a3bc98a53c32f3b546f6017716d26dedc44d0e07
97 Commits
| Author | SHA1 | Message | Date | |
|---|---|---|---|---|
 bvandeusen
|
a3bc98a53c |
feat(translation): Post translation columns + settings + migration (#143 step 1)
Post gains post_title_translated / description_translated / translated_source_lang / translation_engine_version / translated_at — filled by the translate sweep so viewing is instant. ImportSettings gains translation_enabled (OFF by default), interpreter_base_url (EMPTY — no default host; the operator points it at their own Interpreter proxy behind a reverse proxy) and translation_target_lang (en), exposed + validated via /settings/import. Migration 0083. Settings defaults + patch + validation test. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
ab63d94249 |
feat(ml): presentation auto-hide settings + review table (#141 step 3)
MLSettings gains presentation_auto_apply_enabled / _threshold (default 0.90) + presentation_conflict_threshold (default 0.50): banner/editor auto-hide with a FLAT threshold (decoupled from content-head graduation), plus the "also looks like content" conflict cut. New presentation_review table (image, presentation tag, conflict tag + score, created/resolved_at) records auto-hides flagged for review. Migration 0082 (columns + table), ml_admin API (editable + get_settings + _validate bounds), settings roundtrip/bounds test. The sweep that reads these knobs + the Settings UI land in step 4. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
7d3a3b4a83 |
revert(ml): keep head auto-apply precision at 0.97 (operator: general tuning was fine)
Milestone 139 raised head_auto_apply_precision 0.97→0.98; operator confirmed the general-tag confidence was already well tuned, so revert that. The support floor (min_positives 30→50) and CCIP match confidence (0.92→0.95) stay. Migration 0081 (not yet deployed) edited to drop the precision bump. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
cbc3e11a53 |
feat(ml): stricter auto-apply defaults to cut misfires (milestone 139)
head_auto_apply_precision 0.97→0.98, head_auto_apply_min_positives 30→50, ccip_auto_apply_threshold 0.92→0.95 (operator-asked). Model defaults change for fresh installs; migration 0081 bumps the existing singleton row IFF still at the old default (won't clobber a deliberate operator change). ml_admin bounds already permit these. Fixed a stale comment in the auto-apply test. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
2cfbb284d5 |
feat(heads): incremental retraining — refit only changed tags (#1317 phase 2, m138)
train_all_heads is now incremental by default: a per-tag training-data fingerprint (positive + rejection count/latest-timestamp, stored on tag_head.train_fingerprint) means a manual Retrain refits ONLY the tags whose data changed — O(what you touched), not O(all heads). The nightly scheduled_train_heads passes full=True to reconcile sampled-negative + hygiene drift across every head. First incremental run after deploy still refits everyone (NULL fingerprints), stamping them, then it's incremental. The refit decision + fingerprint are split into sklearn-free helpers (_head_fingerprints, _heads_needing_retrain) so the incremental logic is unit-tested directly (train_head itself needs scikit-learn). Migration 0080. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
f24dc81764 |
feat(ccip): schema for precomputed incremental character prototypes (#1317, m138 step 1)
Foundation for making CCIP character references a precomputed, INCREMENTAL artifact instead of a request-path rebuild (kills the per-accept ~4s suggestions stall; cost will scale with change, not library size): - character_prototype: a character's reference CCIP vectors, capped to MLSettings.ccip_prototype_cap so match cost doesn't grow with popularity. - ccip_prototype_state: per-character fingerprint (ref count + max region id) + updated_at → drives per-character incremental rebuilds and the matcher cache's reload-only-what-advanced. - MLSettings.ccip_ref_signature (cheap global change gate) + ccip_prototype_cap. Migration 0079. Schema + models only — the builder service, refresh task/beat, and matcher rewrite land in the following steps. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
62ec70b9e4 |
feat(ml): detector config in MLSettings with working defaults (#134 step 1)
Move the crop-proposer config (per-proposer enable + weights + conf, caps, dedupe IoU) into the DB so it's UI-tunable and can be announced to the GPU agent in the lease (like the embedder model) — no restart, agent env becomes bootstrap-only. Migration 0078 adds the columns with working server_defaults so existing rows + fresh installs crop out-of-the-box with all three proposers ON (operator: default-on): person=yolo11n.pt, anatomy=booru_yolo yolov11m_aa22 (URL, license unstated/private-homelab-OK), panel=mosesb best.pt. Plain columns, no CHECK enum. Steps 2 (lease announce + agent apply) and 3 (Settings UI) follow. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
87d53db0cb |
feat(artist): editable display name + rename surface; drop name-uniqueness (#130 step 1)
First step of decoupling artist identity/storage/display. migration 0077 drops uq_artist_name so the display name is free text (two genuinely different creators can share a name); the slug stays the immutable, unique storage/identity key (the on-disk path component — untouched, so nothing moves). ArtistService.rename + PATCH /api/artists/<id> change the name ONLY. Frontend: inline pencil-edit on the artist header (mirrors TagCard), slug/route unaffected so no navigation. Fixes the operator's 'no surface to rename an artist' + the name-collision fragility. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
0563b2d750 |
feat(pixiv): ledger models + migration 0076 + PixivIngester adapter (#129 step 3)
pixiv_seen_media / pixiv_failed_media mirror the Patreon/SubscribeStar ledgers (keys are always synthesized <illust_id>:p<num> / <illust_id>:ugoira — pximg URLs carry no content hash). PixivIngester wires client/downloader/ ledgers into ingest_core with drift label 'Pixiv app API' and the new body_canary=False opt-out: caption-less pixiv artists are common, so the zero-bodies #862 alarm would false-positive here — the client's response-shape drift checks cover that failure class instead. auth_token joins the uniform adapter constructor (pixiv is the first token-auth native platform). verify_pixiv_credential = one OAuth refresh, no feed walk. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
e9891ee9f3 |
feat(tags): system tags — is_system column, seeded hygiene tags, protection guards
Training hygiene step 1 (milestone #128). Migration 0075 adds tag.is_system and seeds wip / banner / editor screenshot (kind=general), ADOPTING an existing same-(name,kind) tag case-insensitively instead of duplicating. These rows drive the upcoming training exclusions, so they are protected: rename and merge-away refuse system tags (merge-INTO stays allowed — folding an operator's old hygiene tag into the system row is the intended move; merge is the only tag-delete path, so that guard covers deletion). is_system rides every tag serialization. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
19b962f1a7 |
feat(b3): ml-worker becomes optional — embed-only role, decoupled GPU coordination, cpu-embed switch
The ml-worker's ONLY processing role is now the CPU whole-image embed fallback (tag_and_embed renamed embed_image — Camie tagging was retired #1189 and the name kept implying otherwise; videos were already handled agent-style: frame sampling + mean-pool). Detection/cropping/CCIP stay GPU-agent-only, and their completion is judged per-pipeline: ccip by gpu_job rows, siglip by concept regions at the current model version — never by image_record.siglip_embedding. A CPU embed therefore can NEVER close crop work for the agent (regression test pins this; only the whole-image 'embed' job, the same artifact, is satisfied). Making removal actually safe (operator will drop the container): - GPU-queue coordination (enqueue_gpu_backfill, recover_orphaned_gpu_jobs, reprocess_gpu_jobs) moved verbatim to tasks/gpu_queue.py on the maintenance quick lane — it lived on the 'ml' queue only by module colocation, which made the ml-worker a hard dependency of the whole agent pipeline. - New ml_settings.cpu_embed_enabled (migration 0074, default ON so agent-less installs keep working): OFF stops the four import hooks queueing embed work nothing will consume and no-ops the manual backfill; switch lives on the renamed 'CPU embedding backfill' card. - NB heads training / auto-apply still run on the ml image (sklearn) — a stack that removes the container gives those up too. Deploy note: in-flight messages under the old task names are dropped by the new workers; the 60s orphan sweep + hourly backfill re-fire under the new names immediately. Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
eaea4308fc |
chore: retire the tag-eval harness — it proved the heads system, job done (operator-approved)
The head-vs-centroid eval (#1130) existed to prove the 'frozen embedding + trained head' spine; the operator accepted the tagging system and dropped the harness. Removed per rule 22: TagEvalCard + store, /api/tag_eval blueprint, tag_eval_run ml task, recover-stalled-tag-eval-runs sweep + beat entry, TagEvalRun model + table (migration 0073), and its tests. The eval's data loaders + metric helpers were NOT eval-specific — the nightly heads trainer runs on them — so they moved verbatim to services/ml/training_data.py (heads.py import updated; behavior unchanged). Co-Authored-By: Claude Fable 5 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM |
||
|
bvandeusen
|
a7abcc41ca |
feat(triage): failed-processing triage — probe errored files, flag defects, recover (#125 C1-C3)
An errored GPU job's stored reason is a suspicion; the file probe is the
verdict. A 15-min beat sweep (triage_gpu_errors) runs verify_integrity's own
probe (sha256 + decode) on each errored image ONCE and writes both verdicts:
ImageRecord.integrity_status and the new GpuJob.triage_status ('defect' |
'file_ok', migration 0072). Every classification logs at WARNING so it
surfaces in Logs/System Activity.
- 'defect' rows are excluded from /retry_errors (re-running a known-bad file
burns agent time re-minting the tombstone); response now reports
defects_kept and the GpuAgentCard toast says so.
- GET /api/gpu/errors: triage view — reason buckets (classify_reason),
probe verdicts, per-job detail. POST /errors/triage runs the sweep now.
- POST /api/gpu/errors/<id>/recover: reuses the Layer-2 refetch pattern —
delete the defective copy + record (full cascade takes the tombstones too)
and re-poll its subscription Source so a fresh copy re-imports and re-enters
the pipeline; 'no_source' when nothing pollable resolves.
- New 'Failed processing' card (GpuTriageCard) in Maintenance: verdict counts,
reason summary, probe-now, defect list with thumbnails + per-image Recover.
Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01CDgx8bQS5YrGRK76v8HUnM
|
||
|
bvandeusen
|
c22f37d64d |
feat(gallery): sort by earliest post date across all posts (new default)
The gallery's newest/oldest sort keys off image_record.effective_date = COALESCE(primary post's post_date, created_at). The primary post is often the repost/download the file came from, so the grid led with download dates rather than when content was first posted (operator-flagged). Add a second materialized sort key, earliest_post_date = MIN(post_date) across ALL of an image's provenance posts (every post it appears in), else created_at — the original publish date. Mirrors the effective_date pattern so the sort stays a forward index scan. - alembic 0071: add earliest_post_date + index (DESC, id DESC); backfill created_at baseline then MIN over image_provenance ⋈ post. - importer: recompute earliest_post_date whenever a dated post is linked (MIN over the image's provenance, which now includes the just-added row). - gallery_service: new sorts posted_new / posted_old key off earliest_post_date; cursor + year/month grouping follow the active column transparently. - api: accept posted_new|posted_old; DEFAULT is now posted_new so the grid leads with original publish date. newest/oldest (effective_date) still available. - frontend: sort dropdown gains "Newest/Oldest post date" (default Newest post date); existing effective-date sorts relabelled "Newest/Oldest added". - tests: service test asserts posted_new/posted_old key off earliest_post_date; frontend default-sort omission test updated. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
181f1c6a27 |
perf(gpu-queue): partial indexes + two-phase lease so leasing stays O(batch)
The throughput bottleneck was curator-side, not the network. lease() claimed the lowest-id pending/expired jobs with `... ORDER BY id LIMIT n`, but with only a plain `status` index Postgres walked the primary key from id=1, skipping the entire prefix of already done/error rows before reaching pending ones. As `done` grew (69k+), every lease became an O(done) scan — leasing crawled, the DB saturated, and even /status (the queue GROUP BY count) stalled the agent. - Migration 0070 adds two partial indexes over just the live slice: pending rows indexed by id (hot path), and leased rows by lease_expires_at (crash-recovery + orphan sweep). They stay tiny no matter how large the done/error history. - lease() split into two phases so each uses a partial index: claim pending first (id-ordered, O(batch)); reclaim expired leases only when pending can't fill the batch. Same semantics (SKIP LOCKED, attempts++, expired reclaim). - Model __table_args__ declares the indexes so ORM and schema agree. - Test: a done-prefix at low ids must not stop the lease reaching pending. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
359bc5a283 |
feat(ml): default to SigLIP 2 (new installs) + model dropdown, no free-text (#1203)
- Migration 0069: new installs default to SigLIP 2 (so400m, 512px, 1152-d drop-in) — UPDATE applies ONLY where no image is embedded yet (fresh install), so an existing library is NOT silently invalidated; it switches deliberately via the dropdown → Re-embed → Retrain. Column server_defaults moved to SigLIP 2. - GET /api/ml/embedder-models: server-authoritative supported list (SigLIP 2 512 recommended / 384 faster / SigLIP 1 384 original) so the UI never free-types. - GpuAgentCard: the two name/version text fields → a single model dropdown; Save sets name+version from the picked option (the current model is always selectable even if off-list). - embedder.py DEFAULT_MODEL_NAME unchanged (stays the baked local-dir SigLIP 1) to avoid a local-dir/weights mismatch; SigLIP 2 loads by HF name, cached on the ml-worker's persistent HF_HOME. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
bc6d43d3f2 |
refactor(ml): drop dead tagger/suggestion settings + columns (#1199)
Hygiene follow-up to the Camie retirement (#1189) — these were left inert to bound that change; nothing reads them now. Migration 0068 drops: - ml_settings: tagger_store_floor, tagger_model_version, suggestion_threshold_ character/general (already dead pre-retirement — scoring uses per-head thresholds), video_min_tag_frames (only the deleted video-prediction aggregator used it). - image_record: tagger_model_version (no writer), centroid_scores (dead JSON cache, no reader). Also: ml_admin _EDITABLE/GET/_validate pruned (dropped the store-floor invariant + video_min_tag_frames check); MLThresholdSliders trimmed to a video-embedding card (interval + max frames only); importer no longer resets the dropped cols; download_models drops the Camie fetch; stale CASCADE comments in cleanup_service no longer name the removed tables. Tests updated. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
485387ff0b |
refactor(ml): retire the Camie tagger + allowlist bulk-apply (#1189)
Heads + CCIP are the tag source and head auto-apply is the earned propagation.
The Camie tagger ran only to feed the allowlist bulk-apply (its ImagePrediction
rows had no other consumer), and the allowlist was a SECOND, un-earned auto-apply
path firing in parallel with heads on every accept — exactly the un-earned spray
the v2 pivot replaced. Retire both.
Behavior change: accepting a suggestion now applies the tag to THAT image only
(source='ml_accepted', a head-training positive) — it no longer allowlists +
fans the tag across the library via Camie. Propagation is heads' earned
auto-apply. (Loses instant cold-start propagation for booru-vocab tags; that was
un-earned and bypassed the precision gate.)
- tag_and_embed is now EMBED-ONLY (no Camie load/infer, no ImagePrediction
writes); backfill enqueues it for images with no embedding.
- Removed: services/ml/tagger.py, apply_allowlist_tags + helpers + daily beat +
every enqueue caller (accept/alias/merge/per-image), api/allowlist.py +
blueprint, ImagePrediction + TagAllowlist models/tables (migration 0067),
AllowlistTable.vue + allowlist store, the accept coverage-projection payload.
- AllowlistService gutted to accept/dismiss/undismiss/reject (the rejection store
the rail still needs); accept returns nothing, API returns {accepted, tag_id}.
- tag merge no longer repoints/triggers the allowlist; _keep_as_alias now keys on
ML-applied image_tag sources (incl. head_auto) instead of the allowlist.
- UI: MLBackfillCard relabelled to embedding-only; accept toast simplified;
MaintenancePanel drops the allowlist tile.
Left for a follow-up hygiene pass (now-inert, harmless): the dead settings
columns (tagger_store_floor, tagger_model_version, suggestion_threshold_*,
video_min_tag_frames), image_record.tagger_model_version, MLThresholdSliders
trim, and the Camie model download in download_models.py.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
|
||
|
bvandeusen
|
3d77a38a25 |
refactor(ml): remove the dead per-tag centroid subsystem (#1189)
The v2 pivot replaced per-tag SigLIP centroids with learned heads + CCIP. Centroids were still recomputed (on every tag merge + a daily beat) but NOTHING read them — suggestions come from heads+CCIP and apply_allowlist_tags applies via Camie predictions, not centroids. Pure dead wiring; remove it. Removed: CentroidService, recompute_centroid/recompute_centroids tasks, the daily beat, POST /api/ml/recompute-centroids, the recompute-on-merge trigger, the tag_reference_embedding table + model, the centroid_similarity_threshold + min_reference_images settings (migration 0066), the CentroidRecomputeCard + its store action + MaintenancePanel tile, and the centroid slider in MLThresholdSliders. _keep_as_alias drops its vestigial has-centroid branch (the allowlist branch already covers "could re-emit"); tag merge no longer clears a table that no longer exists. NOT touched (still live, parallel to heads): the Camie tagger, ImagePrediction, and the allowlist bulk-apply — accepting a suggestion still allowlists + applies it across the library. The tag-eval "centroid" baseline metric is unrelated (in-memory) and stays. (image_record.centroid_scores JSON column also remains — separate legacy field, its own micro-cleanup.) Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
4daa3f2790 |
feat(ml): operator model swap — GPU re-embed + embedder as a setting (#1190)
Make the SigLIP embedder an operator choice (drop-in to SigLIP 2:
google/siglip2-so400m-patch16-512 is a verified 1152-d model at 512px → no
schema change, better small-cue fidelity). A swap = set model + re-embed +
retrain, all operator-driven; the GPU agent does the re-embed so it's fast.
- settings: embedder_model_name is now a setting (migration 0065) alongside the
existing embedder_model_version; both editable + validated (non-empty) in the
ml admin API. The server embedder loads by HF name (AutoImageProcessor/Model,
model-agnostic), preferring the pre-downloaded local dir for the default so
existing deploys don't re-download; rebuilds on a name change.
- agent: new 'embed' job = whole-image SigLIP embedding (mean-pool video frames)
under the lease-announced model → POST /jobs/submit_embedding writes
image_record.siglip_embedding + siglip_model_version. The lease now announces
the model FROM THE SETTING (not a constant).
- re-embed routing: enqueue_gpu_backfill('embed') selects unembedded + stale-
version images; 'siglip' now re-embeds concept crops whose version != current
(so a swap re-triggers crops, not just the never-embedded back-catalogue). The
CPU ml-worker backfill no longer re-embeds on a version mismatch (it can't
churn the library at 512px) — the GPU agent owns version re-embeds. Daily
'embed' + 'siglip' beats self-heal.
- scoring: score_image only bags embeddings in the CURRENT model's space (whole-
image gated by siglip_model_version, concept regions by embedding_version) so a
mid-swap stale vector isn't scored by new-space heads; legacy NULL = current.
- UI: GpuAgentCard "Embedding model (advanced)" — edit name/version, Save, and
"Re-embed library (GPU)" (queues embed + siglip); points at SigLIP 2.
Tests: lease announces model + submit_embedding round-trip; enqueue 'embed'
selects stale/unembedded; stale-version excluded from scoring; embedder model
settable + empty rejected; siglip gate updated to current-version concept.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
|
||
|
bvandeusen
|
b91a230f12 |
feat(ccip): automation + reference quality — keep identity flowing hands-free (#114)
Works through the optional CCIP ideas + the "keep moving even if I forget" ask:
AUTOMATION (no button needed):
- Hourly beat auto-enqueues CCIP backfill — new images get embedded (and errored
ones retried) on their own; the queue never goes idle waiting for a click.
- CCIP auto-apply: a daily sweep tags confident matches (source='ccip_auto') so
identity tags keep flowing. ON by default (opt-out, like head auto-apply);
ml_settings.ccip_auto_apply_enabled + _threshold (0.92, above the suggest cut),
migration 0064. Vectorized (one matmul + reduceat per image), reversible, skips
already-applied/rejected. Switch + threshold in the GPU agent card; GET/PATCH
/api/ml/settings; auto_applied count in /api/ccip/overview.
REFERENCE QUALITY (the over-fire root cause):
- character_references now draws ONLY from single-character images — on a
multi-character image the tag is image-level, so every figure would otherwise
pollute each character's prototypes (a 2-char image tagged 'Velma' made
Daphne's figure a Velma reference). This is the contamination behind residual
over-firing.
- Cached on a cheap signature (char-tag count + ccip-region count/max-id) so the
reference load isn't redone on every modal open.
Tests: multi-character image not used as a reference; auto-apply tags a confident
match as ccip_auto.
NEXT (not done, confirmed): comic-panel cropping + SigLIP concept crops ("spot
interesting content").
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
|
||
|
bvandeusen
|
625336b6b4 |
feat(ccip): tunable match threshold, default 0.85 (#114)
Live data showed the v1 flat 0.75 cosine over-fired — ~64% of matched images got
3-10 character guesses dominated by the most-referenced characters (a 27-ref
character clears a low bar on many images). A sweep showed 0.85 collapses the
noise (noisy multi-matches 47→3) while keeping the confident single-character
matches.
- ml_settings.ccip_match_threshold (migration 0063, default 0.85); match_image
reads it (override still accepted). DEFAULT_SIM_THRESHOLD fallback 0.75→0.85.
- Exposed in GET/PATCH /api/ml/settings (validated 0.5–0.999).
- Slider in the GPU agent card ("Character-match strictness") — tune live, no
redeploy, same observe-and-tune loop as auto-apply.
Test: a ~0.9-cosine figure matches at 0.85, dropped at 0.95.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
|
||
|
bvandeusen
|
b735432d02 |
feat(gpu): video-ready regions + the HTTP GPU-job queue engine (#114 slice 3)
Answers "how are videos/all media handled by the GPU worker": a job is per ITEM, but the agent fans a VIDEO into per-frame instances (ffmpeg in the agent, the existing cadence), each stored with a timestamp — so a video becomes a BAG of frame embeddings (fixes the mean-embedding muddle) instead of one washed-out vector. Stills → frame_time NULL; animated GIF/WebP treated like short video. - image_region.frame_time (migration 0061, not yet deployed so folded in): the source frame's seconds for video/animated media; NULL for stills. RegionService passes it through. A whole frame is just kind='frame'. - gpu_job + GpuJobService (migration 0062): the durable work list that keeps the desktop agent HTTP-only — enqueue (dedupes (image,task)) / lease (FOR UPDATE SKIP LOCKED, re-claims expired leases so the queue self-heals) / heartbeat / complete / fail (re-queues until MAX_ATTEMPTS then 'error'). The server enqueues; the agent leases+submits over the web API; Redis/Postgres stay private. Tests: enqueue dedupe, lease-then-skip-when-held, expired-lease reclaim, scoped heartbeat, complete, fail-requeue-then-error. region test now covers frame_time. NEXT: the thin HTTP API (lease/submit/heartbeat) + bearer-token auth, then the agent container + control UI. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
0ea7ecdea5 |
feat(regions): image_region storage + service for the crop pipeline (#114 slice 2)
The storage backbone both crop jobs write to and read from. image_region =
normalized bbox (rx/ry/rw/rh) + kind ('face'/'figure' → CCIP character id;
'concept' → SigLIP head bag) + the crop's embedding (nullable Vector(768) CCIP /
Vector(1152) SigLIP, one per kind) + version stamps for compute-once gating. The
bbox doubles as grounded-tag provenance. Migration 0061.
RegionService.replace_regions (scoped BY KIND so the figure + concept pipelines
don't clobber each other) + get_regions — the GPU agent's results endpoint will
call the writer; the character matcher + bag scorer read. Server-side, no GPU.
Tests: replace/get round-trip, kind-scoped replacement, CCIP vector round-trip.
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
|
||
|
bvandeusen
|
48c8811d69 |
feat(heads): auto-apply observability + on by default (#114 auto-apply B)
Auto-apply is now ON by default (operator-asked: opt-OUT, not opt-in) — migration 0059 + model default flipped. The support (>=30) + measured-precision gates keep it safe and every auto-tag is reversible. Observability so the operator can tune from real data: - MISFIRE = an auto-applied (source='head_auto') tag the operator later removes. UNDER-FIRE = a tag with a head the operator adds by hand (the head missed it). Both captured at correction time in TagService.add_to_image/remove_from_image (source is lost on delete) into durable per-tag counters (head_metric), keyed by tag so they survive head retrain/prune. - Daily snapshot_head_metrics writes a per-concept time-series point (head_metrics_snapshot): auto-applied volume + cumulative misfires/under-fires + head quality; 180-day retention; daily beat. - GET /api/heads/metrics: per-concept current counts + realized misfire rate + head quality, plus the snapshot time-series — the report to tune the precision target + support floor. Migration 0060. Tests: misfire/under-fire counting (and the negatives — manual removal isn't a misfire, headless manual add isn't an under-fire), snapshot time-series, metrics API. What's the autofire threshold? There's no single number — each graduated head derives its OWN probability cutoff from its PR curve: the operating point that holds precision >= head_auto_apply_precision (0.97) at max recall. The global knobs are that target + the >=30 support floor. NEXT (slice 3): UI — enable toggle, dry-run preview, per-concept trends. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
74fef908d2 |
feat(heads): earned auto-apply — sweep mechanism, off by default (#114 auto-apply A)
Graduated heads can now apply their tag without a human — gated so it's safe:
- FIRING GATE: a head fires only when the master switch (head_auto_apply_enabled,
default OFF) is on AND it has >= head_auto_apply_min_positives (default 30)
clean labels. A precise-looking but under-supported low-N head can't spray tags.
- auto_apply_sweep (heads.py): streams every embedded image in chunks, scores
against the eligible heads (numpy, no sklearn), applies each head's tag where
score >= its auto_apply_threshold and the tag isn't already applied/rejected,
with source='head_auto' (distinguishable + reversible). dry_run counts only.
- HeadAutoApplyRun (migration 0059) tracks each sweep / preview; apply_head_tags
task (ml queue) + scheduled_apply_head_tags daily beat (no-op unless enabled)
+ recovery sweep + retention(20).
- API: POST /api/heads/auto-apply {dry_run} (202 / 409 running / 400 disabled),
GET /api/heads/auto-apply (recent runs + per-concept report). Settings
head_auto_apply_enabled + min_positives via /api/ml/settings.
Tests: sweep applies above threshold, dry-run writes nothing, skips under-
supported + ungraduated heads; API disabled/dry-run/conflict guards.
NEXT (slice 2): the observability the operator asked for — per-concept misfire
(auto-applied-then-removed) + under-fire tracking, time-series snapshots, and a
reporting API to tune. Slice 3: the UI (enable, preview, trends).
Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com>
Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa
|
||
|
bvandeusen
|
22c3b54746 |
feat(heads): production per-concept heads — train + score backend (#114 A)
The eval (#1130) proved the frozen-embedding + trained-head spine; this lands its production form (the first of three slices that make heads the suggestion source, replacing Camie + centroid). - tag_head: one logistic-regression head per general/character concept with enough labelled positives. Weights (pgvector), honest CV-derived suggest threshold + earned-auto-apply point, and per-concept quality metrics. - head_training_run: persisted batch lifecycle (mirrors tag_eval_run) so the admin card shows live + historical status across navigation. - services/ml/heads.py: TRAIN (sync, ml worker, reuses tag_eval's proven data loaders + metric math so production heads match measured eval numbers) and SCORE (async, API worker — numpy via pgvector, no scikit-learn): score one image's embedding against all heads → the rail's suggestions, cached on (count, max trained_at) so a retrain invalidates without per-request loads. - tasks.ml.train_heads (ml queue, commits per head so a kill leaves progress) + recover_stalled_head_training_runs sweep + retention(20) + 5-min beat (rule 89). - api/heads.py: POST /api/heads/train (one run at a time, 409 guard) + GET /api/heads (count, graduated, last-trained, running, per-concept table, recent runs). - ml_settings: head_min_positives + head_auto_apply_precision, tunable via /api/ml/settings. Scoring isn't wired into the rail yet (slice C) and the admin UI is slice B — this slice makes training + scoring exist and CI-verifiable. 'precision' column stored as precision_cv (SQL reserved word). Migration 0058. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> Claude-Session: https://claude.ai/code/session_01Ttrj5P7upUTueSfoJcxEqa |
||
|
bvandeusen
|
b69c70ab2b |
feat(tag-eval): "keep" records a confirmation so doubts stop resurfacing
"Keep" on a doubted positive was a no-op, so the same confirmed-correct images came back in "head doubts" every run (operator-flagged: reinforcement keeps surfacing the same images). Add tag_positive_confirmation (mirror of tag_suggestion_rejection): keep → POST /images/<id>/tags/<tag_id>/confirm, and the eval excludes confirmed positives from the doubts list — exactly as rejected items already drop out of the suggest list. The tag stays a positive either way (confirmation is a "reviewed" marker, not a training change). - model TagPositiveConfirmation + migration 0057; confirm endpoint (idempotent). - tag_eval: _confirmed_ids + exclude from head_doubts_positive examples. - store.confirmTag + card "keep" calls it. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
6e3c5f697f |
feat(ml): tag-eval backend — head-vs-centroid learning-curve eval (persisted)
Slice 1 of milestone #114 (tagging v2). Proves the frozen-embedding + trained- head spine on the operator's own data, reusing the SigLIP embeddings already stored on image_record — no re-embedding, no GPU. Per concept: train a logistic-regression HEAD (positives + negatives = explicit rejections + sampled unlabeled) vs the old single-CENTROID baseline; report cross-validated precision/recall/AP for both, a LEARNING CURVE (AP/F1 as tagged positives grow 10→30→100→300), and example image ids (head-would-suggest / head-doubts-positive) to eyeball. Persisted so the report SURVIVES navigation (operator-flagged): the run + full report live in a new tag_eval_run row (mirrors library_audit_run); the admin card will rehydrate from GET on mount, not transient state. - models.TagEvalRun + migration 0056; runs on the ml queue (only worker with numpy/sklearn) — numpy/sklearn lazy-imported so the API can still enqueue. - services/ml/tag_eval (compute + start helper, one-running guard), tasks.ml .tag_eval_run, api/tag-eval (POST create, GET history light / detail w/ report). - recover_stalled_tag_eval_runs sweep + retention (keep last 20) + 5-min beat (rule 89). scikit-learn added to requirements-ml. - tests: param normalization + the rehydrate read-path + create/conflict. Frontend admin card (trigger + render persisted report) follows next. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
5269cd0709 |
feat(provenance): capture which archive an extracted image came from (#87)
Images pulled out of a .zip/.rar previously kept no record of WHICH archive
they came from — the member->archive link was computed during extraction and
discarded, leaving only image->post. So the provenance modal could only scope
attachments to the whole post, showing every archive a 'High Resolution Files'
bundle carried instead of the one a given file lives in.
- ImageProvenance.from_attachment_id: nullable FK -> post_attachment.id
(SET NULL), migration 0055.
- importer: _import_archive stamps from_attachment_id on every member's
provenance row for the post (new + superseded + deduped members), resolving
the archive's own PostAttachment by (post, sha). Post-pass UPDATE, NULL-only
and idempotent, so it doesn't touch the dedup/supersede branches and the
backfill is safe to re-run. Nested members link to the outer stored archive.
- provenance_service.for_image: when the originating post's provenance row
records from_attachment_id, return ONLY that archive; else fall back to the
primary-post scoping from
|
||
|
bvandeusen
|
f678819093 |
feat(subscribestar): seen/failed ledger models + migration 0054 (#889)
Phase 1, step 1 of moving SubscribeStar off gallery-dl onto the native core ingester (milestone: SubscribeStar native). Mirror of the Patreon ledger: SubscribeStarSeenMedia (skip already-ingested media on routine walks; recovery bypasses) and SubscribeStarFailedMedia (dead-letter so persistently-failing media stops re-burning backfill chunks). Per operator decision, dedicated per-platform tables (not a generalized shared ledger). filehash is String(128): a CDN content hash when the URL carries one, else a synthesized <post_id>:<filename> key. UNIQUE (source_id, filehash) upsert key. Registered in models/__init__; migration 0054 creates both tables (down 0053). Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
bvandeusen
|
369e3de684 |
feat(ml): cadence-based video frame sampling + min-frame tag aggregation (#747)
Video tag noise root cause: frames were a FIXED count (6) max-pooled — a tag firing on one frame survived at peak confidence, and a fixed count under-samples long multi-scene videos so real scene-local tags looked like noise. Redesign (operator-steered): - Sample at a fixed CADENCE — one frame every `video_frame_interval_seconds` (default 4) across the 5–95% window — so a tag's frame-presence reflects real screen time independent of video length. Capped at `video_max_frames` (default 64): a long video stretches the spacing instead of exploding into hundreds of inferences, bounding per-video cost on the single ml-worker (per-frame ffmpeg timeout also cut 60s→30s). - Aggregate with `_aggregate_video_predictions`: keep a tag only if it appears in >= `video_min_tag_frames` sampled frames (≈ that many × interval seconds on screen — duration-independent noise rejection), with confidence = MEAN over the frames it appears in (not max). Clamps the threshold to the sample count so a 1–2-frame short video still tags. - All three knobs are DB-backed ml_settings (migration 0053), patchable via /api/ml/settings + sliders in the ML settings card — replaces the VIDEO_ML_FRAMES env var (product-not-project). Tests: aggregation drops one-frame noise + means corroborated tags + clamps on short videos; settings round-trip + min>max validation. Replaced the _maxpool_predictions unit test. NOTE: this is the QUALITY half of #747. The perf half — the ml-worker runs CPU-only — is GPU enablement, tracked separately in #872. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
bvandeusen
|
f154603811 |
feat(import): Tier-1 video near-dup by duration+aspect (#871)
Videos deduped on sha256 only (pHash is images-only), so a different encode/remux of the same clip imported as a distinct record — the "same video from multiple sources" clutter surfaced by #859. Tier-1 metadata fingerprint: identity = container duration (±1.0s) + matching aspect ratio, scoped to the same artist; quality axis = pixel dimensions (mirrors image pHash: larger_exists→skip+link, smaller_exists→supersede). Codec/bitrate are deliberately NOT part of identity (the point is matching across re-encodes). Tight tolerances because a wrong video merge is destructive. - image_record.duration_seconds (Float, nullable; migration 0052). NULL for images. - safe_probe.probe_video also reads format=duration (one extra ffprobe field on the call that already runs); ProbeResult.duration. - _find_similar_video(duration,w,h,artist) shared by both import pipelines. - _import_media (filesystem/archive path): captures duration, video near-dup branch, persists duration. - attach_in_place (download path — handles #859's videos, previously didn't probe video at all): best-effort probe for dims+duration (LENIENT — never newly rejects a downloaded video on probe failure), video near-dup branch, persists duration. - _supersede carries duration onto the kept row. Reuses SkipReason.duplicate_phash so the existing download/external dup-cleanup (path-safe unlink, #859) applies unchanged. Tests: skip-smaller, supersede-larger (+ duration adopted), and distinct-durations-not-merged (false-merge guard). Follow-up (Phase 2, #871): a backfill to re-probe NULL-duration existing videos so the current library participates in dedup; retroactive merge of existing dups is a separate destructive maintenance action. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
bvandeusen
|
96c29c370b |
feat(ingest): localize inline post-body images to local copies (Phase 2)
Render a post body faithfully by serving our stored copies of inline images instead of hotlinking the public CDN. The join key is the CDN filehash (32-hex MD5) shared between a body <img src> and the media URL we downloaded (the same identity extract_media dedups by): - utils.paths.filehash_from_url — one source of truth for the extractor; patreon_client._filehash now delegates so capture- and render-time hashing cannot drift. - ImageRecord gains source_url (provenance) + source_filehash (indexed match key); migration 0051. - the per-media sidecar carries the file's source_url; the importer persists it (NULL-only) on the ImageRecord via _apply_sidecar. - post_feed_service.get_post remaps body <img src> -> /images/<path> for every inline image whose filehash maps to a stored image of THIS artist; unmatched / pre-Phase-2 images keep hotlinking. Pre-existing on-disk images have no filehash yet, so they fall back to hotlinking until re-downloaded; localization is forward-looking. Co-Authored-By: Claude Opus 4.8 <noreply@anthropic.com> |
||
|
bvandeusen
|
8dbf29f803 |
feat(external): per-host enable toggles in Settings (Phase 4d)
Operator lever: disable a single file host (e.g. mega.nz when it's banning) without touching the others. Five booleans on import_settings (extdl_<host>_enabled, default true — works out of the box, rule #26); the worker already reads them via getattr so no worker change. Migration 0050 + model fields + settings GET/PATCH (uniform boolean validation) + a 'External file-host downloads' card in the subscriptions Settings tab. Completes Phase 4. Refs FC #830. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
d96918d777 |
feat(posts): extract + record external file-host links (Phase 3)
Capture off-platform links (mega/gdrive/mediafire/dropbox/pixeldrain) embedded in post bodies so they're never silently dropped, and surface them in the post view. The download worker (Phase 4) walks these rows. - link_extract.py: pure extractor — <a href> + bare URLs, unwraps Patreon redirect shims, PRESERVES the full url incl. #fragment (mega's key), dedups. Reusable by every platform (runs off Post.description). - external_link model + migration 0049: post_id/artist_id/host/url/label/status /attempts/last_error/attachment_id/timing; CHECK whitelists (full enum incl. worker statuses up front) + (post_id,url) unique. - importer._sync_external_links: insert-missing on both import paths (_apply_sidecar + upsert_post_record) so a re-import never resets a link's status; runs for all platforms. - post_feed_service.get_post: returns external_links (detail-only). - PostCard: renders the links (host chip + label + status) once expanded. - tests: extractor (5 hosts, fragment, shim unwrap, dedup), importer (record + no-dup on reimport), serializer. Refs FC #830. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
7bb765b6ed |
feat(series): pending staging for add-from-post (#789 Phase 2)
Add-from-post no longer appends straight into the run — it STAGES the post's
pages as pending (per-page status; page_number NULL), grouped by source post,
so the operator drops junk (text-free alts, bumpers) and places the keepers
into the sequence with clean series-global numbering.
- migration 0048: series_page.status ('placed' default | 'pending') + nullable
page_number.
- series_service: placed/pending split everywhere (list_pages returns the
placed run + a `pending` section grouped by source post; reorder/cover/
list_series operate on placed only); add_post stages pending; new
place_pending(image_ids, before_image_id=None) flips pending→placed spliced
before a page (or appended) and renumbers; junk removal reuses remove_images.
- api/tags: /add-post now returns staged count; new POST /series/<id>/pending/
place.
- frontend: PostSeriesMenu navigates to the series after staging; seriesManage
store surfaces `pending` + placePending; SeriesManageView gains a pending
tray (per-post groups, place-all / place-one / drop-junk).
- tests: pending staging, place (append + insert-before), ignore-already-
placed, drop-junk, route guard; updated add_post + match-accept expectations.
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
||
|
bvandeusen
|
59746d213d |
feat(series): flat series sequence + cosmetic chapter dividers (#789 Phase 1)
Reframe a series from "ordered chapters that own pages" to ONE flat, series-global ordered run of pages with optional cosmetic chapter DIVIDERS over it. A chapter no longer wraps content — it's a labeled divider anchored to the page that begins it; a page's chapter is derived as the nearest preceding divider. This is what lets installments assembled from multiple sources sit in one continuous, correctly-numbered sequence (operator's Goblin Juice case). - migration 0047: flatten each series to a series-global page_number (preserving today's reading order); convert each existing chapter to a divider anchored at its first page (keeping title/stated_part); drop series_page.chapter_id; reshape series_chapter (anchor_page_id UNIQUE FK, drop chapter_number/is_placeholder/stated_page_start/end). Loss-safe for content; drops empty placeholder chapters + a redundant page-1 divider. - series_page: page_number is now the series-global order; no chapter_id. - series_chapter: anchored divider (anchor_page_id, title, stated_part). - series_service: flat list_pages (one run + derived dividers + per-page source_post + part_gaps), series-wide reorder/renumber, divider CRUD (create/update/move/delete); retired per-chapter reorder/merge/placement. - api/tags: drop chapter_id from add; /chapters endpoints are divider create/update/delete (removed chapter reorder/merge/page-reorder). - series_match_service: series "end" reads max(series_page.stated_page); accept appends via add_post. tag_service series-merge appends src's pages after tgt's max so the merged series stays one clean run. - frontend: seriesManage store + SeriesManageView → one continuous drag-reorder grid with inline divider bars + series-global page numbers; reader walks the flat run, headings from dividers; PostSeriesMenu copy. - tests reworked across the series suite for the divider model. Phase 2 (pending staging for add-from-post) is separate. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
3610ba495f |
feat(ml): drop image_record.tagger_predictions — image_prediction is sole store (#768 step 3)
Read cutover verified in prod (suggestions + allowlist read image_prediction; backfill complete at 908k rows / 51k images). Removes the old JSON column and everything that fed it: - ImageRecord.tagger_predictions column removed; migration 0046 DROPs it. tagger_model_version kept as the "tagged / current?" signal the backfill sweep reads (needs-tagging check switched to tagger_model_version IS NULL). - tag_and_embed no longer dual-writes the JSON — image_prediction is the only write path. - importer re-import reset drops the JSON line (image_prediction rows are already deleted on re-import). - Retired the one-time #768 backfill task + the #764 prune task, their admin endpoints, and their Maintenance cards (Backfill/PrunePredictionsCard). - Tests seed/assert via image_prediction; stale column refs removed. Disk reclaim is NOT automatic: DROP COLUMN is a catalog change. Run `VACUUM FULL image_record` off-hours afterward to return the ~100 GB to the OS so DB backups go small (#739). image_prediction (~90 MB) stays in pg_dump — it's the source of truth now. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
65211a3f2f |
fix(migration): make 0045 DDL-only; backfill image_prediction via batched task (#768)
The inline INSERT…SELECT backfill in migration 0045 wrapped the table creation and a ~100 GB pass over image_record.tagger_predictions in one transaction: nothing committed until the end, it was unmonitorable, and an earlier MATERIALIZED-CTE form spilled the full 100 GB to temp on NFS. A deploy got stuck on it for ~2h with image_prediction never appearing. Split the concerns: - 0045 now creates ONLY the table + indexes (instant DDL → web boots). - New backend.app.tasks.admin.backfill_image_predictions_task copies the >= store-floor predictions from the JSON into image_prediction, batched by id window and committed per chunk: live progress, resumable (re-enqueues from the last committed id), idempotent (ON CONFLICT DO NOTHING). json_each stays in the DB executor streaming each window — no Python-side 100 GB load, no materialization. - POST /api/admin/maintenance/backfill-predictions + a Maintenance-tab card to trigger the one-time run after upgrading. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
e6d5f67f11 |
perf(migration): 0045 streams json_each via inline CASE guard (no temp spill)
The MATERIALIZED-CTE scalar guard forced Postgres to materialize all object rows with their full JSON (~100 GB) to temp before json_each — on NFS that's a huge spill and pathologically slow (risks disk-full). Replace with an inline CASE that feeds json_each an empty object for non-object rows: same scalar guard, but a single streaming pass with no materialization. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
a712cef92d |
fix(migration): 0045 backfill guards json_each against non-object rows
Some image_record rows store tagger_predictions as a JSON scalar/null rather than an object; json_each throws 'cannot deconstruct a scalar' on those, rolling back the whole migration. Filter to json_typeof = 'object' in a MATERIALIZED CTE so the guard runs before json_each ever evaluates a scalar. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
75eab188c8 |
fix(migration): 0045 backfill filters to >= store floor (supersedes #764 prune)
The #764 in-place prune (rewrite tagger_predictions to >=0.70) is too slow on 100 GB of TOAST and fails at its soft limit (interrupts a query mid-flight -> 'another command is already in progress'). #768 supersedes it: extract only the >=floor predictions into image_prediction via this set-based backfill, then drop the column (step 3) — reading 100 GB once + writing ~840k small rows beats rewriting 100 GB in place. So this backfill no longer assumes the prune ran: it filters by ml_settings.tagger_store_floor (default 0.70) itself, handling the full or partially-pruned JSON identically. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
79089b50b0 |
feat(ml): image_prediction table + backfill + dual-write (#768 step 1)
Normalize tagger predictions out of the image_record.tagger_predictions JSON blob into a queryable per-prediction table. Step 1 of the cutover (expand): additive + low-risk — reads still use the JSON, this just adds the table and keeps it populated. - ImagePrediction(image_record_id, raw_name, category, score) — stores the RAW tagger vocab name (not tag_id) so read-time alias→canonical resolution is unchanged. Indexed for per-image reads + by (raw_name, score). - Migration 0045: create table + set-based backfill from the JSON via json_each (fast post-#764-prune). The old column stays (vestigial) and is dropped in a later follow-up — DROP needs an ACCESS EXCLUSIVE lock on the hot image_record table, so it waits for a quiesced-worker window. - tag_and_embed dual-writes the rows (delete-then-insert, idempotent); tagger_store_floor already applied in infer(). Next: switch suggestion + allowlist reads to the table, then drop the JSON write. Plan-task #768. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
3f92669f12 |
feat(ml): DB-backed tagger_store_floor (default 0.70), the ingest confidence floor
Promotes the prediction store-floor from the TAGGER_STORE_FLOOR env (default 0.05) to a DB-backed, Settings-UI-tunable ml_settings column (default 0.70). Storing every tag down to 0.05 from a ~10k-tag tagger is what grew image_record's TOAST to ~100 GB; the suggestion path already filters at 0.70 and the centroid/learned path covers lower-confidence preferred tags, so the sub-0.70 tail is redundant. Foundation for plan-task #764 (backfill + reclaim land next; this only changes the write gate for NEW imports). - ml_settings.tagger_store_floor (migration 0044, default 0.70) - tagger.Tagger.infer(store_floor=...); ml task passes settings.tagger_store_floor - ML admin GET/PATCH expose it; PATCH rejects a category suggestion threshold below the floor (nothing below the floor is stored, so the gap surfaces nothing) — server backstop for the UI slider clamp - Settings → ML: store-floor slider + caption; category sliders min-bound to it Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
a8f624a0f1 |
fix(posts): link duplicate items to every post + prune bare shells
The native Patreon backfill flooded the feed with bare 'Post <id>' shells (1589 for Anduo). Root cause: PostAttachment.sha256 was GLOBALLY unique, so a non-art file reused across posts only ever linked to the first one, and _capture_attachment created the Post before that dedup check — leaving later posts with no image and no attachment. Duplicate IMAGES had the mirror gap: attach_in_place returned duplicate_hash/duplicate_phash before _apply_sidecar, so the second post got no provenance row, and the feed only rendered via primary_post_id (one post per image). Operator requirement: a duplicate item must show on EVERY post it appears in. Unify the fix as link-not-suppress: - importer: on duplicate_hash / duplicate_phash(larger_exists), append an image_provenance row for the new post (keep primary on the first). Both the download path (attach_in_place) and the filesystem path (_import_media). - post_feed_service: render thumbnails by image_provenance UNION primary_post_id, so a cross-posted image shows on every post (and legacy primary-only images still show). - PostAttachment: per-post uniqueness — drop UNIQUE(sha256), add partial UNIQUE(post_id, sha256) + partial UNIQUE(sha256) WHERE post_id IS NULL (migration 0043); _capture_attachment dedups per-(post,sha) over the shared sha-addressed blob, so no post is left bare. - cleanup: new prune-bare-posts maintenance action (cleanup_service _bare_post_conditions shared by preview/count/delete per preview/apply parity; admin endpoint; PostMaintenanceCard). Deletes posts with zero image links (primary or provenance) AND zero attachments. Run after the feed fix so a hidden provenance link spares the post instead of deleting it. Tests: dup image shows on both posts; dup attachment shows on both posts; feed renders provenance-linked duplicates; prune-bare delete-path == preview. Operator redeploys (migration 0043) then runs the prune to clear the shells. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
978959bdc4 |
feat(series): manage-view redesign — big pages, editable Part #, slide-over picker (FC-6.4)
Operator feedback: thumbnails too small to judge order, no obvious way to mark
'this installment is Part 2', and the permanent two-pane picker was busy and
competed with the ordering work.
- Full-width parts, each a card with a big page grid (150px, contain so whole
pages are visible) and drag-to-reorder; positional page number as a badge.
- Editable Part # (hero field) backed by new series_chapter.stated_part —
separate from the auto-managed chapter_number, mirroring the page_number vs
stated_page split so reorder/delete renumbering can't wipe a hand-set part.
Missing-Part hints when consecutive parts' stated_part jump >1.
- Each part labels its source post (derived from pages' primary_post_id) and
shows the printed-page range with clear labels.
- Picker demoted to an on-demand right slide-over ('Add pages') with a target-
part selector; part actions (move/merge/delete) collapsed into an overflow ⋮.
alembic 0042 adds series_chapter.stated_part (nullable int).
Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
|
||
|
bvandeusen
|
c0fd80e694 |
feat(series): assisted-continuation matcher + suggestion queue — backend (FC-6.3)
Confirm-only "this post may continue this series" matcher. - series_suggestion table (post_id, series_tag_id, score, signals jsonb, status pending|added|dismissed, UNIQUE(post,series)); migration 0041 + two settings knobs (series_suggest_enabled, series_suggest_threshold). - series_match_service: weighted additive score (title-stem / same-artist / page-continuity / shared-distinctive-tags), no single signal gating. The title "pattern" is derived on the fly from the post titles already in a series, so it sharpens as more are confirmed (no persisted state to drift). Candidates are bounded to the post's artist. match_post upserts pending suggestions (UNIQUE + on-conflict, respecting prior added/dismissed decisions). - accept reuses add_post_as_chapter then marks 'added'; dismiss marks 'dismissed'. - rescan_series_suggestions_task: settings-gated, time-boxed + self-resuming from a post-id cursor (maintenance_long lane), like normalize_tags_task. - API: GET /series/suggestions, POST .../<id>/accept|dismiss, POST .../rescan. - Settings: enabled + threshold exposed via /settings/import. - Tests: pure scoring helpers + matcher/accept/dismiss/rescan lifecycle + UNIQUE dedup. Frontend (Suggestions tab + settings card) lands next. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
1804a2c622 |
feat(series): chapter layer over series_page — backend (FC-6.1)
Adds an ordered chapter layer to series. Reading order becomes (series_chapter.chapter_number, series_page.page_number); a chapter may be a placeholder reserving a slot, and carries an optional parsed stated-page range used to flag missing-page gaps. An image still lives in at most one series ⇒ one chapter (image_id stays UNIQUE). - models: series_chapter; series_page gains chapter_id (NOT NULL, cascade) + stated_page. Migration 0040 backfills every existing series into one auto-chapter holding its current flat pages — no data loss. - SeriesService: chapter CRUD (create/update/reorder/delete/merge), page→chapter assignment, reorder_pages, chapter-aware set_cover; list_pages now returns chapters[] + gaps[] alongside a back-compat flat pages[]. Legacy series-wide reorder operates on the single default chapter and rejects multi-chapter series. - API: chapter endpoints under /api/series/<tag>/chapters; POST pages accepts an optional chapter_id. - TagService.merge now repoints series_chapter too, so a merged series' chapters (and their pages) survive the source tag's deletion instead of cascading away. - Tests: new chapter suite; updated the 4 direct SeriesPage(...) constructions to supply chapter_id. Frontend (chapter-aware manage view + reader) lands next; until then the existing UI keeps working via the flat pages[] + single default chapter. Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |
||
|
bvandeusen
|
f2e9ae07dc |
fix(audit): chunk + self-resume library scans (stop the 2h queue-hog timeouts)
scan_library_for_rule ran one 2-hour pass that timed out on large libraries and held the concurrency-1 maintenance queue the whole time, starving vacuum/backup/ normalize (operator-flagged — it was the dominant entry in the 24h failures). It now runs ~10-min chunks and re-enqueues itself until the library is exhausted, matching the operator's preferred pattern (reasonable timeout → retry queued → other things process between). New columns (alembic 0039): resume_after_id persists the keyset cursor so a chunk continues where the last left off; last_progress_at lets the recovery sweep tell a progressing multi- chunk audit from a dead one (it now measures staleness from last_progress_at, not started_at). Matches accumulate across chunks. soft/hard limits dropped 2h→15/16.7 min so the in-chunk budget fires first; a soft-limit backstop re-enqueues to resume instead of erroring the whole run. Tests: time-box → re-enqueue (status stays running); resume carries prior matches and appends new ones. Existing full-scan tests unchanged (small sets finish in one chunk). Co-Authored-By: Claude Opus 4.8 (1M context) <noreply@anthropic.com> |