Public-flip audit: env-driven paths, scrub audit-ticket prefixes, terser README
the host bind paths + LAN host pins replaced with env defaults. Repository URLs → git.sulkta.com. Audit-changelog scaffolding stripped from inline comments (technical reasoning preserved). README sheds marketing scaffolding. AI-speak in load-bearing prompts/SOULs left alone — that IS the product.
This commit is contained in:
Sulkta
parent
566ca05ea8
commit
564233cbff
20 changed files with 233 additions and 258 deletions
|
|
@ -7,7 +7,7 @@ version = "0.0.1"
|
|||
edition = "2024"
|
||||
license = "MIT"
|
||||
authors = ["Sulkta"]
|
||||
repository = "http://127.0.0.1:3001/Sulkta/skald"
|
||||
repository = "https://git.sulkta.com/Sulkta/skald"
|
||||
|
||||
[workspace.dependencies]
|
||||
tokio = { version = "1", features = ["full"] }
|
||||
|
|
|
|||
26
Dockerfile
26
Dockerfile
|
|
@ -1,26 +1,16 @@
|
|||
# Multi-stage build for skald.
|
||||
# Multi-stage build for skald. Postgres ships in the same image
|
||||
# until the tool stabilises; to split, swap the runtime base to
|
||||
# debian:bookworm-slim, drop entrypoint.sh, point DATABASE_URL at
|
||||
# an external pg.
|
||||
#
|
||||
# Stage 1: compile the rust binary against rust:1-bookworm.
|
||||
# Stage 2: pgvector/pgvector:pg17 (debian-bookworm postgres with
|
||||
# pgvector preinstalled) + tini + the skald binary.
|
||||
#
|
||||
# v0.1 ships postgres inside the same container ("singleton till we
|
||||
# have a real working tool"). When we extract the DB out, swap the
|
||||
# runtime base to debian:bookworm-slim, drop entrypoint.sh, point
|
||||
# DATABASE_URL at the external pg.
|
||||
#
|
||||
# Build context is the workspace root:
|
||||
# docker build -t skald:latest .
|
||||
# Build context is the workspace root: `docker build -t skald:latest .`
|
||||
|
||||
# ─── builder ──────────────────────────────────────────────────────
|
||||
FROM rust:1.95-bookworm AS builder
|
||||
WORKDIR /build
|
||||
|
||||
# Cache the dependency graph: copy manifests first, fetch + build
|
||||
# stubs, THEN drop in real sources. The vendored clawdforge SDK
|
||||
# needs its own manifest + source available during the cache layer
|
||||
# (path dep — Cargo resolves it at workspace load time, not at
|
||||
# crate-compile time).
|
||||
# Dependency-cache layer: copy manifests + vendored path-dep first,
|
||||
# build stubs, then drop in real sources. clawdforge is a path dep
|
||||
# resolved at workspace load time.
|
||||
COPY Cargo.toml Cargo.lock ./
|
||||
COPY skald-core/Cargo.toml skald-core/Cargo.toml
|
||||
COPY skald/Cargo.toml skald/Cargo.toml
|
||||
|
|
|
|||
21
LICENSE
Normal file
21
LICENSE
Normal file
|
|
@ -0,0 +1,21 @@
|
|||
MIT License
|
||||
|
||||
Copyright (c) 2026 Sulkta
|
||||
|
||||
Permission is hereby granted, free of charge, to any person obtaining a copy
|
||||
of this software and associated documentation files (the "Software"), to deal
|
||||
in the Software without restriction, including without limitation the rights
|
||||
to use, copy, modify, merge, publish, distribute, sublicense, and/or sell
|
||||
copies of the Software, and to permit persons to whom the Software is
|
||||
furnished to do so, subject to the following conditions:
|
||||
|
||||
The above copyright notice and this permission notice shall be included in all
|
||||
copies or substantial portions of the Software.
|
||||
|
||||
THE SOFTWARE IS PROVIDED "AS IS", WITHOUT WARRANTY OF ANY KIND, EXPRESS OR
|
||||
IMPLIED, INCLUDING BUT NOT LIMITED TO THE WARRANTIES OF MERCHANTABILITY,
|
||||
FITNESS FOR A PARTICULAR PURPOSE AND NONINFRINGEMENT. IN NO EVENT SHALL THE
|
||||
AUTHORS OR COPYRIGHT HOLDERS BE LIABLE FOR ANY CLAIM, DAMAGES OR OTHER
|
||||
LIABILITY, WHETHER IN AN ACTION OF CONTRACT, TORT OR OTHERWISE, ARISING FROM,
|
||||
OUT OF OR IN CONNECTION WITH THE SOFTWARE OR THE USE OR OTHER DEALINGS IN THE
|
||||
SOFTWARE.
|
||||
130
README.md
130
README.md
|
|
@ -1,95 +1,79 @@
|
|||
# skald
|
||||
|
||||
Long-form story-writer with canon-keeping, sequel continuity, and
|
||||
(future) self-hosted audiobook narration. Database is the source of
|
||||
truth — the writer is the tooling.
|
||||
self-hosted audiobook narration. The database is the source of truth;
|
||||
the binary is the tooling.
|
||||
|
||||
Named for the Old Norse poets who composed and memorized kings'
|
||||
sagas across generations.
|
||||
Named for the Old Norse poets who composed and memorized kings' sagas
|
||||
across generations.
|
||||
|
||||
## Status: v0.1 — scaffold
|
||||
|
||||
What's wired:
|
||||
## What's wired
|
||||
|
||||
- Rust workspace (`skald-core` + `skald`)
|
||||
- Postgres schema for stories, characters, canon facts, chapters,
|
||||
passages, generation runs, audit findings, tags
|
||||
- pgvector extension installed for future similarity search
|
||||
- pgvector for future similarity search
|
||||
- `skald import-markdown` ingests a story file (chapters + bible)
|
||||
into the schema
|
||||
- `skald serve` exposes `/health` and runs migrations on boot
|
||||
- Single-container deploy: postgres + skald in one image
|
||||
|
||||
Wired (this commit):
|
||||
|
||||
- clawdforge Rust SDK vendored at `vendor/clawdforge/` (upstream:
|
||||
`Sulkta-OSS/clawdforge` `clients/rust/`)
|
||||
- `skald-core::forge` — three-pass orchestration shell (gen / cleanup /
|
||||
audit). Prompts are TODO stubs; pipeline plumbing is in place.
|
||||
|
||||
Not yet wired:
|
||||
|
||||
- Web UI (the inbox + browse + queue surface)
|
||||
- Prompt templates for the three passes (heavy prompt-engineering
|
||||
work — own session)
|
||||
- `skald-core::context` — assemble the LLM context blob from DB rows
|
||||
(bible + characters + parent prose summaries + similarity-matched
|
||||
passages)
|
||||
- Embeddings backfill + ivfflat index
|
||||
- TTS sidecar container + post-render audit chain (see
|
||||
`docs/tts-pipeline.md`)
|
||||
|
||||
## v0.1 smoke
|
||||
|
||||
```bash
|
||||
docker compose -p skald up -d
|
||||
docker exec skald skald import-markdown \
|
||||
--path /seed/coast-down.md \
|
||||
--title "The Coast-Down"
|
||||
|
||||
curl http://localhost:7780/health
|
||||
# → { ok: true, db_ok: true, story_count: 1, ... }
|
||||
```
|
||||
- `skald serve` exposes `/health` + the web inspector and runs
|
||||
migrations on boot
|
||||
- `skald continue` runs gen → cleanup → audit per chapter, with
|
||||
multi-chapter batching (cap 20)
|
||||
- `skald rewrite` re-authors a chapter in a named author's voice
|
||||
- `skald audit` runs whole-story prose-quality audit; `skald dedup`
|
||||
is the surgical fix half of the loop
|
||||
- `skald prepare-narration` annotates a chapter with `[breath]` /
|
||||
`[pause:Xs]` / `[scene]` beats and per-character `[voice:...]`
|
||||
tags
|
||||
- `skald narrate` renders a chapter to audio via one of three TTS
|
||||
engines (F5-TTS, Kokoro-82M, Tortoise-TTS) — see `engines/`
|
||||
- Named-author "soul" personas via `skald authors seed`; author
|
||||
voice replaces the model's base system prompt for gen/cleanup/
|
||||
rewrite/dedup/narrate_prep
|
||||
|
||||
## Schema (cheat sheet)
|
||||
|
||||
```
|
||||
stories → meta + status + parent/root for series
|
||||
characters → real or fictional, story-scoped
|
||||
canon_facts → setting, mystery, theme, rule, historical_anchor, hook
|
||||
chapters → full prose body
|
||||
chapter_summaries → short summaries for cheap context loading
|
||||
passages → paragraph-level + embedding vector(1536)
|
||||
generation_runs → every LLM call logged
|
||||
audit_findings → canon audit output (severity + area)
|
||||
tags → arbitrary labels
|
||||
stories meta + status + parent/root for series
|
||||
authors persona identity (slug, display_name, model)
|
||||
author_revisions versioned souls; one current per author
|
||||
characters real or fictional, story-scoped, voice-mappable
|
||||
canon_facts setting, mystery, theme, rule, historical_anchor, hook
|
||||
chapters full prose body + optional body_md_tts annotation
|
||||
chapter_summaries short summaries for cheap context loading
|
||||
passages paragraph-level + embedding vector(1536)
|
||||
voices TTS voice rows (F5 ref clips / Kokoro / Tortoise names)
|
||||
pronunciation_overrides per-story + global respellings for proper nouns
|
||||
generation_runs every LLM call logged
|
||||
audit_findings audit pass output (severity + area)
|
||||
narration_runs per-chapter TTS renders
|
||||
```
|
||||
|
||||
## Architecture (v0.1 + the plan)
|
||||
## Quickstart
|
||||
|
||||
```
|
||||
┌─────────────────────────────────┐
|
||||
│ skald container │
|
||||
│ ┌───────────┐ ┌────────────┐ │
|
||||
│ │ postgres │ │ skald-rust │ │
|
||||
│ │ pgvector │←─│ axum + cli │ │
|
||||
│ │ localhost │ │ :7780 │ │
|
||||
│ └───────────┘ └─────┬──────┘ │
|
||||
└─────────────────────────┼────────┘
|
||||
│ HTTP (future)
|
||||
↓
|
||||
┌──────────┐
|
||||
│clawdforge│
|
||||
└─────┬────┘
|
||||
↓
|
||||
opus calls
|
||||
```sh
|
||||
docker compose up -d
|
||||
docker exec skald skald import-markdown --path /seed/<story>.md \
|
||||
--title "<title>"
|
||||
curl http://localhost:7780/health
|
||||
```
|
||||
|
||||
v1.0+: extract postgres to its own container on db-net. skald
|
||||
becomes pure stateless rust, connects via `DATABASE_URL`. Migration
|
||||
is a connection-string change + a network move; the binary doesn't
|
||||
care where the DB lives.
|
||||
The compose file expects `POSTGRES_PASSWORD` and (optionally)
|
||||
`CLAWDFORGE_URL` + `CLAWDFORGE_TOKEN` in `.env`. Story markdown
|
||||
goes into `./seed/`; postgres data persists in `./pgdata/`.
|
||||
|
||||
## Architecture
|
||||
|
||||
v0.1 ships postgres inside the skald container — singleton until
|
||||
the tool stabilises. To extract postgres later, swap the runtime
|
||||
base to `debian:bookworm-slim`, drop `entrypoint.sh`, and point
|
||||
`DATABASE_URL` at the external pg. The binary doesn't care where
|
||||
the DB lives.
|
||||
|
||||
The generation passes call out to `clawdforge` (a bearer-token-gated
|
||||
HTTP wrapper around `claude -p`). The Rust client is vendored at
|
||||
`vendor/clawdforge/`. TTS calls go HTTP+JSON to the per-engine
|
||||
sidecars under `engines/`.
|
||||
|
||||
## License
|
||||
|
||||
MIT.
|
||||
MIT — see `LICENSE`.
|
||||
|
|
|
|||
51
compose.yml
51
compose.yml
|
|
@ -1,37 +1,42 @@
|
|||
# Standalone compose stack for skald v0.1. Postgres lives in the
|
||||
# same container — single deployable unit "till we have a real
|
||||
# working tool" (Sulkta's call, 2026-05-13).
|
||||
# Standalone compose stack for skald. Postgres lives inside the
|
||||
# skald container — single deployable unit until the tool stabilises.
|
||||
#
|
||||
# To deploy on the host:
|
||||
# sudo mkdir -p /srv/appdata/skald/{pgdata,seed}
|
||||
# sudo cp <story>.md /srv/appdata/skald/seed/
|
||||
# sudo cp skald.env /srv/appdata/secrets/skald.env # POSTGRES_PASSWORD=...
|
||||
# docker compose -p skald up -d
|
||||
# Set in .env (or the environment):
|
||||
# POSTGRES_PASSWORD=... # required
|
||||
# CLAWDFORGE_URL=http://...:8800 # if running gen / cleanup / audit
|
||||
# CLAWDFORGE_TOKEN=cf_...
|
||||
# SKALD_DATA=./pgdata # optional override; defaults to ./pgdata
|
||||
# SKALD_SEED=./seed # optional override; defaults to ./seed
|
||||
#
|
||||
# To import the first story:
|
||||
# docker exec skald skald import-markdown \
|
||||
# docker compose exec skald skald import-markdown \
|
||||
# --path /seed/<story>.md \
|
||||
# --title "<title>"
|
||||
name: skald
|
||||
|
||||
services:
|
||||
skald:
|
||||
image: registry.example.local:5000/skald:latest
|
||||
build: .
|
||||
image: skald:latest
|
||||
container_name: skald
|
||||
restart: unless-stopped
|
||||
ports:
|
||||
- "7780:7780"
|
||||
env_file:
|
||||
- /srv/appdata/secrets/skald.env
|
||||
environment:
|
||||
POSTGRES_PASSWORD: ${POSTGRES_PASSWORD:?POSTGRES_PASSWORD is required}
|
||||
POSTGRES_USER: ${POSTGRES_USER:-skald}
|
||||
POSTGRES_DB: ${POSTGRES_DB:-skald}
|
||||
DATABASE_URL: ${DATABASE_URL:-postgresql://skald:${POSTGRES_PASSWORD}@localhost:5432/skald}
|
||||
CLAWDFORGE_URL: ${CLAWDFORGE_URL:-}
|
||||
CLAWDFORGE_TOKEN: ${CLAWDFORGE_TOKEN:-}
|
||||
SKALD_MODEL: ${SKALD_MODEL:-opus}
|
||||
F5_TTS_URL: ${F5_TTS_URL:-}
|
||||
KOKORO_URL: ${KOKORO_URL:-}
|
||||
TORTOISE_URL: ${TORTOISE_URL:-}
|
||||
RUST_LOG: ${RUST_LOG:-info}
|
||||
SKALD_LOG_FORMAT: ${SKALD_LOG_FORMAT:-json}
|
||||
volumes:
|
||||
# Postgres data — persist across container recreates.
|
||||
- /srv/appdata/skald/pgdata:/var/lib/postgresql/data
|
||||
# Markdown corpus to import via `docker exec skald skald import-markdown`.
|
||||
- /srv/appdata/skald/seed:/seed:ro
|
||||
environment:
|
||||
RUST_LOG: ${RUST_LOG:-info}
|
||||
SKALD_LOG_FORMAT: json
|
||||
labels:
|
||||
org.sulkta.domain: "sulkta"
|
||||
org.sulkta.owner: "Sulkta"
|
||||
org.sulkta.managed-by: "compose"
|
||||
org.sulkta.role: "skald"
|
||||
- ${SKALD_DATA:-./pgdata}:/var/lib/postgresql/data
|
||||
# Markdown corpus to import via `skald import-markdown`.
|
||||
- ${SKALD_SEED:-./seed}:/seed:ro
|
||||
|
|
|
|||
|
|
@ -40,19 +40,18 @@ generalise. Examples:
|
|||
all three at once), preset choice ergonomics, character→tortoise-
|
||||
voice seed assignments.
|
||||
|
||||
When deploying an engine to the host, the build dir at
|
||||
`/srv/appdata/<engine>/build/` tracks the engine's branch:
|
||||
To deploy a tuned engine, check out the engine's branch in the build
|
||||
dir and `docker compose up -d --build`:
|
||||
|
||||
```bash
|
||||
cd /srv/appdata/kokoro/build
|
||||
git fetch && git checkout engine/kokoro
|
||||
docker compose -p <name> up -d --build
|
||||
docker compose up -d --build
|
||||
```
|
||||
|
||||
## GPU coordination (2070 Super)
|
||||
## GPU coordination
|
||||
|
||||
The 8GB card is the bottleneck. F5 + Kokoro can co-reside (~5GB +
|
||||
~1GB). Tortoise pushes the budget over and needs the GPU largely
|
||||
to itself — the `engine/tortoise` branch will carry the script
|
||||
that stops kokoro + f5 before a tortoise run and restarts them
|
||||
after. Replace with proper coordination once we have more VRAM.
|
||||
On an 8GB card F5 + Kokoro can co-reside (~5GB + ~1GB). Tortoise
|
||||
pushes the budget over and needs the GPU largely to itself — the
|
||||
`engine/tortoise` branch carries a script to stop kokoro + f5
|
||||
before a Tortoise run and restart them after. Replace with proper
|
||||
coordination once more VRAM is available.
|
||||
|
|
|
|||
|
|
@ -1,9 +1,6 @@
|
|||
# Sulkta build of F5-TTS — upstream ghcr.io/swivid/f5-tts:main was
|
||||
# shipped with torch 2.11/torchaudio 2.4 ABI mismatch on 2026-05-13,
|
||||
# breaking import torchaudio at boot. We rebuild on a known-good
|
||||
# pytorch base + pip install f5-tts.
|
||||
#
|
||||
# Image tag in localhost-registry: registry.example.local:5000/f5-tts:<ver>
|
||||
# F5-TTS rebuild on a known-good pytorch base. Upstream
|
||||
# ghcr.io/swivid/f5-tts:main shipped a torch/torchaudio ABI mismatch
|
||||
# that broke `import torchaudio` at boot; this image bypasses that.
|
||||
#
|
||||
# License: Apache 2.0 (code) / CC-BY-NC (Emilia-trained weights).
|
||||
# Personal use OK; redistribution gray-area — flagged.
|
||||
|
|
@ -30,12 +27,11 @@ RUN pip install --no-cache-dir 'f5-tts>=1.0.0'
|
|||
# Pre-warm the HF cache directory.
|
||||
RUN mkdir -p /cache/hf /audio /voices
|
||||
|
||||
COPY f5_server.py /app/f5_server.py
|
||||
COPY server.py /app/server.py
|
||||
WORKDIR /app
|
||||
|
||||
EXPOSE 7860
|
||||
|
||||
# Skald talks to our purpose-built FastAPI server, not Gradio.
|
||||
# Models load at startup (first request would otherwise pay the
|
||||
# cold-start cost). uvicorn on :7860 to keep the port stable.
|
||||
CMD ["uvicorn", "f5_server:app", "--host", "0.0.0.0", "--port", "7860"]
|
||||
# Purpose-built FastAPI server, not Gradio. Models load at startup
|
||||
# so the first request doesn't pay the cold-start cost.
|
||||
CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "7860"]
|
||||
|
|
|
|||
|
|
@ -1,19 +1,21 @@
|
|||
# F5-TTS standalone stack on the host.
|
||||
# F5-TTS sidecar.
|
||||
#
|
||||
# License posture (acknowledged 2026-05-13): code is Apache 2.0, but
|
||||
# the pretrained model weights are CC-BY-NC (Emilia training data).
|
||||
# Personal listen is fine; public sharing is a flagged gray area.
|
||||
# Sulkta's call: ship anyway.
|
||||
# Code is Apache 2.0; pretrained F5TTS_v1_Base weights are CC-BY-NC
|
||||
# (Emilia training data). Personal use is fine; redistribution is a
|
||||
# flagged gray area.
|
||||
#
|
||||
# Runtime: 8GB GPU is plenty (F5 inference ~4-6GB peak).
|
||||
# First run downloads ~2GB of model weights from HuggingFace into
|
||||
# the hf-cache volume; subsequent runs are warm.
|
||||
#
|
||||
# First-run cost: ~2GB model download from HuggingFace into hf-cache,
|
||||
# happens on first inference request. Subsequent runs are warm.
|
||||
# Set in .env (or override):
|
||||
# F5_HOST_PORT=7792
|
||||
# F5_DATA=./data # ${F5_DATA}/hf-cache + voices + audio
|
||||
name: f5-tts
|
||||
|
||||
services:
|
||||
f5-tts:
|
||||
image: registry.example.local:5000/f5-tts:0.3
|
||||
build: .
|
||||
image: f5-tts:0.3
|
||||
container_name: f5-tts
|
||||
restart: unless-stopped
|
||||
deploy:
|
||||
|
|
@ -24,20 +26,11 @@ services:
|
|||
count: all
|
||||
capabilities: [gpu]
|
||||
ports:
|
||||
- "127.0.0.1:7792:7860"
|
||||
- "127.0.0.1:7792:7860"
|
||||
- "${F5_HOST_PORT:-7792}:7860"
|
||||
volumes:
|
||||
# HF model weights cache — persists ~2GB after first download.
|
||||
- /srv/appdata/f5-tts/hf-cache:/cache/hf
|
||||
# Reference voice clips (lj_speech.wav, etc).
|
||||
- /srv/appdata/f5-tts/voices:/voices:ro
|
||||
# Rendered audio output — skald writes story narrations here.
|
||||
- /srv/appdata/f5-tts/audio:/audio
|
||||
- ${F5_DATA:-./data}/hf-cache:/cache/hf
|
||||
- ${F5_DATA:-./data}/voices:/voices:ro
|
||||
- ${F5_DATA:-./data}/audio:/audio
|
||||
environment:
|
||||
HF_HOME: /cache/hf
|
||||
HF_HUB_DISABLE_TELEMETRY: "1"
|
||||
labels:
|
||||
org.sulkta.domain: "sulkta"
|
||||
org.sulkta.owner: "Sulkta"
|
||||
org.sulkta.managed-by: "compose"
|
||||
org.sulkta.role: "f5-tts"
|
||||
|
|
|
|||
|
|
@ -1,13 +1,8 @@
|
|||
# Sulkta build of Kokoro-82M TTS.
|
||||
# Kokoro-82M TTS. Apache 2.0 code AND weights — clean stack vs
|
||||
# F5-TTS's CC-BY-NC asterisk.
|
||||
#
|
||||
# License: Apache 2.0 (code AND model weights). Clean stack — no
|
||||
# CC-BY-NC asterisk like F5-TTS's Emilia weights. This is the
|
||||
# narrator engine for sleep-quality audiobook reads; F5-TTS stays
|
||||
# around for voice-cloning cases.
|
||||
#
|
||||
# Kokoro is small enough to run on CPU but we use the cuda base
|
||||
# anyway to stay consistent with f5-tts and so it'll pick up the
|
||||
# GPU when no other tenant has it.
|
||||
# Kokoro runs fine on CPU but we use the cuda base to stay
|
||||
# consistent with f5-tts and pick up the GPU when free.
|
||||
FROM pytorch/pytorch:2.6.0-cuda12.4-cudnn9-runtime
|
||||
|
||||
ENV DEBIAN_FRONTEND=noninteractive \
|
||||
|
|
@ -27,9 +22,9 @@ RUN pip install --no-cache-dir 'kokoro>=0.9.0' 'fastapi>=0.115.0' 'uvicorn>=0.32
|
|||
|
||||
RUN mkdir -p /cache/hf /audio
|
||||
|
||||
COPY kokoro_server.py /app/kokoro_server.py
|
||||
COPY server.py /app/server.py
|
||||
WORKDIR /app
|
||||
|
||||
EXPOSE 7860
|
||||
|
||||
CMD ["uvicorn", "kokoro_server:app", "--host", "0.0.0.0", "--port", "7860"]
|
||||
CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "7860"]
|
||||
|
|
|
|||
|
|
@ -1,16 +1,18 @@
|
|||
# Kokoro-82M TTS stack on the host.
|
||||
# Kokoro-82M TTS sidecar.
|
||||
#
|
||||
# Audiobook-quality narrator engine (Apache 2.0 code + weights —
|
||||
# clean stack vs F5-TTS's CC-BY-NC asterisk). Sibling to f5-tts;
|
||||
# both share /srv/appdata/f5-tts/audio so skald's audio
|
||||
# route serves outputs from either engine through the same path.
|
||||
# Apache 2.0 code AND model weights — clean stack for share/publish.
|
||||
# Audiobook-quality narrator; F5-TTS stays around for voice-cloning.
|
||||
#
|
||||
# License: Apache 2.0 top to bottom. Right for share/publish.
|
||||
# Set in .env (or override):
|
||||
# KOKORO_HOST_PORT=7794
|
||||
# KOKORO_DATA=./data # ${KOKORO_DATA}/hf-cache
|
||||
# AUDIO_DIR=../f5-tts/data/audio # shared output dir across engines
|
||||
name: kokoro
|
||||
|
||||
services:
|
||||
kokoro:
|
||||
image: registry.example.local:5000/kokoro:0.5
|
||||
build: .
|
||||
image: kokoro:0.5
|
||||
container_name: kokoro
|
||||
restart: unless-stopped
|
||||
deploy:
|
||||
|
|
@ -21,17 +23,10 @@ services:
|
|||
count: all
|
||||
capabilities: [gpu]
|
||||
ports:
|
||||
- "127.0.0.1:7794:7860"
|
||||
- "127.0.0.1:7794:7860"
|
||||
- "${KOKORO_HOST_PORT:-7794}:7860"
|
||||
volumes:
|
||||
- /srv/appdata/kokoro/hf-cache:/cache/hf
|
||||
# Shared with f5-tts so skald's /audio route covers both.
|
||||
- /srv/appdata/f5-tts/audio:/audio
|
||||
- ${KOKORO_DATA:-./data}/hf-cache:/cache/hf
|
||||
- ${AUDIO_DIR:-./data/audio}:/audio
|
||||
environment:
|
||||
HF_HOME: /cache/hf
|
||||
HF_HUB_DISABLE_TELEMETRY: "1"
|
||||
labels:
|
||||
org.sulkta.domain: "sulkta"
|
||||
org.sulkta.owner: "Sulkta"
|
||||
org.sulkta.managed-by: "compose"
|
||||
org.sulkta.role: "kokoro"
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
"""Kokoro-82M FastAPI server, sibling to f5_server.
|
||||
"""Kokoro-82M FastAPI server, sibling to the f5-tts server.
|
||||
|
||||
Same /synthesize contract as F5 so skald can route between engines
|
||||
just by which URL it points at. The semantic difference: Kokoro
|
||||
|
|
@ -234,7 +234,7 @@ def _startup() -> None:
|
|||
|
||||
@app.get("/healthz")
|
||||
def healthz() -> dict:
|
||||
# Shape matches f5_server's so the same Rust HealthResponse
|
||||
# Shape matches the f5-tts server's so the same Rust HealthResponse
|
||||
# struct deserializes both: model/vocoder/loaded fields are
|
||||
# required by skald-core::narrate::HealthResponse.
|
||||
return {
|
||||
|
|
|
|||
|
|
@ -1,17 +1,13 @@
|
|||
# Sulkta build of Tortoise-TTS.
|
||||
# Tortoise-TTS. Apache 2.0 code + weights.
|
||||
#
|
||||
# Voice roster (built-in, no cloning needed): angie, daniel, deniro,
|
||||
# emma, freeman, geralt, halle, jlaw, lj, mol, myself, pat, pat2,
|
||||
# rainbow, snakes, tim_reynolds, tom, train_atkins, train_dotrice,
|
||||
# ~26 built-in voices (no cloning): angie, daniel, deniro, emma,
|
||||
# freeman, geralt, halle, jlaw, lj, mol, myself, pat, pat2, rainbow,
|
||||
# snakes, tim_reynolds, tom, train_atkins, train_dotrice,
|
||||
# train_dreams, train_grace, train_kennard, train_lescault,
|
||||
# train_mouse, weaver, william. ~26 voices baked in.
|
||||
# train_mouse, weaver, william.
|
||||
#
|
||||
# License: Apache 2.0 (code) + Apache 2.0 (model weights). Clean
|
||||
# stack for share/publish.
|
||||
#
|
||||
# Speed: slow. Trade for quality. Standard preset is ~10x slower
|
||||
# than Kokoro; high_quality is ~30x slower. Worth it for the
|
||||
# audiobook-quality bar.
|
||||
# Slow: standard preset is ~10x slower than Kokoro; high_quality is
|
||||
# ~30x. Trade for quality.
|
||||
|
||||
FROM pytorch/pytorch:2.6.0-cuda12.4-cudnn9-runtime
|
||||
|
||||
|
|
@ -37,9 +33,9 @@ RUN pip install --no-cache-dir \
|
|||
|
||||
RUN mkdir -p /cache/hf /cache/tortoise-models /audio
|
||||
|
||||
COPY tortoise_server.py /app/tortoise_server.py
|
||||
COPY server.py /app/server.py
|
||||
WORKDIR /app
|
||||
|
||||
EXPOSE 7860
|
||||
|
||||
CMD ["uvicorn", "tortoise_server:app", "--host", "0.0.0.0", "--port", "7860"]
|
||||
CMD ["uvicorn", "server:app", "--host", "0.0.0.0", "--port", "7860"]
|
||||
|
|
|
|||
|
|
@ -1,19 +1,22 @@
|
|||
# Tortoise-TTS stack on the host. Audiobook-quality engine with 25+
|
||||
# named voices (no cloning). Apache 2.0 top to bottom.
|
||||
# Tortoise-TTS sidecar. 25+ named voices, no cloning needed.
|
||||
# Apache 2.0 top to bottom.
|
||||
#
|
||||
# Slow: ~10x kokoro wall clock at 'standard' preset. Worth it for
|
||||
# the quality bar. Sulkta's call 2026-05-14: "use higgs (now tortoise)
|
||||
# and we will only let it use the full gpu for runs" — translated:
|
||||
# runs are batched, slow is acceptable.
|
||||
# Slow: ~10x kokoro wall-clock at 'standard' preset. Worth it for the
|
||||
# quality bar; runs are batched.
|
||||
#
|
||||
# Co-resides with kokoro on the 2070 Super since tortoise is ~5GB
|
||||
# and kokoro is ~1GB (8GB total). If OOM hits during a render,
|
||||
# we'll add a coordination layer to pause kokoro first.
|
||||
# Co-resides with kokoro on an 8GB card (tortoise ~5GB + kokoro ~1GB).
|
||||
# OOM during a render: add a coordinator that pauses kokoro first.
|
||||
#
|
||||
# Set in .env (or override):
|
||||
# TORTOISE_HOST_PORT=7795
|
||||
# TORTOISE_DATA=./data # ${TORTOISE_DATA}/{hf-cache,models}
|
||||
# AUDIO_DIR=../f5-tts/data/audio # shared output dir across engines
|
||||
name: tortoise
|
||||
|
||||
services:
|
||||
tortoise:
|
||||
image: registry.example.local:5000/tortoise:0.1
|
||||
build: .
|
||||
image: tortoise:0.1
|
||||
container_name: tortoise
|
||||
restart: unless-stopped
|
||||
deploy:
|
||||
|
|
@ -24,20 +27,12 @@ services:
|
|||
count: all
|
||||
capabilities: [gpu]
|
||||
ports:
|
||||
- "127.0.0.1:7795:7860"
|
||||
- "127.0.0.1:7795:7860"
|
||||
- "${TORTOISE_HOST_PORT:-7795}:7860"
|
||||
volumes:
|
||||
- /srv/appdata/tortoise/hf-cache:/cache/hf
|
||||
- /srv/appdata/tortoise/models:/cache/tortoise-models
|
||||
# Shared audio dir with f5/kokoro so skald serves all engines'
|
||||
# outputs through the same /audio route.
|
||||
- /srv/appdata/f5-tts/audio:/audio
|
||||
- ${TORTOISE_DATA:-./data}/hf-cache:/cache/hf
|
||||
- ${TORTOISE_DATA:-./data}/models:/cache/tortoise-models
|
||||
- ${AUDIO_DIR:-./data/audio}:/audio
|
||||
environment:
|
||||
HF_HOME: /cache/hf
|
||||
HF_HUB_DISABLE_TELEMETRY: "1"
|
||||
TORTOISE_MODELS_DIR: /cache/tortoise-models
|
||||
labels:
|
||||
org.sulkta.domain: "sulkta"
|
||||
org.sulkta.owner: "Sulkta"
|
||||
org.sulkta.managed-by: "compose"
|
||||
org.sulkta.role: "tortoise-tts"
|
||||
|
|
|
|||
|
|
@ -1,4 +1,4 @@
|
|||
"""Tortoise-TTS FastAPI server. Sibling to kokoro_server.
|
||||
"""Tortoise-TTS FastAPI server. Sibling to the kokoro server.
|
||||
|
||||
Same /synthesize contract as the kokoro server so skald only has to
|
||||
route by voice.source. Differences:
|
||||
|
|
@ -71,7 +71,7 @@ def _get_voice(name: str) -> tuple:
|
|||
return _voice_cache[name]
|
||||
|
||||
|
||||
# ─── tag splitter (lifted from kokoro_server) ───────────────────
|
||||
# ─── tag splitter (lifted from the kokoro server) ───────────────
|
||||
|
||||
|
||||
class Node:
|
||||
|
|
@ -209,7 +209,7 @@ def _startup() -> None:
|
|||
|
||||
@app.get("/healthz")
|
||||
def healthz() -> dict:
|
||||
# Shape matches f5_server/kokoro_server so skald's HealthResponse
|
||||
# Shape matches the f5-tts + kokoro servers so skald's HealthResponse
|
||||
# struct deserializes all three.
|
||||
return {
|
||||
"ok": True,
|
||||
|
|
|
|||
|
|
@ -1,14 +1,9 @@
|
|||
#!/usr/bin/env bash
|
||||
# Skald container entrypoint.
|
||||
#
|
||||
# Boots the embedded postgres via the pgvector image's own
|
||||
# docker-entrypoint, waits for it to accept connections, then execs
|
||||
# `skald` in the foreground. Tini is PID 1 (so it can reap zombies +
|
||||
# forward signals); we are PID 2; postgres becomes our child.
|
||||
#
|
||||
# This is explicitly "DB in the same container, for now" — when we
|
||||
# split the DB out (see project notes), the entrypoint reduces to
|
||||
# `exec /usr/local/bin/skald "$@"` and the pg startup goes away.
|
||||
# `skald` in the foreground. Tini is PID 1; postgres becomes our
|
||||
# child. When the DB is extracted to its own container, this reduces
|
||||
# to `exec /usr/local/bin/skald "$@"`.
|
||||
|
||||
set -eo pipefail
|
||||
|
||||
|
|
|
|||
|
|
@ -8,24 +8,23 @@ use serde::{Deserialize, Serialize};
|
|||
|
||||
#[derive(Debug, Clone, Serialize, Deserialize)]
|
||||
pub struct ForgeConfig {
|
||||
/// Base URL of the clawdforge HTTP service. Defaults to
|
||||
/// `http://clawdforge.example.local:8800` in production; override
|
||||
/// for tests via env.
|
||||
/// Base URL of the clawdforge HTTP service. The calling binary
|
||||
/// resolves this from `CLAWDFORGE_URL`.
|
||||
pub base_url: String,
|
||||
|
||||
/// App-level bearer token. Resolved by the binary from
|
||||
/// `CLAWDFORGE_TOKEN`; should never be logged or `Display`ed.
|
||||
pub app_token: String,
|
||||
|
||||
/// Model alias passed to clawdforge → `claude -p --model`. Skald
|
||||
/// is opinionated: always opus max effort. Default reflects that.
|
||||
/// Model alias passed to clawdforge → `claude -p --model`.
|
||||
/// Defaults to opus.
|
||||
pub model: String,
|
||||
}
|
||||
|
||||
impl Default for ForgeConfig {
|
||||
fn default() -> Self {
|
||||
Self {
|
||||
base_url: "http://clawdforge.example.local:8800".into(),
|
||||
base_url: "http://localhost:8800".into(),
|
||||
app_token: String::new(),
|
||||
model: "opus".into(),
|
||||
}
|
||||
|
|
|
|||
|
|
@ -27,7 +27,7 @@ use uuid::Uuid;
|
|||
|
||||
#[derive(Debug, Clone)]
|
||||
pub struct F5Config {
|
||||
/// e.g. http://127.0.0.1:7792
|
||||
/// e.g. http://localhost:7792
|
||||
pub base_url: String,
|
||||
/// Inference subprocess timeout. Long-form chapters (3000 words)
|
||||
/// take 60-180s on an 8GB GPU; cap at 1800s to match clawdforge.
|
||||
|
|
|
|||
|
|
@ -30,8 +30,8 @@ use uuid::Uuid;
|
|||
about = "Long-form story-writer. Database is the source of truth; the writer is the tooling."
|
||||
)]
|
||||
struct Cli {
|
||||
/// Postgres connection URL. Defaults to `postgresql://skald:skald@localhost:5432/skald`.
|
||||
#[arg(long, env = "DATABASE_URL", default_value = "postgresql://skald:skald@localhost:5432/skald")]
|
||||
/// Postgres connection URL. Read from `DATABASE_URL` if unset.
|
||||
#[arg(long, env = "DATABASE_URL")]
|
||||
database_url: String,
|
||||
|
||||
#[command(subcommand)]
|
||||
|
|
|
|||
|
|
@ -458,16 +458,15 @@ async fn cleanup_superseded_renders(pool: &PgPool, chapter_id: Uuid, current_run
|
|||
/// kokoro_* → KOKORO_URL
|
||||
/// tortoise_* → TORTOISE_URL
|
||||
/// anything else (lj_speech etc.) → F5_TTS_URL
|
||||
/// Each env var has a LAN-default for the host.
|
||||
fn engine_url_for(source: &str) -> anyhow::Result<String> {
|
||||
if source.starts_with("kokoro") {
|
||||
Ok(std::env::var("KOKORO_URL")
|
||||
.unwrap_or_else(|_| "http://127.0.0.1:7794".into()))
|
||||
let (env_var, engine) = if source.starts_with("kokoro") {
|
||||
("KOKORO_URL", "kokoro")
|
||||
} else if source.starts_with("tortoise") {
|
||||
Ok(std::env::var("TORTOISE_URL")
|
||||
.unwrap_or_else(|_| "http://127.0.0.1:7795".into()))
|
||||
("TORTOISE_URL", "tortoise")
|
||||
} else {
|
||||
Ok(std::env::var("F5_TTS_URL")
|
||||
.unwrap_or_else(|_| "http://127.0.0.1:7792".into()))
|
||||
}
|
||||
("F5_TTS_URL", "f5-tts")
|
||||
};
|
||||
std::env::var(env_var).map_err(|_| {
|
||||
anyhow::anyhow!("{env_var} not set — point at the {engine} sidecar")
|
||||
})
|
||||
}
|
||||
|
|
|
|||
|
|
@ -201,8 +201,12 @@ async fn new_story_create(
|
|||
// parent to compare against. So you get a single first-chapter
|
||||
// gen + cleanup pass and status flows to 'complete'.
|
||||
if form.fire == "now" {
|
||||
let database_url = std::env::var("DATABASE_URL")
|
||||
.unwrap_or_else(|_| "postgresql://skald:skald@localhost:5432/skald".into());
|
||||
let Ok(database_url) = std::env::var("DATABASE_URL") else {
|
||||
return Err((
|
||||
StatusCode::INTERNAL_SERVER_ERROR,
|
||||
"DATABASE_URL not set — cannot spawn background gen".into(),
|
||||
));
|
||||
};
|
||||
let author_owned = if author_slug.is_empty() {
|
||||
None
|
||||
} else {
|
||||
|
|
@ -337,7 +341,12 @@ async fn continue_create(
|
|||
// If user clicked "fire now," spawn a background gen task.
|
||||
// Otherwise the sequel sits in seed state until CLI fires it.
|
||||
if form.fire == "now" {
|
||||
let database_url = std::env::var("DATABASE_URL").unwrap_or_else(|_| "postgresql://skald:skald@localhost:5432/skald".into());
|
||||
let Ok(database_url) = std::env::var("DATABASE_URL") else {
|
||||
return Err((
|
||||
StatusCode::INTERNAL_SERVER_ERROR,
|
||||
"DATABASE_URL not set — cannot spawn background gen".into(),
|
||||
));
|
||||
};
|
||||
let author_owned = if author_slug.is_empty() { None } else { Some(author_slug.to_string()) };
|
||||
let direction_owned = direction.clone();
|
||||
let chapters = parse_chapters(&form.chapters);
|
||||
|
|
@ -512,8 +521,12 @@ async fn chapter_narrate_fire(
|
|||
let chapter_id =
|
||||
chapter_id.ok_or((StatusCode::NOT_FOUND, "chapter not found".into()))?;
|
||||
|
||||
let database_url = std::env::var("DATABASE_URL")
|
||||
.unwrap_or_else(|_| "postgresql://skald:skald@localhost:5432/skald".into());
|
||||
let Ok(database_url) = std::env::var("DATABASE_URL") else {
|
||||
return Err((
|
||||
StatusCode::INTERNAL_SERVER_ERROR,
|
||||
"DATABASE_URL not set — cannot spawn background narrate".into(),
|
||||
));
|
||||
};
|
||||
tokio::spawn(async move {
|
||||
if let Err(e) = crate::narrate::run(&database_url, chapter_id, None, 1.0).await {
|
||||
tracing::error!(chapter_id = %chapter_id, error = %e, "background narrate failed");
|
||||
|
|
@ -671,7 +684,7 @@ fn render_shell(stories: &[StoryRow], current: Option<Uuid>, main: Markup) -> Ma
|
|||
}
|
||||
footer.footbar {
|
||||
span { "skald · v0.3 · written down · "
|
||||
a href="http://127.0.0.1:3001/Sulkta/skald" { "Sulkta/skald" }
|
||||
a href="https://git.sulkta.com/Sulkta/skald" { "Sulkta/skald" }
|
||||
}
|
||||
}
|
||||
}
|
||||
|
|
|
|||
Loading…
Add table
Add a link
Reference in a new issue