L'électron rare
d2fed6087a
- Introduced `mcp-server.js` to expose KXKM personas as MCP tools, supporting chat, persona listing, web search, and status checks. - Implemented `mcp-server-smoke.js` for testing the MCP server functionality, ensuring compatibility with both new and legacy message formats. - Created `setup-voice-clone.sh` for managing voice cloning environment setup, including bootstrapping, sample generation, and smoke testing. - Added `state.json` to track project status and task outputs for various batches. - Generated summary files for deep cycle and overall project status, capturing performance and security findings.
257 lines
6.6 KiB
Bash
257 lines
6.6 KiB
Bash
#!/usr/bin/env bash
|
|
set -euo pipefail
|
|
|
|
ROOT_DIR="$(cd "$(dirname "${BASH_SOURCE[0]}")/.." && pwd)"
|
|
JSON_MODE=0
|
|
VERBOSE=0
|
|
YES=0
|
|
PERSONA="${VOICE_PERSONA:-pharmacius}"
|
|
PYTHON_CMD="${VOICE_CLONE_PYTHON:-}"
|
|
VENV_DIR="${VOICE_CLONE_VENV_DIR:-$ROOT_DIR/.venvs/voice-clone}"
|
|
|
|
usage() {
|
|
cat <<'EOF'
|
|
Usage: scripts/setup-voice-clone.sh <subcommand> [options]
|
|
|
|
Subcommands:
|
|
bootstrap Create/update the local XTTS venv and install runtime deps
|
|
sample Generate a local persona sample with Piper using the local venv
|
|
smoke Run a synthetic XTTS smoke using the local venv and sample
|
|
all Run bootstrap, sample and smoke in sequence
|
|
help Show this help
|
|
|
|
Options:
|
|
--persona NICK Persona/sample to target (default: pharmacius)
|
|
--python CMD Python interpreter used to create the venv (default: python3.12 if available)
|
|
--venv-dir PATH Virtualenv directory (default: .venvs/voice-clone)
|
|
--json Emit machine-readable log lines when possible
|
|
--verbose Print executed commands
|
|
--yes Accepted for CLI parity
|
|
--help Show this help
|
|
EOF
|
|
}
|
|
|
|
log() {
|
|
local message="$1"
|
|
if [[ "$JSON_MODE" -eq 1 ]]; then
|
|
python3 - "$message" <<'PY'
|
|
import json, sys
|
|
print(json.dumps({"level": "info", "msg": sys.argv[1]}))
|
|
PY
|
|
else
|
|
printf '[voice-clone-setup] %s\n' "$message"
|
|
fi
|
|
}
|
|
|
|
die() {
|
|
local message="$1"
|
|
if [[ "$JSON_MODE" -eq 1 ]]; then
|
|
python3 - "$message" <<'PY'
|
|
import json, sys
|
|
print(json.dumps({"level": "error", "msg": sys.argv[1]}))
|
|
PY
|
|
else
|
|
printf '[voice-clone-setup] error: %s\n' "$message" >&2
|
|
fi
|
|
exit 1
|
|
}
|
|
|
|
run_cmd() {
|
|
if [[ "$VERBOSE" -eq 1 ]]; then
|
|
log "$*"
|
|
fi
|
|
"$@"
|
|
}
|
|
|
|
resolve_python_cmd() {
|
|
if [[ -n "$PYTHON_CMD" ]]; then
|
|
printf '%s\n' "$PYTHON_CMD"
|
|
return 0
|
|
fi
|
|
|
|
if command -v python3.12 >/dev/null 2>&1; then
|
|
printf '%s\n' "python3.12"
|
|
return 0
|
|
fi
|
|
|
|
if command -v python3 >/dev/null 2>&1; then
|
|
printf '%s\n' "python3"
|
|
return 0
|
|
fi
|
|
|
|
die "python3.12 or python3 is required"
|
|
}
|
|
|
|
resolve_voice_sample_dir() {
|
|
if [[ -n "${KXKM_VOICE_SAMPLES_DIR:-}" ]]; then
|
|
python3 - "$KXKM_VOICE_SAMPLES_DIR" <<'PY'
|
|
import os, sys
|
|
print(os.path.abspath(sys.argv[1]))
|
|
PY
|
|
return 0
|
|
fi
|
|
|
|
if [[ -n "${KXKM_LOCAL_DATA_DIR:-}" ]]; then
|
|
python3 - "$KXKM_LOCAL_DATA_DIR" <<'PY'
|
|
import os, sys
|
|
print(os.path.abspath(os.path.join(sys.argv[1], "voice-samples")))
|
|
PY
|
|
return 0
|
|
fi
|
|
|
|
python3 - "$ROOT_DIR" <<'PY'
|
|
import os, sys
|
|
print(os.path.join(sys.argv[1], "data", "voice-samples"))
|
|
PY
|
|
}
|
|
|
|
bootstrap_env() {
|
|
local python_cmd="$1"
|
|
command -v uv >/dev/null 2>&1 || die "uv is required"
|
|
|
|
if [[ ! -x "$VENV_DIR/bin/python" ]]; then
|
|
log "creating venv -> $VENV_DIR"
|
|
run_cmd uv venv --python "$python_cmd" "$VENV_DIR"
|
|
else
|
|
log "reusing venv -> $VENV_DIR"
|
|
fi
|
|
|
|
log "installing XTTS runtime deps into $(basename "$VENV_DIR")"
|
|
run_cmd uv pip install --python "$VENV_DIR/bin/python" torch torchaudio torchcodec 'transformers<5' coqui-tts piper-tts
|
|
}
|
|
|
|
generate_sample() {
|
|
local python_bin="$1"
|
|
local sample_dir="$2"
|
|
mkdir -p "$sample_dir"
|
|
log "generating sample for $PERSONA"
|
|
(cd "$ROOT_DIR" && PYTHON_BIN="$python_bin" node scripts/generate-voice-samples.js --persona "$PERSONA")
|
|
}
|
|
|
|
smoke_xtts() {
|
|
local python_bin="$1"
|
|
local sample_dir="$2"
|
|
local sample_basename
|
|
sample_basename="$(printf '%s' "$PERSONA" | tr '[:upper:]' '[:lower:]' | sed -E 's/[^a-z0-9_-]/_/g' | cut -c1-64)"
|
|
local sample_path="$sample_dir/${sample_basename}.wav"
|
|
local tmp_output
|
|
local tmp_json
|
|
|
|
[[ "${COQUI_TOS_AGREED:-}" == "1" ]] || die "XTTS smoke requires COQUI_TOS_AGREED=1 after reviewing the CPML terms"
|
|
[[ -f "$sample_path" ]] || die "sample missing: $sample_path"
|
|
tmp_output="$(mktemp "${TMPDIR:-/tmp}/kxkm-voice-clone-audio.XXXXXX")"
|
|
tmp_json="$(mktemp "${TMPDIR:-/tmp}/kxkm-voice-clone-json.XXXXXX")"
|
|
trap 'rm -f "${tmp_output:-}" "${tmp_json:-}"' EXIT
|
|
|
|
log "running XTTS smoke for $PERSONA"
|
|
(
|
|
cd "$ROOT_DIR"
|
|
"$python_bin" scripts/tts_clone_voice.py \
|
|
--text "Bonjour, ceci est un test de clonage pour ${PERSONA}." \
|
|
--reference "$sample_path" \
|
|
--output "$tmp_output"
|
|
) | tee "$tmp_json" >/dev/null
|
|
|
|
python3 - "$tmp_json" "$tmp_output" <<'PY'
|
|
import json
|
|
import pathlib
|
|
import sys
|
|
|
|
lines = pathlib.Path(sys.argv[1]).read_text(encoding="utf-8").splitlines()
|
|
payload = None
|
|
for line in reversed(lines):
|
|
candidate = line.strip()
|
|
if not candidate:
|
|
continue
|
|
try:
|
|
payload = json.loads(candidate)
|
|
break
|
|
except json.JSONDecodeError:
|
|
continue
|
|
|
|
if payload is None:
|
|
raise SystemExit("XTTS smoke produced no JSON status line")
|
|
|
|
output_path = pathlib.Path(sys.argv[2])
|
|
if payload.get("status") != "completed":
|
|
raise SystemExit(payload.get("error") or "XTTS smoke failed")
|
|
if not output_path.exists() or output_path.stat().st_size == 0:
|
|
raise SystemExit("XTTS smoke produced no audio output")
|
|
print(json.dumps({"status": "completed", "output": str(output_path), "bytes": output_path.stat().st_size}))
|
|
PY
|
|
}
|
|
|
|
SUBCOMMAND="${1:-all}"
|
|
if [[ $# -gt 0 ]]; then
|
|
shift
|
|
fi
|
|
|
|
while [[ $# -gt 0 ]]; do
|
|
case "$1" in
|
|
--persona)
|
|
shift
|
|
[[ $# -gt 0 ]] || die "--persona requires a value"
|
|
PERSONA="$1"
|
|
;;
|
|
--python)
|
|
shift
|
|
[[ $# -gt 0 ]] || die "--python requires a value"
|
|
PYTHON_CMD="$1"
|
|
;;
|
|
--venv-dir)
|
|
shift
|
|
[[ $# -gt 0 ]] || die "--venv-dir requires a value"
|
|
VENV_DIR="$1"
|
|
;;
|
|
--json) JSON_MODE=1 ;;
|
|
--verbose) VERBOSE=1 ;;
|
|
--yes) YES=1 ;;
|
|
--help|-h) usage; exit 0 ;;
|
|
*) usage; die "unknown option: $1" ;;
|
|
esac
|
|
shift
|
|
done
|
|
|
|
case "$SUBCOMMAND" in
|
|
bootstrap|sample|smoke|all|help) ;;
|
|
*) usage; die "unknown subcommand: $SUBCOMMAND" ;;
|
|
esac
|
|
|
|
if [[ "$SUBCOMMAND" == "help" ]]; then
|
|
usage
|
|
exit 0
|
|
fi
|
|
|
|
PYTHON_CMD="$(resolve_python_cmd)"
|
|
VENV_DIR="$(cd "$ROOT_DIR" && python3 - "$VENV_DIR" <<'PY'
|
|
import os, sys
|
|
print(os.path.abspath(sys.argv[1]))
|
|
PY
|
|
)"
|
|
VOICE_SAMPLE_DIR="$(resolve_voice_sample_dir)"
|
|
VENV_PYTHON="$VENV_DIR/bin/python"
|
|
|
|
log "persona=$PERSONA"
|
|
log "python=$PYTHON_CMD"
|
|
log "venv=$VENV_DIR"
|
|
log "voice-sample-dir=$VOICE_SAMPLE_DIR"
|
|
|
|
case "$SUBCOMMAND" in
|
|
bootstrap)
|
|
bootstrap_env "$PYTHON_CMD"
|
|
;;
|
|
sample)
|
|
[[ -x "$VENV_PYTHON" ]] || die "venv missing: run bootstrap first"
|
|
generate_sample "$VENV_PYTHON" "$VOICE_SAMPLE_DIR"
|
|
;;
|
|
smoke)
|
|
[[ -x "$VENV_PYTHON" ]] || die "venv missing: run bootstrap first"
|
|
smoke_xtts "$VENV_PYTHON" "$VOICE_SAMPLE_DIR"
|
|
;;
|
|
all)
|
|
bootstrap_env "$PYTHON_CMD"
|
|
generate_sample "$VENV_PYTHON" "$VOICE_SAMPLE_DIR"
|
|
smoke_xtts "$VENV_PYTHON" "$VOICE_SAMPLE_DIR"
|
|
;;
|
|
esac
|