ac3

[NOTICKET] feat(knowledge-extraction): subdomains + counts — v3 complete172d8dc7

[NOTICKET] docs(knowledge): record the DomainContext v3 reshape6a9fb784

[NOTICKET] docs(knowledge): v3 delta in the workstream docs58e20027

[NOTICKET] feat(knowledge-parsing): Mistral OCR block normalizer (section default)81b4f4ee

[NOTICKET] feat(knowledge-parsing): wire Mistral OCR backend end-to-end6856ea8d

[NOTICKET] feat(knowledge-parsing): add 1-based page_no (S6b)717452f6

[NOTICKET] docs(knowledge): reflect Mistral OCR as the primary parser854d3732

[NOTICKET] feat(knowledge-eval): freeze the Mistral baseline (M1)a358448d

[NOTICKET] feat(knowledge-extraction): carry page_no across the seam (S6b)0466f67a

[NOTICKET] docs(knowledge): surface plan, persistence contract, DomainContext v3ef8ee47e

[NOTICKET] eval(knowledge): commit the untracked result files25e7e7a9

[NOTICKET] docs(knowledge): reframe as a Product-B experiment; persistence contract v20c20b40c

[NOTICKET] docs(knowledge): finalise the DDL handoff (stage A)c3db7d0b

[NOTICKET] feat(knowledge-ingest): job lifecycle + ingestion routes (N10)b14034b4

[NOTICKET] docs(knowledge): reflect the shipped ingestion slice in the workstream48e558b4

[NOTICKET] feat(knowledge): seam version gate (N6b) + ingest auth posture (N11)67af8187

[NOTICKET] build(deps): drop mineru, pin torch to the CPU wheel (N7b)26ba1ab7

[NOTICKET] docs(knowledge): MinerU is dropped, not demoted — sync every status claim6281877b

[NOTICKET] feat(knowledge-parsing): parse_one() single-doc entry point + doc_id param (N9)7a447359

[NOTICKET] feat(knowledge-parsing): read MISTRAL_API_KEY through app settings85c9c289

[NOTICKET] feat(knowledge): create the tables and wire the pipeline end to endc2164ae5

[NOTICKET] feat(knowledge): read + review routes, closing the loop (N10b, N12b)f1188789

[NOTICKET] docs(knowledge): sync the canonical docs with the shipped surface840b74d9

[NOTICKET] fix(knowledge): the diff baseline never reached the ingest path65f84bf1

[NOTICKET] docs(knowledge): plan for how agents consume the entries (N14)0b1bd45f

[NOTICKET] docs(knowledge): one name for the rules; correct the stale v3 banner73495cb8

[NOTICKET] docs(knowledge): plan a real domain context (v4)5e6400c0

[NOTICKET] feat(knowledge): kind CHECK gains 'document'; finalise the v4 workstream1b8660e7

[NOTICKET] feat(knowledge-domain): compose a real domain context (V4 phase 2)e43be1c9

[NOTICKET] fix(knowledge-extraction): parenthetical abbreviations split one concept in twod19a981e

[NOTICKET] feat(knowledge): DocumentBrief split + cross-document measure identityb5de3b4e

[NOTICKET] fix(knowledge): entry-id collisions and orphaned run rows25aafc04

[NOTICKET] docs(knowledge): mistralai cannot install in the Space (found by the e2e run)b39e6b9b

[NOTICKET] feat(knowledge-domain): KnowledgeService — the interface agents read through (K1-K4)ae0e4560

[NOTICKET] fix(knowledge-domain): a domain entry was being read as a document brief1160b03b

[NOTICKET] feat(knowledge-domain): the declared half — domain TBox + review (V4-9, V4-10)601937c7

[NOTICKET] feat(knowledge): wire knowledge into the prompts and traceability (K5, K9, K11)037efcf0

[NOTICKET] feat(knowledge): arm the consumption path end to end (K5, K6, K9, K11)a6bed7f5

[NOTICKET] build(deps): mistralai into main deps, pydantic pinned forwardcafa1543

sofhiaazzhr changed pull request status to open Sep 9

ishaq101 changed pull request status to merged Sep 9

Upload images, audio, and videos by dragging in the text input, pasting, or clicking here.

Tap or paste here to upload images

· Sign up or log in to comment