mirror of
https://github.com/nomadkaraoke/karaoke-gen.git
synced 2026-11-20 13:07:57 +09:00
feat(auto-correct): proactive multi-model suggestions + cost logging + de-AI-ified review UI (#816)
## Summary Makes AI lyrics auto-correction run **automatically on every job** (multi-model), so suggestions are cached and ready the moment the reviewer opens the lyrics UI — plus token/cost instrumentation and a UI cleanup that removes all "AI"/"model" wording. Measured cost first (`scripts/measure_autocorrect_cost.py`): **~$0.13/track** multi-model on normal songs — cheap enough to run proactively. ## Changes **Cost instrumentation** - Capture per-model `usage` from Anthropic + Gemini; log greppable `auto-correct usage` / `auto-correct cost … total_est_cost_usd=…` lines from `suggest()`. - `backend/services/auto_correct/pricing.py` (Anthropic rates authoritative; Gemini estimated) + `scripts/measure_autocorrect_cost.py` offline harness. **Proactive generation (every job)** - `lyrics_worker._run_proactive_auto_correct()` runs the multi-model path after `mark_lyrics_complete` (downstream already triggered → non-blocking), gated by `AUTO_CORRECT_PROACTIVE_ENABLED`, bounded by a 180s timeout, and **swallows every error** — it never blocks or fails the karaoke job (flaky AI / out of credits / timeout all degrade gracefully). - Cache-key aligned with the UI request, so the review page's on-load call is an instant cache hit (no double spend). - `ci.yml`: wires `AUTO_CORRECT_*` env + the `anthropic-api-key` secret onto `lyrics-transcription-job` (same SA as the API service — no IAM change). **Review UI** - Always multi-model (removed the model-selection toggle); button renamed **"Auto-correct"**; stripped model/consensus/cached attribution badges; reworded strings to drop "AI"/"model" (translated to all 32 locales). - Auto-runs once on review load so suggestions are ready without a click; the button stays for the paste-references-later flow. ## Testing - Backend: **3460 passed** (`backend/tests/`, emulator excluded) — incl. new `test_pricing.py` + `test_proactive_auto_correct.py` (gating, graceful failure, cache alignment). - Frontend: **1006 jest tests pass**; type-clean; 33-locale key parity verified. ## Notes / follow-ups - Proactive activates on the **next deploy** (the lyrics-job env/secret apply via this PR's ci.yml on merge). - Left "AI" wording in two *separate* features untouched (different subsystems): `lyricsReview.agenticMetrics` and `lyricsReview.modals.correctionAnnotation` — flagged for a separate decision. - The production E2E spec's click-to-open flow may want a tweak now that suggestions auto-run on load. @coderabbitai ignore --------- Co-authored-by: Claude Opus 4.8 (1M context) <noreply@anthropic.com>
This commit is contained in:
1 parent
0dd5d404e5
commit
364e4ebd5f
53 files changed
+1308
-386
No files matched your search
@@ -1676,10 +1676,17 @@ jobs:
|
||||
--quiet
|
||||
|
||||
# Lyrics transcription job (CPU, us-central1)
|
||||
# Also wire the auto-correct env + Anthropic key so the worker can
|
||||
# proactively pre-generate multi-model suggestions (cached for the
|
||||
# review UI). --update-* merges, preserving the job's existing vars.
|
||||
# Same service account as the API service, which already reads
|
||||
# anthropic-api-key, so no new IAM is required.
|
||||
gcloud run jobs update lyrics-transcription-job \
|
||||
--image "$IMAGE" \
|
||||
--region us-central1 \
|
||||
--project ${{ env.PROJECT_ID }} \
|
||||
--update-env-vars "AUTO_CORRECT_MODEL=claude-fable-5,AUTO_CORRECT_COMPARE_MODELS=claude-fable-5;gemini-3.1-pro-preview,AUTO_CORRECT_PROACTIVE_ENABLED=true" \
|
||||
--update-secrets "ANTHROPIC_API_KEY=anthropic-api-key:latest" \
|
||||
--quiet
|
||||
|
||||
# Audio separation job (GPU, us-east4) — uses GPU image
|
||||
|
||||
@@ -95,6 +95,14 @@ class Settings(BaseSettings):
|
||||
# Models used by the multi-model "compare" mode (semicolon-separated —
|
||||
# Cloud Run --set-env-vars is comma-delimited). Empty = single-model only.
|
||||
auto_correct_compare_models: str = os.getenv("AUTO_CORRECT_COMPARE_MODELS", "")
|
||||
# Proactively generate (and cache) auto-correct suggestions in the lyrics
|
||||
# worker once transcription + references are ready, so the review UI has
|
||||
# them instantly. Best-effort: failures never block or fail the job. Only
|
||||
# meaningful where ANTHROPIC_API_KEY + AUTO_CORRECT_COMPARE_MODELS are set
|
||||
# (the lyrics-transcription-job, via ci.yml). Default off.
|
||||
auto_correct_proactive_enabled: bool = os.getenv(
|
||||
"AUTO_CORRECT_PROACTIVE_ENABLED", "false"
|
||||
).lower() in ("true", "1", "yes")
|
||||
|
||||
# Cloud Tasks (for scalable worker coordination)
|
||||
# When enabled, workers are triggered via Cloud Tasks for guaranteed delivery
|
||||
|
||||
@@ -0,0 +1,94 @@
|
||||
"""Token pricing for auto-correct model calls.
|
||||
|
||||
Used to turn the per-call token usage we capture from each provider into an
|
||||
estimated USD cost, for both the production usage logs and the offline
|
||||
measurement script (``scripts/measure_autocorrect_cost.py``).
|
||||
|
||||
Prices are USD per 1,000,000 tokens.
|
||||
|
||||
- **Anthropic** prices are authoritative (published Claude API pricing).
|
||||
``output`` is the billed output rate, which already includes adaptive
|
||||
*thinking* tokens — they are billed as output and are not separable from
|
||||
the final answer in the Anthropic usage object.
|
||||
- **Gemini** prices are a best-known ESTIMATE for the Vertex preview tier and
|
||||
are clearly marked below. Verify against current Vertex pricing before
|
||||
relying on the dollar figure. Gemini thinking tokens ARE separable
|
||||
(``thoughts_token_count``) and are billed at the output rate, so we fold
|
||||
them into the output total.
|
||||
|
||||
The raw token counts are always logged alongside the dollar figure, so the
|
||||
cost here is a derived convenience — if a rate is stale, recompute from the
|
||||
logged tokens rather than trusting the historical USD value.
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
from typing import Optional, TypedDict
|
||||
|
||||
|
||||
class TokenUsage(TypedDict):
|
||||
"""Normalized, provider-agnostic usage for one model call.
|
||||
|
||||
``output_tokens`` is the total billed output (final answer + thinking).
|
||||
``thinking_tokens`` is informational only (None when the provider does
|
||||
not report it separately) — it is already counted inside ``output_tokens``.
|
||||
"""
|
||||
|
||||
input_tokens: int
|
||||
output_tokens: int
|
||||
cache_read_tokens: int
|
||||
thinking_tokens: Optional[int]
|
||||
|
||||
|
||||
class _Rate(TypedDict):
|
||||
input: float
|
||||
output: float
|
||||
cache_read: float
|
||||
|
||||
|
||||
# USD per 1,000,000 tokens.
|
||||
MODEL_PRICING: dict[str, _Rate] = {
|
||||
# --- Anthropic (authoritative) ---
|
||||
"claude-fable-5": {"input": 10.0, "output": 50.0, "cache_read": 1.0},
|
||||
"claude-opus-4-8": {"input": 5.0, "output": 25.0, "cache_read": 0.5},
|
||||
"claude-sonnet-4-6": {"input": 3.0, "output": 15.0, "cache_read": 0.3},
|
||||
"claude-haiku-4-5": {"input": 1.0, "output": 5.0, "cache_read": 0.1},
|
||||
# --- Gemini (ESTIMATE — Vertex 3.x Pro preview, <=200K-token context
|
||||
# tier; confirm against current Google pricing before trusting USD) ---
|
||||
"gemini-3.1-pro-preview": {"input": 2.0, "output": 12.0, "cache_read": 0.2},
|
||||
}
|
||||
|
||||
|
||||
def get_rate(model: str) -> Optional[_Rate]:
|
||||
"""Return the rate card for ``model``, or None if we have no pricing.
|
||||
|
||||
Falls back to a prefix match on the model family (e.g. a dated
|
||||
``claude-fable-5-2026...`` snapshot still resolves to ``claude-fable-5``)
|
||||
so a minor model-id change doesn't silently drop cost tracking.
|
||||
"""
|
||||
if model in MODEL_PRICING:
|
||||
return MODEL_PRICING[model]
|
||||
for known, rate in MODEL_PRICING.items():
|
||||
if model.startswith(known):
|
||||
return rate
|
||||
return None
|
||||
|
||||
|
||||
def estimate_cost_usd(model: str, usage: Optional[TokenUsage]) -> Optional[float]:
|
||||
"""Estimate the USD cost of one model call from its token usage.
|
||||
|
||||
Returns None when usage is unavailable or the model has no known pricing,
|
||||
so callers can distinguish "free/zero" from "unknown".
|
||||
"""
|
||||
if not usage:
|
||||
return None
|
||||
rate = get_rate(model)
|
||||
if rate is None:
|
||||
return None
|
||||
input_tokens = usage.get("input_tokens") or 0
|
||||
output_tokens = usage.get("output_tokens") or 0
|
||||
cache_read = usage.get("cache_read_tokens") or 0
|
||||
return (
|
||||
input_tokens * rate["input"]
|
||||
+ output_tokens * rate["output"]
|
||||
+ cache_read * rate["cache_read"]
|
||||
) / 1_000_000.0
|
||||
@@ -13,6 +13,7 @@ from google import genai
|
||||
from google.genai import types
|
||||
|
||||
from backend.config import get_settings
|
||||
from backend.services.auto_correct.pricing import TokenUsage, estimate_cost_usd
|
||||
from backend.services.auto_correct.prompts import (
|
||||
RESPONSE_SCHEMA,
|
||||
build_system_prompt,
|
||||
@@ -131,11 +132,17 @@ class AutoCorrectService:
|
||||
|
||||
t0 = time.time()
|
||||
per_model: dict[str, list[Suggestion]] = {}
|
||||
# Token usage per successful model call (None when the provider's
|
||||
# usage object couldn't be read). Drives the cost log below.
|
||||
per_model_usage: dict[str, Optional[TokenUsage]] = {}
|
||||
warnings: list[str] = []
|
||||
if len(models) == 1:
|
||||
raw = self._call_model(models[0], system_prompt, user_prompt, job_id=job_id)
|
||||
raw, usage = self._call_model(
|
||||
models[0], system_prompt, user_prompt, job_id=job_id
|
||||
)
|
||||
suggestions, vw = self._validate(raw, flat, settings)
|
||||
per_model[models[0]] = suggestions
|
||||
per_model_usage[models[0]] = usage
|
||||
warnings.extend(vw)
|
||||
else:
|
||||
import concurrent.futures as cf
|
||||
@@ -149,7 +156,7 @@ class AutoCorrectService:
|
||||
}
|
||||
for fut, m in futures.items():
|
||||
try:
|
||||
raw = fut.result()
|
||||
raw, usage = fut.result()
|
||||
suggestions, vw = self._validate(raw, flat, settings)
|
||||
except AutoCorrectServiceError as exc:
|
||||
# One model failing (call OR malformed output) must not
|
||||
@@ -157,6 +164,7 @@ class AutoCorrectService:
|
||||
warnings.append(f"model {m} failed: {exc}")
|
||||
continue
|
||||
per_model[m] = suggestions
|
||||
per_model_usage[m] = usage
|
||||
warnings.extend(f"[{m}] {w}" for w in vw)
|
||||
if not per_model:
|
||||
raise AutoCorrectServiceError(
|
||||
@@ -170,6 +178,7 @@ class AutoCorrectService:
|
||||
"auto-correct job=%s models=%s words=%d suggestions=%d dropped_warnings=%d elapsed=%.1fs",
|
||||
job_id, model_label, len(flat), len(merged), len(warnings), elapsed,
|
||||
)
|
||||
self._log_usage(job_id, per_model_usage)
|
||||
result = AutoCorrectResult(
|
||||
suggestions=merged,
|
||||
model=model_label,
|
||||
@@ -180,6 +189,50 @@ class AutoCorrectService:
|
||||
self._cache_put(cache_path, result)
|
||||
return result
|
||||
|
||||
@staticmethod
|
||||
def _log_usage(
|
||||
job_id: str, per_model_usage: dict[str, Optional[TokenUsage]]
|
||||
) -> None:
|
||||
"""Emit greppable per-model and total cost lines for one suggest() call.
|
||||
|
||||
Token counts are ground truth; the USD figures are derived from
|
||||
``pricing.MODEL_PRICING`` and use ESTIMATED Gemini rates (see that
|
||||
module). Query in Cloud Logging via ``auto-correct usage`` /
|
||||
``auto-correct cost``. Observability only — never raises.
|
||||
"""
|
||||
total_cost = 0.0
|
||||
any_cost = False
|
||||
for model, usage in per_model_usage.items():
|
||||
cost = estimate_cost_usd(model, usage)
|
||||
if cost is not None:
|
||||
total_cost += cost
|
||||
any_cost = True
|
||||
if usage is None:
|
||||
logger.info(
|
||||
"auto-correct usage job=%s model=%s tokens=unavailable",
|
||||
job_id, model,
|
||||
)
|
||||
continue
|
||||
thinking = usage.get("thinking_tokens")
|
||||
logger.info(
|
||||
"auto-correct usage job=%s model=%s input_tokens=%d "
|
||||
"output_tokens=%d thinking_tokens=%s cache_read_tokens=%d "
|
||||
"est_cost_usd=%s",
|
||||
job_id,
|
||||
model,
|
||||
usage.get("input_tokens", 0),
|
||||
usage.get("output_tokens", 0),
|
||||
"na" if thinking is None else thinking,
|
||||
usage.get("cache_read_tokens", 0),
|
||||
"unknown" if cost is None else f"{cost:.4f}",
|
||||
)
|
||||
logger.info(
|
||||
"auto-correct cost job=%s models=%s total_est_cost_usd=%s",
|
||||
job_id,
|
||||
", ".join(per_model_usage.keys()),
|
||||
f"{total_cost:.4f}" if any_cost else "unknown",
|
||||
)
|
||||
|
||||
# ---- suggestion cache (GCS, best-effort) ----
|
||||
|
||||
def _get_storage(self):
|
||||
@@ -334,14 +387,59 @@ class AutoCorrectService:
|
||||
|
||||
def _call_model(
|
||||
self, model: str, system_prompt: str, user_prompt: str, *, job_id: str
|
||||
) -> Any:
|
||||
) -> tuple[Any, Optional[TokenUsage]]:
|
||||
"""Return (parsed_json, token_usage). Usage is None if unreadable."""
|
||||
if model.startswith("claude"):
|
||||
return self._call_anthropic(model, system_prompt, user_prompt, job_id=job_id)
|
||||
return self._call_gemini(model, system_prompt, user_prompt, job_id=job_id)
|
||||
|
||||
@staticmethod
|
||||
def _anthropic_usage(response: Any) -> Optional[TokenUsage]:
|
||||
"""Best-effort usage from an Anthropic response.
|
||||
|
||||
``output_tokens`` already includes adaptive thinking tokens (billed as
|
||||
output); they are not separable here, so thinking_tokens is None.
|
||||
Returns None on any shape problem — usage must never break the call.
|
||||
"""
|
||||
try:
|
||||
u = response.usage
|
||||
return TokenUsage(
|
||||
input_tokens=int(getattr(u, "input_tokens", 0) or 0),
|
||||
output_tokens=int(getattr(u, "output_tokens", 0) or 0),
|
||||
cache_read_tokens=int(getattr(u, "cache_read_input_tokens", 0) or 0),
|
||||
thinking_tokens=None,
|
||||
)
|
||||
except Exception:
|
||||
logger.debug("auto-correct: could not read Anthropic usage", exc_info=True)
|
||||
return None
|
||||
|
||||
@staticmethod
|
||||
def _gemini_usage(response: Any) -> Optional[TokenUsage]:
|
||||
"""Best-effort usage from a Gemini (Vertex) response.
|
||||
|
||||
Gemini reports thinking separately (``thoughts_token_count``); it is
|
||||
billed at the output rate, so it's folded into output_tokens and also
|
||||
reported separately for visibility. Returns None on any shape problem.
|
||||
"""
|
||||
try:
|
||||
u = response.usage_metadata
|
||||
prompt = int(getattr(u, "prompt_token_count", 0) or 0)
|
||||
candidates = int(getattr(u, "candidates_token_count", 0) or 0)
|
||||
thoughts = int(getattr(u, "thoughts_token_count", 0) or 0)
|
||||
cached = int(getattr(u, "cached_content_token_count", 0) or 0)
|
||||
return TokenUsage(
|
||||
input_tokens=prompt,
|
||||
output_tokens=candidates + thoughts,
|
||||
cache_read_tokens=cached,
|
||||
thinking_tokens=thoughts,
|
||||
)
|
||||
except Exception:
|
||||
logger.debug("auto-correct: could not read Gemini usage", exc_info=True)
|
||||
return None
|
||||
|
||||
def _call_anthropic(
|
||||
self, model: str, system_prompt: str, user_prompt: str, *, job_id: str
|
||||
) -> Any:
|
||||
) -> tuple[Any, Optional[TokenUsage]]:
|
||||
import os
|
||||
|
||||
import anthropic
|
||||
@@ -384,9 +482,10 @@ class AutoCorrectService:
|
||||
raise AutoCorrectServiceError(
|
||||
"AI model call failed", status_code=502
|
||||
) from exc
|
||||
usage = self._anthropic_usage(response)
|
||||
try:
|
||||
text = next(b.text for b in response.content if b.type == "text")
|
||||
return json.loads(text)
|
||||
return json.loads(text), usage
|
||||
except (StopIteration, json.JSONDecodeError, TypeError) as exc:
|
||||
logger.exception(
|
||||
"auto-correct model returned non-JSON job=%s model=%s", job_id, model
|
||||
@@ -397,7 +496,7 @@ class AutoCorrectService:
|
||||
|
||||
def _call_gemini(
|
||||
self, model: str, system_prompt: str, user_prompt: str, *, job_id: str
|
||||
) -> Any:
|
||||
) -> tuple[Any, Optional[TokenUsage]]:
|
||||
client = genai.Client(
|
||||
vertexai=True,
|
||||
project=self.settings.google_cloud_project,
|
||||
@@ -425,8 +524,9 @@ class AutoCorrectService:
|
||||
raise AutoCorrectServiceError(
|
||||
"AI model call failed", status_code=502
|
||||
) from exc
|
||||
usage = self._gemini_usage(response)
|
||||
try:
|
||||
return json.loads(response.text)
|
||||
return json.loads(response.text), usage
|
||||
except (json.JSONDecodeError, TypeError) as exc:
|
||||
logger.exception(
|
||||
"auto-correct model returned non-JSON job=%s model=%s", job_id, model
|
||||
|
||||
@@ -85,7 +85,7 @@ def test_cache_hit_skips_model_call() -> None:
|
||||
def test_cache_miss_calls_model_and_writes_cache() -> None:
|
||||
service = AutoCorrectService()
|
||||
with patch.object(service, "_cache_get", return_value=None), \
|
||||
patch.object(service, "_call_model", return_value=RAW) as call, \
|
||||
patch.object(service, "_call_model", return_value=(RAW, None)) as call, \
|
||||
patch.object(service, "_cache_put") as put:
|
||||
result = _suggest(service)
|
||||
call.assert_called_once()
|
||||
@@ -197,7 +197,7 @@ def test_corrupt_cache_entry_is_a_miss() -> None:
|
||||
def test_storage_failure_never_breaks_request() -> None:
|
||||
service = AutoCorrectService()
|
||||
with patch.object(service, "_get_storage", side_effect=RuntimeError("no gcs")), \
|
||||
patch.object(service, "_call_model", return_value=RAW):
|
||||
patch.object(service, "_call_model", return_value=(RAW, None)):
|
||||
result = _suggest(service)
|
||||
assert result.cached is False
|
||||
assert len(result.suggestions) == 1
|
||||
@@ -49,7 +49,8 @@ def _run_compare(responses_by_model, compare_models_env="model-a;model-b",
|
||||
resp = responses_by_model[model]
|
||||
if isinstance(resp, Exception):
|
||||
raise resp
|
||||
return resp
|
||||
# _call_model now returns (parsed_json, token_usage).
|
||||
return resp, None
|
||||
|
||||
with patch.object(
|
||||
service.settings, "auto_correct_compare_models", compare_models_env
|
||||
|
||||
@@ -0,0 +1,173 @@
|
||||
"""Tests for auto-correct token pricing and usage extraction/logging."""
|
||||
from __future__ import annotations
|
||||
|
||||
from types import SimpleNamespace
|
||||
from unittest.mock import patch
|
||||
|
||||
from backend.services.auto_correct.pricing import (
|
||||
MODEL_PRICING,
|
||||
estimate_cost_usd,
|
||||
get_rate,
|
||||
)
|
||||
from backend.services.auto_correct.service import AutoCorrectService
|
||||
from backend.services.auto_correct.settings import AutoCorrectSettings
|
||||
|
||||
|
||||
# ---- pricing ----
|
||||
|
||||
def test_estimate_cost_fable_known_rates() -> None:
|
||||
# 5000 in @ $10/M + 4000 out @ $50/M = 0.05 + 0.20 = 0.25
|
||||
usage = {
|
||||
"input_tokens": 5000,
|
||||
"output_tokens": 4000,
|
||||
"cache_read_tokens": 0,
|
||||
"thinking_tokens": None,
|
||||
}
|
||||
cost = estimate_cost_usd("claude-fable-5", usage)
|
||||
assert cost == round(0.05 + 0.20, 10)
|
||||
|
||||
|
||||
def test_estimate_cost_includes_cache_read() -> None:
|
||||
usage = {
|
||||
"input_tokens": 1_000_000,
|
||||
"output_tokens": 0,
|
||||
"cache_read_tokens": 1_000_000,
|
||||
"thinking_tokens": None,
|
||||
}
|
||||
# 1M input @ $10 + 1M cache_read @ $1 = 11.0
|
||||
assert estimate_cost_usd("claude-fable-5", usage) == 11.0
|
||||
|
||||
|
||||
def test_estimate_cost_none_usage_returns_none() -> None:
|
||||
assert estimate_cost_usd("claude-fable-5", None) is None
|
||||
|
||||
|
||||
def test_estimate_cost_unknown_model_returns_none() -> None:
|
||||
usage = {
|
||||
"input_tokens": 1000,
|
||||
"output_tokens": 1000,
|
||||
"cache_read_tokens": 0,
|
||||
"thinking_tokens": None,
|
||||
}
|
||||
assert estimate_cost_usd("some-unknown-model-9", usage) is None
|
||||
|
||||
|
||||
def test_get_rate_prefix_match_for_dated_snapshot() -> None:
|
||||
# A dated snapshot id still resolves to its family rate card.
|
||||
assert get_rate("claude-fable-5-20260601") == MODEL_PRICING["claude-fable-5"]
|
||||
|
||||
|
||||
# ---- usage extraction ----
|
||||
|
||||
def test_anthropic_usage_extraction() -> None:
|
||||
response = SimpleNamespace(
|
||||
usage=SimpleNamespace(
|
||||
input_tokens=4321,
|
||||
output_tokens=3210,
|
||||
cache_read_input_tokens=100,
|
||||
)
|
||||
)
|
||||
usage = AutoCorrectService._anthropic_usage(response)
|
||||
assert usage == {
|
||||
"input_tokens": 4321,
|
||||
"output_tokens": 3210, # already includes thinking
|
||||
"cache_read_tokens": 100,
|
||||
"thinking_tokens": None,
|
||||
}
|
||||
|
||||
|
||||
def test_gemini_usage_folds_thoughts_into_output() -> None:
|
||||
response = SimpleNamespace(
|
||||
usage_metadata=SimpleNamespace(
|
||||
prompt_token_count=4000,
|
||||
candidates_token_count=500,
|
||||
thoughts_token_count=2500,
|
||||
cached_content_token_count=0,
|
||||
)
|
||||
)
|
||||
usage = AutoCorrectService._gemini_usage(response)
|
||||
assert usage == {
|
||||
"input_tokens": 4000,
|
||||
"output_tokens": 3000, # 500 visible + 2500 thinking
|
||||
"cache_read_tokens": 0,
|
||||
"thinking_tokens": 2500,
|
||||
}
|
||||
|
||||
|
||||
def test_usage_extraction_never_raises_on_bad_shape() -> None:
|
||||
# Missing usage attribute entirely -> None, not an exception.
|
||||
assert AutoCorrectService._anthropic_usage(SimpleNamespace()) is None
|
||||
assert AutoCorrectService._gemini_usage(SimpleNamespace()) is None
|
||||
|
||||
|
||||
# ---- logging ----
|
||||
|
||||
SEGMENTS = [
|
||||
{"id": "seg-1", "words": [{"id": "w0", "text": "glory"}]},
|
||||
]
|
||||
REFS = {"genius": {"segments": [{"text": "chlorine"}]}}
|
||||
|
||||
|
||||
def _raw():
|
||||
return {
|
||||
"suggestions": [
|
||||
{
|
||||
"op": "replace",
|
||||
"start_idx": 0,
|
||||
"end_idx": 0,
|
||||
"new_text": "chlorine",
|
||||
"reason": "r",
|
||||
"category": "mishearing",
|
||||
"confidence": 0.9,
|
||||
}
|
||||
]
|
||||
}
|
||||
|
||||
|
||||
def test_suggest_logs_per_model_and_total_cost(caplog) -> None:
|
||||
service = AutoCorrectService()
|
||||
usage = {
|
||||
"input_tokens": 5000,
|
||||
"output_tokens": 4000,
|
||||
"cache_read_tokens": 0,
|
||||
"thinking_tokens": None,
|
||||
}
|
||||
with patch.object(service, "_cache_get", return_value=None), \
|
||||
patch.object(service, "_cache_put"), \
|
||||
patch.object(service, "_call_model", return_value=(_raw(), usage)), \
|
||||
patch.object(service.settings, "auto_correct_model", "claude-fable-5"):
|
||||
with caplog.at_level("INFO"):
|
||||
service.suggest(
|
||||
job_id="job-1",
|
||||
segments=SEGMENTS,
|
||||
reference_lyrics=REFS,
|
||||
artist="A",
|
||||
title="T",
|
||||
settings=AutoCorrectSettings(),
|
||||
)
|
||||
text = caplog.text
|
||||
assert "auto-correct usage job=job-1 model=claude-fable-5" in text
|
||||
assert "input_tokens=5000" in text
|
||||
assert "output_tokens=4000" in text
|
||||
assert "est_cost_usd=0.2500" in text
|
||||
assert "auto-correct cost job=job-1" in text
|
||||
assert "total_est_cost_usd=0.2500" in text
|
||||
|
||||
|
||||
def test_suggest_logs_unavailable_usage_without_crashing(caplog) -> None:
|
||||
service = AutoCorrectService()
|
||||
with patch.object(service, "_cache_get", return_value=None), \
|
||||
patch.object(service, "_cache_put"), \
|
||||
patch.object(service, "_call_model", return_value=(_raw(), None)), \
|
||||
patch.object(service.settings, "auto_correct_model", "claude-fable-5"):
|
||||
with caplog.at_level("INFO"):
|
||||
service.suggest(
|
||||
job_id="job-2",
|
||||
segments=SEGMENTS,
|
||||
reference_lyrics=REFS,
|
||||
artist="A",
|
||||
title="T",
|
||||
settings=AutoCorrectSettings(),
|
||||
)
|
||||
assert "auto-correct usage job=job-2 model=claude-fable-5 tokens=unavailable" in caplog.text
|
||||
assert "total_est_cost_usd=unknown" in caplog.text
|
||||
@@ -35,7 +35,7 @@ REFS = {"genius": {"segments": [{"text": "Sippin' on straight chlorine"}]}}
|
||||
|
||||
def _suggest(raw_response, settings=None):
|
||||
service = AutoCorrectService()
|
||||
with patch.object(service, "_call_model", return_value=raw_response), \
|
||||
with patch.object(service, "_call_model", return_value=(raw_response, None)), \
|
||||
patch.object(service, "_cache_get", return_value=None), \
|
||||
patch.object(service, "_cache_put"):
|
||||
return service.suggest(
|
||||
@@ -217,8 +217,8 @@ def test_non_dict_suggestion_entries_dropped() -> None:
|
||||
|
||||
def test_call_model_dispatches_claude_to_anthropic() -> None:
|
||||
service = AutoCorrectService()
|
||||
with patch.object(service, "_call_anthropic", return_value={"suggestions": []}) as anth, \
|
||||
patch.object(service, "_call_gemini", return_value={"suggestions": []}) as gem:
|
||||
with patch.object(service, "_call_anthropic", return_value=({"suggestions": []}, None)) as anth, \
|
||||
patch.object(service, "_call_gemini", return_value=({"suggestions": []}, None)) as gem:
|
||||
service._call_model("claude-fable-5", "s", "u", job_id="j")
|
||||
anth.assert_called_once()
|
||||
gem.assert_not_called()
|
||||
@@ -262,8 +262,8 @@ def test_anthropic_schema_strips_genai_pollution(monkeypatch) -> None:
|
||||
|
||||
monkeypatch.setattr(anthropic_sdk, "Anthropic", FakeClient)
|
||||
service = AutoCorrectService()
|
||||
result = service._call_anthropic("claude-fable-5", "s", "u", job_id="j")
|
||||
assert result == {"suggestions": []}
|
||||
raw, _usage = service._call_anthropic("claude-fable-5", "s", "u", job_id="j")
|
||||
assert raw == {"suggestions": []}
|
||||
schema = captured["output_config"]["format"]["schema"]
|
||||
assert "property_ordering" not in schema
|
||||
assert "property_ordering" not in schema["properties"]["suggestions"]["items"]
|
||||
|
||||
@@ -0,0 +1,97 @@
|
||||
"""Tests for proactive auto-correct generation in the lyrics worker.
|
||||
|
||||
The helper must be best-effort: gated by a flag, never raise, never block, and
|
||||
use the multi-model default settings that line up with the review UI's request.
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
from types import SimpleNamespace
|
||||
from unittest.mock import AsyncMock, MagicMock, patch
|
||||
|
||||
import pytest
|
||||
|
||||
from backend.workers.lyrics_worker import _run_proactive_auto_correct
|
||||
|
||||
|
||||
CORRECTIONS = {
|
||||
"corrected_segments": [
|
||||
{"id": "seg-1", "words": [{"id": "w0", "text": "glory"}]},
|
||||
],
|
||||
"reference_lyrics": {"genius": {"segments": [{"text": "chlorine"}]}},
|
||||
}
|
||||
|
||||
|
||||
def _storage(corrections=CORRECTIONS, exists=True):
|
||||
storage = MagicMock()
|
||||
storage.file_exists.return_value = exists
|
||||
storage.download_json.return_value = corrections
|
||||
return storage
|
||||
|
||||
|
||||
def _settings(enabled: bool):
|
||||
return SimpleNamespace(auto_correct_proactive_enabled=enabled)
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skips_when_flag_disabled() -> None:
|
||||
storage = _storage()
|
||||
job = SimpleNamespace(artist="A", title="T")
|
||||
with patch("backend.config.get_settings", return_value=_settings(False)), \
|
||||
patch("backend.services.auto_correct.get_auto_correct_service") as svc:
|
||||
await _run_proactive_auto_correct("job-1", job, storage, MagicMock())
|
||||
svc.assert_not_called()
|
||||
storage.file_exists.assert_not_called()
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_skips_when_no_references() -> None:
|
||||
storage = _storage(corrections={"corrected_segments": [{"id": "s", "words": []}],
|
||||
"reference_lyrics": {}})
|
||||
job = SimpleNamespace(artist="A", title="T")
|
||||
with patch("backend.config.get_settings", return_value=_settings(True)), \
|
||||
patch("backend.services.auto_correct.get_auto_correct_service") as svc:
|
||||
await _run_proactive_auto_correct("job-1", job, storage, MagicMock())
|
||||
svc.assert_not_called()
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_happy_path_calls_suggest_multi_model() -> None:
|
||||
storage = _storage()
|
||||
job = SimpleNamespace(artist="A", title="T")
|
||||
service = MagicMock()
|
||||
service.suggest.return_value = SimpleNamespace(
|
||||
suggestions=[1, 2], model="claude-fable-5, gemini-3.1-pro-preview",
|
||||
elapsed_seconds=12.3,
|
||||
)
|
||||
with patch("backend.config.get_settings", return_value=_settings(True)), \
|
||||
patch("backend.services.auto_correct.get_auto_correct_service", return_value=service):
|
||||
await _run_proactive_auto_correct("job-1", job, storage, MagicMock())
|
||||
service.suggest.assert_called_once()
|
||||
kwargs = service.suggest.call_args.kwargs
|
||||
assert kwargs["job_id"] == "job-1"
|
||||
assert kwargs["settings"].compare_models is True
|
||||
assert kwargs["artist"] == "A"
|
||||
assert kwargs["segments"] == CORRECTIONS["corrected_segments"]
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_swallows_service_errors() -> None:
|
||||
storage = _storage()
|
||||
job = SimpleNamespace(artist="A", title="T")
|
||||
service = MagicMock()
|
||||
service.suggest.side_effect = RuntimeError("anthropic down / out of credits")
|
||||
job_log = MagicMock()
|
||||
with patch("backend.config.get_settings", return_value=_settings(True)), \
|
||||
patch("backend.services.auto_correct.get_auto_correct_service", return_value=service):
|
||||
# Must NOT raise — the karaoke job proceeds regardless.
|
||||
await _run_proactive_auto_correct("job-1", job, storage, job_log)
|
||||
job_log.warning.assert_called() # logged, not raised
|
||||
|
||||
|
||||
@pytest.mark.asyncio
|
||||
async def test_swallows_storage_errors() -> None:
|
||||
storage = MagicMock()
|
||||
storage.file_exists.side_effect = RuntimeError("gcs blip")
|
||||
job = SimpleNamespace(artist="A", title="T")
|
||||
with patch("backend.config.get_settings", return_value=_settings(True)):
|
||||
await _run_proactive_auto_correct("job-1", job, storage, MagicMock()) # no raise
|
||||
@@ -170,6 +170,68 @@ def create_lyrics_processor(
|
||||
)
|
||||
|
||||
|
||||
async def _run_proactive_auto_correct(job_id, job, storage, job_log) -> None:
|
||||
"""Pre-generate + cache AI auto-correct suggestions for the review UI.
|
||||
|
||||
Best-effort and fully isolated: gated behind a flag, bounded by a timeout,
|
||||
and swallows every error. It must NEVER raise, fail the job, or block the
|
||||
pipeline — downstream stages have already been triggered before this runs.
|
||||
|
||||
Reads the corrections.json the worker just wrote (the same data the review
|
||||
UI loads on first open) and runs the multi-model suggest() path with default
|
||||
settings, so the UI's on-load call is a cache hit instead of a fresh spend.
|
||||
"""
|
||||
from backend.config import get_settings
|
||||
|
||||
settings = get_settings()
|
||||
if not settings.auto_correct_proactive_enabled:
|
||||
return
|
||||
|
||||
try:
|
||||
from backend.services.auto_correct import get_auto_correct_service
|
||||
from backend.services.auto_correct.settings import AutoCorrectSettings
|
||||
|
||||
corrections_path = f"jobs/{job_id}/lyrics/corrections.json"
|
||||
if not storage.file_exists(corrections_path):
|
||||
job_log.info("Proactive auto-correct skipped: no corrections.json yet")
|
||||
return
|
||||
data = storage.download_json(corrections_path)
|
||||
segments = data.get("corrected_segments") or data.get("segments") or []
|
||||
reference_lyrics = data.get("reference_lyrics") or {}
|
||||
if not segments or not reference_lyrics:
|
||||
# No references → nothing to compare against. The reviewer can paste
|
||||
# references in the UI and trigger auto-correct manually later.
|
||||
job_log.info(
|
||||
"Proactive auto-correct skipped: %d segments, %d reference sources",
|
||||
len(segments), len(reference_lyrics),
|
||||
)
|
||||
return
|
||||
|
||||
service = get_auto_correct_service()
|
||||
job_log.info("Proactive auto-correct: generating suggestions (multi-model)")
|
||||
# Default settings + compare_models=True must match what the review UI
|
||||
# sends, or the cache key won't line up and the UI pays for a fresh call.
|
||||
result = await asyncio.wait_for(
|
||||
asyncio.to_thread(
|
||||
service.suggest,
|
||||
job_id=job_id,
|
||||
segments=segments,
|
||||
reference_lyrics=reference_lyrics,
|
||||
artist=getattr(job, "artist", None),
|
||||
title=getattr(job, "title", None),
|
||||
settings=AutoCorrectSettings(compare_models=True),
|
||||
),
|
||||
timeout=180,
|
||||
)
|
||||
job_log.info(
|
||||
"Proactive auto-correct done: %d suggestions cached (models=%s, %.1fs)",
|
||||
len(result.suggestions), result.model, result.elapsed_seconds,
|
||||
)
|
||||
except Exception as e: # noqa: BLE001 — never let auto-correct affect the job
|
||||
job_log.warning(f"Proactive auto-correct failed (non-fatal): {e}", exc_info=True)
|
||||
logger.warning(f"[job:{job_id}] Proactive auto-correct failed (non-fatal): {e}")
|
||||
|
||||
|
||||
async def process_lyrics_transcription(job_id: str) -> bool:
|
||||
"""
|
||||
Process lyrics transcription and correction for a job using karaoke_gen.LyricsProcessor.
|
||||
@@ -516,7 +578,14 @@ async def process_lyrics_transcription(job_id: str) -> bool:
|
||||
# This will check if audio is also complete and transition to next stage if so
|
||||
job_log.info("Lyrics worker complete, checking if audio is also done...")
|
||||
job_manager.mark_lyrics_complete(job_id)
|
||||
|
||||
|
||||
# Proactively pre-generate + cache AI auto-correct suggestions so
|
||||
# they're ready instantly when the reviewer opens the lyrics UI.
|
||||
# Done AFTER mark_lyrics_complete (downstream is already triggered,
|
||||
# so this never blocks it) and fully isolated — it can never fail
|
||||
# or delay the job beyond its own timeout.
|
||||
await _run_proactive_auto_correct(job_id, job, storage, job_log)
|
||||
|
||||
duration = time.time() - start_time
|
||||
# Store worker-level timing
|
||||
job_manager.update_processing_metadata(job_id, "timing.lyrics_worker_seconds", round(duration, 1))
|
||||
|
||||
+11
-4
@@ -560,20 +560,27 @@ Content-Type: application/json
|
||||
"suggest_adlib_removal": true,
|
||||
"allow_insertions": true,
|
||||
"min_confidence": 0.0,
|
||||
"compare_models": false
|
||||
"compare_models": true
|
||||
}
|
||||
}
|
||||
```
|
||||
|
||||
Opt-in, user-triggered AI correction suggestions for the lyrics review UI.
|
||||
Stateless: one whole-song LLM call per model (default model from
|
||||
`AUTO_CORRECT_MODEL`; Claude models use the Anthropic API via the
|
||||
AI correction suggestions for the lyrics review UI. Stateless: one whole-song
|
||||
LLM call per model (Claude models use the Anthropic API via the
|
||||
`anthropic-api-key` secret, Gemini models use Vertex AI) compares the
|
||||
client's current transcription against the reference sources and returns
|
||||
word-id-keyed suggestions. Nothing is applied server-side — the reviewer
|
||||
accepts/rejects each suggestion in the UI and persistence flows through the
|
||||
existing corrections/complete paths.
|
||||
|
||||
The review UI always requests the multi-model path (`compare_models: true`)
|
||||
and auto-runs once on load, so suggestions appear without a click. To make
|
||||
that load instant, the lyrics worker **proactively pre-generates and caches**
|
||||
the same multi-model run once transcription + references are ready (gated by
|
||||
`AUTO_CORRECT_PROACTIVE_ENABLED`; best-effort — it never blocks or fails the
|
||||
karaoke job). Because the proactive run uses the same default settings the UI
|
||||
sends, the on-load request is normally a cache hit with no extra cost.
|
||||
|
||||
With `"compare_models": true`, every model in `AUTO_CORRECT_COMPARE_MODELS`
|
||||
(semicolon-separated) is queried in parallel; identical suggestions are
|
||||
deduplicated with a per-suggestion `consensus` count, and suggestions that
|
||||
|
||||
@@ -0,0 +1,82 @@
|
||||
# Proactive AI Auto-Correction + UI De-AI-ification
|
||||
|
||||
**Date:** 2026-06-11
|
||||
**Branch:** `feat/sess-20260611-1514-autocorrect-cost-logging`
|
||||
**Context:** Measured cost ~$0.13/track (multi-model). Cheap enough to run on every job.
|
||||
|
||||
## Goal (user's words)
|
||||
|
||||
Run multi-model auto-correction automatically on every karaoke job so suggestions
|
||||
are ready to review/approve/reject by the time anyone reaches the lyrics review UI.
|
||||
Must fail gracefully — never block the karaoke job if AI APIs are flaky / out of
|
||||
credits. Remove all "AI"/"model" mentions from the review UI; button becomes just
|
||||
"Auto-correct"; drop the user-configurable model selection and always use multi-model.
|
||||
Keep the button for the case where references arrive later (user pastes reference
|
||||
lyrics during review, then clicks it).
|
||||
|
||||
## Architecture decisions (validated against the codebase)
|
||||
|
||||
1. **Where proactive runs: the `lyrics-transcription-job` Cloud Run Job**, after
|
||||
`mark_lyrics_complete()` (so downstream screens/render are already triggered and
|
||||
are NOT blocked), awaited (a fire-and-forget task would die when the job
|
||||
entrypoint returns). The job is long-lived, so an extra ~15-30s is fine and the
|
||||
process stays alive to finish the cache write.
|
||||
2. **Cache alignment = the whole point.** `AutoCorrectService` already writes a
|
||||
per-job GCS cache keyed on (models, settings, word-id+text list, ref texts,
|
||||
artist, title). If the proactive run uses the SAME inputs + settings + models the
|
||||
UI will send, the UI's on-load call is a **cache hit** (instant, no second spend).
|
||||
- Proactive settings = `AutoCorrectSettings(compare_models=True)` (all other
|
||||
fields default). Backend defaults == frontend `DEFAULT_AUTO_CORRECT_SETTINGS`
|
||||
once we flip `compare_models` to `true`. Both paths build settings server-side
|
||||
and hash the same dict.
|
||||
- `models` is derived server-side identically in both paths via `_models_for()`.
|
||||
- Proactive reads `corrections.json` (corrected_segments + reference_lyrics) — the
|
||||
exact data the review-data endpoint serves on first load (before any edits).
|
||||
3. **Env/secret wiring (REQUIRED for it to actually work).** The lyrics job currently
|
||||
has neither `ANTHROPIC_API_KEY` nor `AUTO_CORRECT_*`. It uses the SAME SA as the
|
||||
API service (`karaoke-backend@…`), which already reads `anthropic-api-key`
|
||||
(project-level secret access) — so adding the secret to the job is low-risk.
|
||||
Add via `--update-env-vars`/`--update-secrets` in the ci.yml job-update step
|
||||
(merges, doesn't wipe the existing 12 vars).
|
||||
4. **Feature flag:** `auto_correct_proactive_enabled` (config), default **false** in
|
||||
code, set `AUTO_CORRECT_PROACTIVE_ENABLED=true` on the lyrics job via ci.yml.
|
||||
5. **UI auto-run on load:** on review load, if references exist, auto-invoke
|
||||
auto-correct (cache hit → instant). Keep the button for manual re-run after the
|
||||
user pastes references. Always send `compare_models: true`.
|
||||
|
||||
## Changes
|
||||
|
||||
### Backend
|
||||
- `backend/config.py`: add `auto_correct_proactive_enabled`.
|
||||
- `backend/workers/lyrics_worker.py`: after `mark_lyrics_complete`, await a guarded
|
||||
helper `_run_proactive_auto_correct(job_id, ...)` — loads corrections.json, calls
|
||||
`suggest(..., settings=AutoCorrectSettings(compare_models=True))`, `asyncio.wait_for`
|
||||
timeout, swallow ALL exceptions (log only). Never affects job status.
|
||||
- Tests: worker test that it's gated by the flag, swallows errors, and skips when no
|
||||
references.
|
||||
|
||||
### Infra
|
||||
- `.github/workflows/ci.yml`: lyrics-transcription-job update step gains
|
||||
`--update-env-vars "AUTO_CORRECT_MODEL=claude-fable-5,AUTO_CORRECT_COMPARE_MODELS=claude-fable-5;gemini-3.1-pro-preview,AUTO_CORRECT_PROACTIVE_ENABLED=true"`
|
||||
and `--update-secrets "ANTHROPIC_API_KEY=anthropic-api-key:latest"`.
|
||||
|
||||
### Frontend (`frontend/`, next-intl, 33 locales)
|
||||
- Always multi-model: `DEFAULT_AUTO_CORRECT_SETTINGS.compare_models = true`; remove
|
||||
the compare/model toggle from `AutoCorrectModal.tsx`.
|
||||
- Strip model attribution: remove model-name badge + consensus "X/Y models" badge in
|
||||
`AutoCorrectPanel.tsx`; reword the conflict badge to not mention models (keep the
|
||||
conflict-resolution behavior).
|
||||
- Rename button to "Auto-correct"; remove "AI"/"model" from all related strings.
|
||||
- Auto-run on review load when references exist (cache hit → instant); keep button.
|
||||
- i18n: edit `messages/en.json`, then `python frontend/scripts/translate.py
|
||||
--messages-dir frontend/messages --target all` for all 33 locales.
|
||||
- Update `e2e/production/auto-correct-suggestions.spec.ts` button selector.
|
||||
|
||||
## Failure modes (all non-blocking by design)
|
||||
- Anthropic down / no credits → claude call raises, caught; Gemini result still cached
|
||||
(multi-model degrades to whatever answered). UI on-load call same behaviour.
|
||||
- Both providers fail → proactive logs warning, job proceeds normally; UI on-load call
|
||||
returns 502 and the panel shows nothing (or the manual button can retry later).
|
||||
- Proactive task times out → `wait_for` cancels it, job proceeds.
|
||||
- No references at lyrics-complete time → proactive skipped; user pastes refs in UI →
|
||||
clicks button → on-demand call.
|
||||
@@ -62,9 +62,6 @@ export default function AutoCorrectModal({
|
||||
DEFAULT_AUTO_CORRECT_SETTINGS.allow_insertions,
|
||||
)
|
||||
const [confidencePreset, setConfidencePreset] = useState<ConfidencePreset>('all')
|
||||
const [compareModels, setCompareModels] = useState(
|
||||
DEFAULT_AUTO_CORRECT_SETTINGS.compare_models,
|
||||
)
|
||||
|
||||
const isRunning = status === 'running'
|
||||
|
||||
@@ -73,7 +70,9 @@ export default function AutoCorrectModal({
|
||||
suggest_adlib_removal: adlibRemoval,
|
||||
allow_insertions: allowInsertions,
|
||||
min_confidence: PRESET_THRESHOLDS[confidencePreset],
|
||||
compare_models: compareModels,
|
||||
// Always multi-model — kept aligned with the backend proactive run so
|
||||
// the cached result is reused.
|
||||
compare_models: DEFAULT_AUTO_CORRECT_SETTINGS.compare_models,
|
||||
})
|
||||
}
|
||||
|
||||
@@ -133,20 +132,6 @@ export default function AutoCorrectModal({
|
||||
/>
|
||||
</div>
|
||||
|
||||
<div className="flex items-center justify-between gap-4">
|
||||
<div>
|
||||
<Label htmlFor="ac-compare">{t('compareLabel')}</Label>
|
||||
<p className="text-xs text-muted-foreground">
|
||||
{t('compareHint')}
|
||||
</p>
|
||||
</div>
|
||||
<Switch
|
||||
id="ac-compare"
|
||||
checked={compareModels}
|
||||
onCheckedChange={setCompareModels}
|
||||
/>
|
||||
</div>
|
||||
|
||||
<div className="flex items-center justify-between gap-4">
|
||||
<div>
|
||||
<Label>{t('confidenceLabel')}</Label>
|
||||
|
||||
@@ -57,8 +57,6 @@ export default function AutoCorrectPanel({
|
||||
const {
|
||||
suggestions,
|
||||
decisions,
|
||||
model,
|
||||
cached,
|
||||
pendingCount,
|
||||
acceptedCount,
|
||||
accept,
|
||||
@@ -86,26 +84,6 @@ export default function AutoCorrectPanel({
|
||||
accepted: acceptedCount,
|
||||
})}
|
||||
</span>
|
||||
<Badge variant="outline" className="hidden sm:inline-flex text-xs">
|
||||
{model}
|
||||
</Badge>
|
||||
{cached && (
|
||||
<TooltipProvider>
|
||||
<Tooltip>
|
||||
<TooltipTrigger asChild>
|
||||
<Badge
|
||||
variant="outline"
|
||||
className="shrink-0 text-xs border-teal-500/50 text-teal-600 dark:text-teal-400"
|
||||
>
|
||||
{t('cachedBadge')}
|
||||
</Badge>
|
||||
</TooltipTrigger>
|
||||
<TooltipContent className="max-w-xs">
|
||||
{t('cachedTooltip')}
|
||||
</TooltipContent>
|
||||
</Tooltip>
|
||||
</TooltipProvider>
|
||||
)}
|
||||
</div>
|
||||
<div className="flex items-center gap-1 shrink-0">
|
||||
{!isReadOnly && pendingCount > 0 && (
|
||||
@@ -169,31 +147,6 @@ export default function AutoCorrectPanel({
|
||||
>
|
||||
{t(`category_${s.category}`)}
|
||||
</Badge>
|
||||
{s.total_models > 1 && (
|
||||
<TooltipProvider>
|
||||
<Tooltip>
|
||||
<TooltipTrigger asChild>
|
||||
<Badge
|
||||
variant="outline"
|
||||
className={cn(
|
||||
'shrink-0 text-[10px] tabular-nums',
|
||||
s.consensus === s.total_models
|
||||
? 'border-green-500/50 text-green-600 dark:text-green-400'
|
||||
: 'border-amber-500/50 text-amber-600 dark:text-amber-400',
|
||||
)}
|
||||
>
|
||||
{t('consensusBadge', {
|
||||
consensus: s.consensus,
|
||||
total: s.total_models,
|
||||
})}
|
||||
</Badge>
|
||||
</TooltipTrigger>
|
||||
<TooltipContent className="max-w-xs">
|
||||
{t('consensusTooltip', { models: s.models.join(', ') })}
|
||||
</TooltipContent>
|
||||
</Tooltip>
|
||||
</TooltipProvider>
|
||||
)}
|
||||
{s.conflict_group && (
|
||||
<TooltipProvider>
|
||||
<Tooltip>
|
||||
|
||||
@@ -227,6 +227,11 @@ export default function LyricsAnalyzer({
|
||||
updateDataWithHistory,
|
||||
editLog,
|
||||
getAuthToken: () => getAccessToken() ?? undefined,
|
||||
// Auto-run once on load so suggestions are ready without a click. The
|
||||
// backend pre-generates + caches them during lyrics processing, so this
|
||||
// is normally an instant cache hit. Skipped in read-only / local mode,
|
||||
// and only when a jobId exists (the request needs one).
|
||||
autoRunOnLoad: !isReadOnly && !isLocalMode && Boolean(jobId),
|
||||
})
|
||||
|
||||
// Close the run modal once suggestions arrive (panel takes over)
|
||||
|
||||
@@ -6,7 +6,7 @@ import { test, expect, Page } from '@playwright/test'
|
||||
* Verifies the opt-in AI suggestions flow end-to-end against live production:
|
||||
* 1. Lists jobs filtered by status=in_review (admin can see all jobs).
|
||||
* 2. Navigates to the lyrics review page for the first matching job.
|
||||
* 3. Opens the "AI Suggest" modal and runs suggestion generation.
|
||||
* 3. Opens the "Auto-correct" modal and runs suggestion generation.
|
||||
* 4. Waits for the suggestions panel (up to 120s for the LLM call).
|
||||
* 5. Asserts the panel renders with accept/reject affordances (or the
|
||||
* explicit "no suggestions" state), then dismisses it WITHOUT
|
||||
@@ -66,7 +66,7 @@ test.describe('AI auto-correct suggestions (production)', () => {
|
||||
// exceed 30s (cold Cloud Run instance + uncached correction payload).
|
||||
// A "Saved Review Sessions" dialog may open on load (it aria-hides the
|
||||
// rest of the app); close it without restoring before proceeding.
|
||||
const suggestBtn = page.getByRole('button', { name: /AI Suggest/i })
|
||||
const suggestBtn = page.getByRole('button', { name: /Auto-correct/i })
|
||||
const sessionDialog = page.getByRole('dialog', { name: /Saved Review Sessions/i })
|
||||
await expect(suggestBtn.or(sessionDialog).first()).toBeVisible({ timeout: 90_000 })
|
||||
if (await sessionDialog.isVisible().catch(() => false)) {
|
||||
|
||||
@@ -38,6 +38,9 @@ interface UseAutoCorrectArgs {
|
||||
updateDataWithHistory: (newData: CorrectionData, desc?: string) => void
|
||||
editLog: EditLog
|
||||
getAuthToken: () => string | undefined
|
||||
/** Auto-trigger one run on load (when references exist) so suggestions are
|
||||
* ready without a click. Normally a cache hit (backend pre-generates). */
|
||||
autoRunOnLoad?: boolean
|
||||
}
|
||||
|
||||
export function useAutoCorrect({
|
||||
@@ -46,6 +49,7 @@ export function useAutoCorrect({
|
||||
updateDataWithHistory,
|
||||
editLog,
|
||||
getAuthToken,
|
||||
autoRunOnLoad = false,
|
||||
}: UseAutoCorrectArgs) {
|
||||
const [status, setStatus] = useState<AutoCorrectStatus>('idle')
|
||||
const [suggestions, setSuggestions] = useState<AiSuggestion[]>([])
|
||||
@@ -123,6 +127,17 @@ export function useAutoCorrect({
|
||||
abortRef.current?.abort()
|
||||
}, [])
|
||||
|
||||
// Fire exactly one run on load once references + segments are available.
|
||||
// The ref guard ensures it never re-fires as `data` changes during review.
|
||||
const autoRanRef = useRef(false)
|
||||
useEffect(() => {
|
||||
if (!autoRunOnLoad || autoRanRef.current) return
|
||||
if (!jobId || status !== 'idle') return
|
||||
if (!hasReferences || !(data.corrected_segments?.length > 0)) return
|
||||
autoRanRef.current = true
|
||||
void run()
|
||||
}, [autoRunOnLoad, jobId, status, hasReferences, data.corrected_segments, run])
|
||||
|
||||
const logDecision = useCallback(
|
||||
(s: AiSuggestion, op: 'ai_suggestion_accept' | 'ai_suggestion_reject' | 'ai_suggestion_undo') => {
|
||||
addEditEntry(editLog, op, {
|
||||
|
||||
@@ -35,7 +35,10 @@ export const DEFAULT_AUTO_CORRECT_SETTINGS: AutoCorrectSettings = {
|
||||
suggest_adlib_removal: true,
|
||||
allow_insertions: true,
|
||||
min_confidence: 0,
|
||||
compare_models: false,
|
||||
// Always multi-model. Kept in sync with the backend proactive run
|
||||
// (AutoCorrectSettings(compare_models=True)) so the on-load request is a
|
||||
// cache hit. Not user-configurable.
|
||||
compare_models: true,
|
||||
}
|
||||
|
||||
export interface AiSuggestion {
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "قبول الكل",
|
||||
"revertAll": "تراجع عن الكل",
|
||||
"undoAll": "تراجع عن الكل",
|
||||
"resetAllChanges": "إعادة تعيين جميع التغييرات إلى الكلمات الأصلية المصححة بالذكاء الاصطناعي",
|
||||
"resetAllChanges": "إعادة تعيين جميع التغييرات والعودة إلى الكلمات المصححة الأصلية",
|
||||
"undo": "تراجع",
|
||||
"redo": "إعادة",
|
||||
"findReplace": "بحث/استبدال",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "تعذر تحميل الصوت",
|
||||
"audioStillLoading": "لا يزال الصوت قيد التحميل...",
|
||||
"loadingAudio": "جاري تحميل الصوت...",
|
||||
"autoCorrect": "اقتراحات الذكاء الاصطناعي",
|
||||
"autoCorrectDesc": "احصل على اقتراحات للتصحيح بالذكاء الاصطناعي من خلال مقارنة النص المنسوخ مع الكلمات المرجعية. لن يتم تغيير أي شيء حتى توافق على كل اقتراح."
|
||||
"autoCorrect": "التصحيح التلقائي",
|
||||
"autoCorrectDesc": "إنشاء اقتراحات تصحيح من خلال مقارنة التفريغ الصوتي مع الكلمات المرجعية. لن يتم تغيير أي شيء حتى توافق على كل اقتراح."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "إظهار النصائح",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "أخرى",
|
||||
"category_repeated_chorus_fix": "كورس",
|
||||
"collapse": "طي الاقتراحات",
|
||||
"confidenceHint": "مدى ثقة الذكاء الاصطناعي المطلوبة لعرض الاقتراح",
|
||||
"confidenceHint": "مستوى الثقة المطلوب للاقتراح قبل ظهوره",
|
||||
"confidenceLabel": "تصفية الاقتراحات",
|
||||
"confidence_all": "عرض الكل",
|
||||
"confidence_high": "ثقة عالية",
|
||||
"description": "يقارن الذكاء الاصطناعي الكلمات المنسوخة مع المصادر المرجعية ويقترح تعديلات على مستوى الكلمة. يمكنك مراجعة كل اقتراح — ولن يتم تطبيق أي تغيير حتى توافق عليه.",
|
||||
"description": "يقارن التصحيح التلقائي التفريغ الصوتي بالمصادر المرجعية ويقترح تعديلات على مستوى الكلمة. يمكنك مراجعة كل اقتراح — ولن يتم تطبيق أي شيء حتى توافق عليه.",
|
||||
"dismiss": "تجاهل الاقتراحات",
|
||||
"expand": "توسيع الاقتراحات",
|
||||
"insertMarker": "إدراج:",
|
||||
"insertionsHint": "السماح بالاقتراحات التي تدرج الكلمات المفقودة في النسخ",
|
||||
"insertionsLabel": "اقتراح الكلمات المفقودة",
|
||||
"noReferences": "لا تتوفر كلمات مرجعية لهذه الأغنية، لذا لا يمكن إنشاء اقتراحات الذكاء الاصطناعي. أضف الكلمات من اللوحة المرجعية أولاً.",
|
||||
"noSuggestions": "لم يتم العثور على أي مشاكل — لا توجد تغييرات مقترحة من الذكاء الاصطناعي.",
|
||||
"noReferences": "لا تتوفر كلمات مرجعية لهذه الأغنية، لذا لا يمكن إنشاء اقتراحات. أضف الكلمات من اللوحة المرجعية أولاً.",
|
||||
"noSuggestions": "لم يتم العثور على أي مشاكل — لا توجد تغييرات مقترحة.",
|
||||
"reject": "رفض الاقتراح",
|
||||
"rejectAll": "رفض الكل",
|
||||
"rejected": "مرفوض",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "جاري تحليل الكلمات... يستغرق هذا عادةً بين 15 و 60 ثانية.",
|
||||
"staleHint": "تم تعديل هذه الكلمات بعد إنشاء الاقتراحات، لذا لم يعد هذا الاقتراح صالحاً.",
|
||||
"summary": "{total} اقتراحات · {pending} قيد الانتظار · {accepted} مقبولة",
|
||||
"title": "اقتراحات التصحيح بالذكاء الاصطناعي",
|
||||
"title": "اقتراحات التصحيح التلقائي",
|
||||
"undo": "التراجع عن هذا الاقتراح",
|
||||
"cancel": "إلغاء",
|
||||
"close": "إغلاق",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "أبطأ وأعلى تكلفة، لكنه يوضح الاقتراحات التي تتفق عليها عدة نماذج",
|
||||
"compareLabel": "مقارنة عدة نماذج للذكاء الاصطناعي",
|
||||
"conflictBadge": "بديل",
|
||||
"conflictTooltip": "تختلف النماذج حول هذه الكلمات — قبول أحد البدائل يعني رفض الأخرى.",
|
||||
"conflictTooltip": "توجد اقتراحات مختلفة لهذه الكلمات — قبول أحد الخيارات يؤدي إلى رفض الخيارات الأخرى.",
|
||||
"consensusBadge": "{consensus}/{total} من النماذج",
|
||||
"consensusTooltip": "مقترح من: {models}",
|
||||
"cachedBadge": "مخزن مؤقتاً",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Accepta-ho tot",
|
||||
"revertAll": "Reverteix-ho tot",
|
||||
"undoAll": "Desfés-ho tot",
|
||||
"resetAllChanges": "Restableix tots els canvis a les lletres originals corregides per IA",
|
||||
"resetAllChanges": "Restableix tots els canvis a la lletra corregida original",
|
||||
"undo": "Desfés",
|
||||
"redo": "Refés",
|
||||
"findReplace": "Cerca/Reemplaça",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "No s'ha pogut carregar l'àudio",
|
||||
"audioStillLoading": "L'àudio encara s'està carregant...",
|
||||
"loadingAudio": "Carregant l'àudio...",
|
||||
"autoCorrect": "Suggeriments de la IA",
|
||||
"autoCorrectDesc": "Genera suggeriments de correcció de la IA comparant la transcripció amb la lletra de referència. No es canvia res fins que acceptis cada suggeriment."
|
||||
"autoCorrect": "Autocorrecció",
|
||||
"autoCorrectDesc": "Genera suggeriments de correcció comparant la transcripció amb la lletra de referència. No es canvia res fins que acceptes cada suggeriment."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Mostra els consells",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "altres",
|
||||
"category_repeated_chorus_fix": "tornada",
|
||||
"collapse": "Contreu els suggeriments",
|
||||
"confidenceHint": "Quin grau de certesa ha de tenir la IA per mostrar un suggeriment",
|
||||
"confidenceHint": "Quin nivell de confiança ha de tenir un suggeriment abans de mostrar-se",
|
||||
"confidenceLabel": "Filtre de suggeriments",
|
||||
"confidence_all": "Mostra-ho tot",
|
||||
"confidence_high": "Certesa alta",
|
||||
"description": "La IA compara la lletra transcrita amb les fonts de referència i suggereix correccions paraula per paraula. Tu revises cada suggeriment — no s'aplica res fins que ho acceptes.",
|
||||
"description": "L'autocorrecció compara la lletra transcrita amb les fonts de referència i suggereix correccions a nivell de paraula. Tu revises cada suggeriment — no s'aplica res fins que ho acceptes.",
|
||||
"dismiss": "Descarta els suggeriments",
|
||||
"expand": "Desplega els suggeriments",
|
||||
"insertMarker": "insereix:",
|
||||
"insertionsHint": "Permet suggeriments que insereixin paraules que la transcripció ha omès",
|
||||
"insertionsLabel": "Suggereix paraules que falten",
|
||||
"noReferences": "No hi ha cap lletra de referència disponible per a aquesta cançó, per la qual cosa no es poden generar suggeriments de la IA. Primer, afegeix la lletra des del tauler de referència.",
|
||||
"noSuggestions": "No s'ha trobat cap problema — la IA no suggereix cap canvi.",
|
||||
"noReferences": "No hi ha cap lletra de referència disponible per a aquesta cançó, així que no es poden generar suggeriments. Afegeix primer la lletra des del panell de referència.",
|
||||
"noSuggestions": "No s'han trobat problemes — no hi ha cap canvi suggerit.",
|
||||
"reject": "Rebutja el suggeriment",
|
||||
"rejectAll": "Rebutja-ho tot",
|
||||
"rejected": "Rebutjat",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "S'està analitzant la lletra... això sol trigar entre 15 i 60 segons.",
|
||||
"staleHint": "Aquestes paraules s'han editat des que es van generar els suggeriments, per la qual cosa aquest suggeriment ja no és vàlid.",
|
||||
"summary": "{total} suggeriments · {pending} pendents · {accepted} acceptats",
|
||||
"title": "Suggeriments de correcció de la IA",
|
||||
"title": "Suggeriments d'autocorrecció",
|
||||
"undo": "Desfés aquest suggeriment",
|
||||
"cancel": "Cancel·la",
|
||||
"close": "Tanca",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Més lent i més car, però mostra en quins suggeriments coincideixen diversos models",
|
||||
"compareLabel": "Compara diversos models d'IA",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Els models no coincideixen en aquestes paraules — si acceptes una variant, rebutges les altres.",
|
||||
"conflictTooltip": "Hi ha diferents suggeriments per a aquestes paraules — si acceptes una variant, es rebutgen les altres.",
|
||||
"consensusBadge": "{consensus}/{total} models",
|
||||
"consensusTooltip": "Proposat per: {models}",
|
||||
"cachedBadge": "en memòria cau",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Přijmout vše",
|
||||
"revertAll": "Vrátit vše",
|
||||
"undoAll": "Vrátit vše zpět",
|
||||
"resetAllChanges": "Resetovat všechny změny zpět na původní text opravený AI",
|
||||
"resetAllChanges": "Vrátit všechny změny zpět na původní opravený text",
|
||||
"undo": "Zpět",
|
||||
"redo": "Znovu",
|
||||
"findReplace": "Najít/Nahradit",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Nepodařilo se načíst audio",
|
||||
"audioStillLoading": "Audio se stále načítá...",
|
||||
"loadingAudio": "Načítání audia...",
|
||||
"autoCorrect": "Návrhy AI",
|
||||
"autoCorrectDesc": "Nech AI vygenerovat návrhy oprav porovnáním přepisu s referenčním textem. Nic se nezmění, dokud jednotlivé návrhy nepřijmeš."
|
||||
"autoCorrect": "Automatická oprava",
|
||||
"autoCorrectDesc": "Vygenerovat návrhy oprav porovnáním přepisu s referenčním textem. Nic se nezmění, dokud každý návrh nepřijmeš."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Zobrazit tipy",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "ostatní",
|
||||
"category_repeated_chorus_fix": "refrén",
|
||||
"collapse": "Sbalit návrhy",
|
||||
"confidenceHint": "Jak moc si musí být AI jistá, aby zobrazila návrh",
|
||||
"confidenceHint": "Jak vysoká musí být jistota návrhu, než se zobrazí",
|
||||
"confidenceLabel": "Filtr návrhů",
|
||||
"confidence_all": "Zobrazit vše",
|
||||
"confidence_high": "Vysoká jistota",
|
||||
"description": "AI porovná přepsaný text s referenčními zdroji a navrhne opravy jednotlivých slov. Každý návrh si můžeš zkontrolovat — nic se nezmění, dokud ho nepřijmeš.",
|
||||
"description": "Automatická oprava porovnává přepsaný text s referenčními zdroji a navrhuje opravy na úrovni slov. Každý návrh si zkontroluješ — nic se neaplikuje, dokud ho nepřijmeš.",
|
||||
"dismiss": "Skrýt návrhy",
|
||||
"expand": "Rozbalit návrhy",
|
||||
"insertMarker": "vložit:",
|
||||
"insertionsHint": "Povolit návrhy, které vkládají slova, která v přepisu chybí",
|
||||
"insertionsLabel": "Navrhnout chybějící slova",
|
||||
"noReferences": "Pro tuto píseň nejsou k dispozici žádné referenční texty, takže nelze vygenerovat návrhy AI. Nejprve přidej text z referenčního panelu.",
|
||||
"noSuggestions": "Nebyly nalezeny žádné problémy — AI nemá žádné návrhy na změnu.",
|
||||
"noReferences": "Pro tuto píseň nejsou k dispozici žádné referenční texty, takže nelze vygenerovat návrhy. Nejprve přidej text z referenčního panelu.",
|
||||
"noSuggestions": "Nebyly nalezeny žádné problémy — nejsou navrženy žádné změny.",
|
||||
"reject": "Odmítnout návrh",
|
||||
"rejectAll": "Odmítnout vše",
|
||||
"rejected": "Odmítnuto",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analyzuji text… obvykle to trvá 15–60 sekund.",
|
||||
"staleHint": "Tato slova byla upravena od chvíle, kdy byly návrhy vygenerovány, takže tento návrh již neplatí.",
|
||||
"summary": "{total} návrhů · {pending} čeká · {accepted} přijato",
|
||||
"title": "Návrhy oprav AI",
|
||||
"title": "Návrhy automatické opravy",
|
||||
"undo": "Vzít tento návrh zpět",
|
||||
"cancel": "Zrušit",
|
||||
"close": "Zavřít",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Pomalejší a dražší, ale ukáže, na kterých návrzích se shodne více modelů",
|
||||
"compareLabel": "Porovnat více AI modelů",
|
||||
"conflictBadge": "varianta",
|
||||
"conflictTooltip": "Modely se na těchto slovech neshodují — přijetím jedné varianty odmítneš ostatní.",
|
||||
"conflictTooltip": "Pro tato slova existují různé návrhy — přijetím jedné varianty odmítneš ostatní.",
|
||||
"consensusBadge": "{consensus}/{total} modelů",
|
||||
"consensusTooltip": "Navrženo: {models}",
|
||||
"cachedBadge": "z mezipaměti",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Accepter alle",
|
||||
"revertAll": "Gendan alle",
|
||||
"undoAll": "Fortryd alle",
|
||||
"resetAllChanges": "Nulstil alle ændringer tilbage til de originale AI-rettede sangtekster",
|
||||
"resetAllChanges": "Nulstil alle ændringer tilbage til den oprindelige rettede sangtekst",
|
||||
"undo": "Fortryd",
|
||||
"redo": "Gendan",
|
||||
"findReplace": "Find/Erstat",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Kunne ikke indlæse lyd",
|
||||
"audioStillLoading": "Lyden indlæses stadig...",
|
||||
"loadingAudio": "Indlæser lyd...",
|
||||
"autoCorrect": "AI-forslag",
|
||||
"autoCorrectDesc": "Generer AI-rettelsesforslag ved at sammenligne transskriptionen med referenceteksten. Intet ændres, før du accepterer hvert forslag."
|
||||
"autoCorrect": "Autokorrektur",
|
||||
"autoCorrectDesc": "Generer forslag til rettelser ved at sammenligne transskriptionen med referenceteksten. Intet ændres, før du accepterer hvert forslag."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Vis tips",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "andet",
|
||||
"category_repeated_chorus_fix": "omkvæd",
|
||||
"collapse": "Skjul forslag",
|
||||
"confidenceHint": "Hvor sikker AI'en skal være for at vise et forslag",
|
||||
"confidenceHint": "Hvor sikkert et forslag skal være, før det vises",
|
||||
"confidenceLabel": "Forslagsfilter",
|
||||
"confidence_all": "Vis alle",
|
||||
"confidence_high": "Høj sikkerhed",
|
||||
"description": "AI'en sammenligner den transskriberede tekst med referencekilderne og foreslår rettelser på ordniveau. Du gennemgår hvert forslag — intet ændres, før du accepterer det.",
|
||||
"description": "Autokorrektur sammenligner den transskriberede sangtekst med referencekilderne og foreslår rettelser på ordniveau. Du gennemgår hvert forslag — intet anvendes, før du accepterer det.",
|
||||
"dismiss": "Fjern forslag",
|
||||
"expand": "Udvid forslag",
|
||||
"insertMarker": "indsæt:",
|
||||
"insertionsHint": "Tillad forslag, der indsætter ord, som transskriptionen overså",
|
||||
"insertionsLabel": "Foreslå manglende ord",
|
||||
"noReferences": "Der er ingen referencetekster tilgængelige for denne sang, så der kan ikke genereres AI-forslag. Tilføj tekster fra referencepanelet først.",
|
||||
"noSuggestions": "Ingen problemer fundet — AI'en har ingen foreslåede ændringer.",
|
||||
"noReferences": "Der er ingen referencetekster tilgængelige for denne sang, så der kan ikke genereres forslag. Tilføj en sangtekst fra referencepanelet først.",
|
||||
"noSuggestions": "Ingen problemer fundet — der er ingen foreslåede ændringer.",
|
||||
"reject": "Afvis forslag",
|
||||
"rejectAll": "Afvis alle",
|
||||
"rejected": "Afvist",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analyserer tekst… dette tager normalt 15–60 sekunder.",
|
||||
"staleHint": "Disse ord er blevet redigeret, siden forslagene blev genereret, så dette forslag gælder ikke længere.",
|
||||
"summary": "{total} forslag · {pending} afventer · {accepted} accepteret",
|
||||
"title": "AI-rettelsesforslag",
|
||||
"title": "Forslag til autokorrektur",
|
||||
"undo": "Fortryd dette forslag",
|
||||
"cancel": "Annuller",
|
||||
"close": "Luk",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Langsommere og koster mere, men viser, hvilke forslag flere modeller er enige om",
|
||||
"compareLabel": "Sammenlign flere AI-modeller",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Modellerne er uenige om disse ord — hvis du accepterer én variant, afvises de andre.",
|
||||
"conflictTooltip": "Der er forskellige forslag til disse ord — hvis du accepterer én variant, afvises de andre.",
|
||||
"consensusBadge": "{consensus}/{total} modeller",
|
||||
"consensusTooltip": "Foreslået af: {models}",
|
||||
"cachedBadge": "cachet",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Alle akzeptieren",
|
||||
"revertAll": "Alle verwerfen",
|
||||
"undoAll": "Alle rückgängig machen",
|
||||
"resetAllChanges": "Alle Änderungen auf den ursprünglichen KI-korrigierten Text zurücksetzen",
|
||||
"resetAllChanges": "Alle Änderungen auf den ursprünglich korrigierten Text zurücksetzen",
|
||||
"undo": "Rückgängig",
|
||||
"redo": "Wiederholen",
|
||||
"findReplace": "Suchen/Ersetzen",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Audio konnte nicht geladen werden",
|
||||
"audioStillLoading": "Audio wird noch geladen...",
|
||||
"loadingAudio": "Audio wird geladen...",
|
||||
"autoCorrect": "KI-Vorschläge",
|
||||
"autoCorrectDesc": "Generiere KI-Korrekturvorschläge durch den Vergleich der Transkription mit dem Referenztext. Nichts wird geändert, bis du jeden Vorschlag akzeptierst."
|
||||
"autoCorrect": "Autokorrektur",
|
||||
"autoCorrectDesc": "Generiere Korrekturvorschläge, indem du die Transkription mit dem Referenztext vergleichst. Nichts wird geändert, bis du jeden Vorschlag akzeptierst."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Tipps anzeigen",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "Sonstiges",
|
||||
"category_repeated_chorus_fix": "Refrain",
|
||||
"collapse": "Vorschläge einklappen",
|
||||
"confidenceHint": "Wie sicher die KI sein muss, um einen Vorschlag anzuzeigen",
|
||||
"confidenceHint": "Wie sicher ein Vorschlag sein muss, bevor er angezeigt wird",
|
||||
"confidenceLabel": "Vorschlagsfilter",
|
||||
"confidence_all": "Alle anzeigen",
|
||||
"confidence_high": "Hohe Sicherheit",
|
||||
"description": "Die KI vergleicht den transkribierten Text mit den Referenzquellen und schlägt Korrekturen auf Wortebene vor. Du überprüfst jeden Vorschlag — nichts wird angewendet, bis du ihn akzeptierst.",
|
||||
"description": "Die Autokorrektur vergleicht den transkribierten Text mit den Referenzquellen und schlägt Korrekturen auf Wortebene vor. Du überprüfst jeden Vorschlag — nichts wird angewendet, bis du ihn akzeptierst.",
|
||||
"dismiss": "Vorschläge verwerfen",
|
||||
"expand": "Vorschläge ausklappen",
|
||||
"insertMarker": "einfügen:",
|
||||
"insertionsHint": "Erlaube Vorschläge, die Wörter einfügen, die bei der Transkription übersehen wurden",
|
||||
"insertionsLabel": "Fehlende Wörter vorschlagen",
|
||||
"noReferences": "Für diesen Song ist kein Referenztext verfügbar, daher können keine KI-Vorschläge generiert werden. Füge zuerst Text aus dem Referenz-Panel hinzu.",
|
||||
"noSuggestions": "Keine Probleme gefunden — die KI hat keine Änderungsvorschläge.",
|
||||
"noReferences": "Für diesen Song ist kein Referenztext verfügbar, daher können keine Vorschläge generiert werden. Füge zuerst einen Text über das Referenz-Panel hinzu.",
|
||||
"noSuggestions": "Keine Probleme gefunden — es gibt keine Änderungsvorschläge.",
|
||||
"reject": "Vorschlag ablehnen",
|
||||
"rejectAll": "Alle ablehnen",
|
||||
"rejected": "Abgelehnt",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Text wird analysiert… dies dauert normalerweise 15–60 Sekunden.",
|
||||
"staleHint": "Diese Wörter wurden bearbeitet, seit die Vorschläge generiert wurden, daher gilt dieser Vorschlag nicht mehr.",
|
||||
"summary": "{total} Vorschläge · {pending} ausstehend · {accepted} akzeptiert",
|
||||
"title": "KI-Korrekturvorschläge",
|
||||
"title": "Autokorrektur-Vorschläge",
|
||||
"undo": "Diesen Vorschlag rückgängig machen",
|
||||
"cancel": "Abbrechen",
|
||||
"close": "Schließen",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Langsamer und teurer, zeigt aber, bei welchen Vorschlägen mehrere Modelle übereinstimmen",
|
||||
"compareLabel": "Mehrere KI-Modelle vergleichen",
|
||||
"conflictBadge": "Variante",
|
||||
"conflictTooltip": "Die Modelle sind sich bei diesen Wörtern uneinig — wenn du eine Variante akzeptierst, werden die anderen abgelehnt.",
|
||||
"conflictTooltip": "Es gibt verschiedene Vorschläge für diese Wörter — wenn du eine Variante akzeptierst, werden die anderen abgelehnt.",
|
||||
"consensusBadge": "{consensus}/{total} Modelle",
|
||||
"consensusTooltip": "Vorgeschlagen von: {models}",
|
||||
"cachedBadge": "zwischengespeichert",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Αποδοχή Όλων",
|
||||
"revertAll": "Επαναφορά Όλων",
|
||||
"undoAll": "Αναίρεση Όλων",
|
||||
"resetAllChanges": "Επαναφορά όλων των αλλαγών πίσω στους αρχικούς στίχους που διορθώθηκαν από το AI",
|
||||
"resetAllChanges": "Επαναφορά όλων των αλλαγών στους αρχικούς διορθωμένους στίχους",
|
||||
"undo": "Αναίρεση",
|
||||
"redo": "Επανάληψη",
|
||||
"findReplace": "Εύρεση/Αντικατάσταση",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Δεν ήταν δυνατή η φόρτωση του ήχου",
|
||||
"audioStillLoading": "Ο ήχος φορτώνει ακόμα...",
|
||||
"loadingAudio": "Φόρτωση ήχου...",
|
||||
"autoCorrect": "Προτάσεις AI",
|
||||
"autoCorrectDesc": "Δημιούργησε προτάσεις διόρθωσης AI συγκρίνοντας τη μεταγραφή με τους στίχους αναφοράς. Τίποτα δεν αλλάζει μέχρι να αποδεχτείς κάθε πρόταση."
|
||||
"autoCorrect": "Αυτόματη διόρθωση",
|
||||
"autoCorrectDesc": "Δημιούργησε προτάσεις διόρθωσης συγκρίνοντας τη μεταγραφή με τους στίχους αναφοράς. Τίποτα δεν αλλάζει μέχρι να αποδεχτείς κάθε πρόταση."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Εμφάνιση συμβουλών",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "άλλο",
|
||||
"category_repeated_chorus_fix": "ρεφρέν",
|
||||
"collapse": "Σύμπτυξη προτάσεων",
|
||||
"confidenceHint": "Πόσο σίγουρο πρέπει να είναι το AI για να εμφανίσει μια πρόταση",
|
||||
"confidenceHint": "Πόσο βέβαιη πρέπει να είναι μια πρόταση πριν εμφανιστεί",
|
||||
"confidenceLabel": "Φίλτρο προτάσεων",
|
||||
"confidence_all": "Εμφάνιση όλων",
|
||||
"confidence_high": "Υψηλή βεβαιότητα",
|
||||
"description": "Το AI συγκρίνει τους μεταγραμμένους στίχους με τις πηγές αναφοράς και προτείνει διορθώσεις σε επίπεδο λέξης. Εσύ ελέγχεις κάθε πρόταση — τίποτα δεν εφαρμόζεται μέχρι να το αποδεχτείς.",
|
||||
"description": "Η αυτόματη διόρθωση συγκρίνει τους μεταγραμμένους στίχους με τις πηγές αναφοράς και προτείνει διορθώσεις σε επίπεδο λέξης. Ελέγχεις κάθε πρόταση — τίποτα δεν εφαρμόζεται μέχρι να την αποδεχτείς.",
|
||||
"dismiss": "Αγνόηση προτάσεων",
|
||||
"expand": "Ανάπτυξη προτάσεων",
|
||||
"insertMarker": "εισαγωγή:",
|
||||
"insertionsHint": "Επίτρεψε προτάσεις που εισάγουν λέξεις που παραλείφθηκαν κατά τη μεταγραφή",
|
||||
"insertionsLabel": "Πρόταση λέξεων που λείπουν",
|
||||
"noReferences": "Δεν υπάρχουν διαθέσιμοι στίχοι αναφοράς για αυτό το τραγούδι, επομένως δεν μπορούν να δημιουργηθούν προτάσεις AI. Πρόσθεσε στίχους από το πάνελ αναφοράς πρώτα.",
|
||||
"noSuggestions": "Δεν βρέθηκαν προβλήματα — το AI δεν έχει προτεινόμενες αλλαγές.",
|
||||
"noReferences": "Δεν υπάρχουν διαθέσιμοι στίχοι αναφοράς για αυτό το τραγούδι, επομένως δεν μπορούν να δημιουργηθούν προτάσεις. Πρόσθεσε πρώτα στίχους από τον πίνακα αναφοράς.",
|
||||
"noSuggestions": "Δεν βρέθηκαν προβλήματα — δεν υπάρχουν προτεινόμενες αλλαγές.",
|
||||
"reject": "Απόρριψη πρότασης",
|
||||
"rejectAll": "Απόρριψη όλων",
|
||||
"rejected": "Απορρίφθηκε",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Ανάλυση στίχων… αυτό συνήθως διαρκεί 15–60 δευτερόλεπτα.",
|
||||
"staleHint": "Αυτές οι λέξεις επεξεργάστηκαν μετά τη δημιουργία των προτάσεων, επομένως αυτή η πρόταση δεν ισχύει πλέον.",
|
||||
"summary": "{total} προτάσεις · {pending} σε εκκρεμότητα · {accepted} αποδεκτές",
|
||||
"title": "Προτάσεις διόρθωσης AI",
|
||||
"title": "Προτάσεις αυτόματης διόρθωσης",
|
||||
"undo": "Αναίρεση αυτής της πρότασης",
|
||||
"cancel": "Ακύρωση",
|
||||
"close": "Κλείσιμο",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Πιο αργό και κοστίζει περισσότερο, αλλά δείχνει σε ποιες προτάσεις συμφωνούν πολλαπλά μοντέλα",
|
||||
"compareLabel": "Σύγκρινε πολλαπλά μοντέλα AI",
|
||||
"conflictBadge": "παραλλαγή",
|
||||
"conflictTooltip": "Τα μοντέλα διαφωνούν για αυτές τις λέξεις — αν αποδεχτείς μια παραλλαγή, απορρίπτεις τις υπόλοιπες.",
|
||||
"conflictTooltip": "Υπάρχουν διαφορετικές προτάσεις για αυτές τις λέξεις — η αποδοχή μιας παραλλαγής απορρίπτει τις υπόλοιπες.",
|
||||
"consensusBadge": "{consensus}/{total} μοντέλα",
|
||||
"consensusTooltip": "Προτάθηκε από: {models}",
|
||||
"cachedBadge": "προσωρινά αποθηκευμένο",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Accept All",
|
||||
"revertAll": "Revert All",
|
||||
"undoAll": "Undo All",
|
||||
"resetAllChanges": "Reset all changes back to the original AI-corrected lyrics",
|
||||
"resetAllChanges": "Reset all changes back to the original corrected lyrics",
|
||||
"undo": "Undo",
|
||||
"redo": "Redo",
|
||||
"findReplace": "Find/Replace",
|
||||
@@ -821,8 +821,8 @@
|
||||
"loadingAudio": "Loading audio...",
|
||||
"audioStillLoading": "Audio is still loading...",
|
||||
"audioLoadError": "Couldn't load audio",
|
||||
"autoCorrect": "AI Suggest",
|
||||
"autoCorrectDesc": "Generate AI correction suggestions by comparing the transcription against the reference lyrics. Nothing is changed until you accept each suggestion."
|
||||
"autoCorrect": "Auto-correct",
|
||||
"autoCorrectDesc": "Generate correction suggestions by comparing the transcription against the reference lyrics. Nothing is changed until you accept each suggestion."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Show tips",
|
||||
@@ -1282,23 +1282,23 @@
|
||||
"resetOverridesTooltip": "Reset word overrides in this segment"
|
||||
},
|
||||
"autoCorrect": {
|
||||
"title": "AI correction suggestions",
|
||||
"description": "The AI compares the transcribed lyrics against the reference sources and suggests word-level fixes. You review every suggestion — nothing is applied until you accept it.",
|
||||
"noReferences": "No reference lyrics are available for this song, so AI suggestions can’t be generated. Add lyrics from the reference panel first.",
|
||||
"title": "Auto-correct suggestions",
|
||||
"description": "Auto-correct compares the transcribed lyrics against the reference sources and suggests word-level fixes. You review every suggestion — nothing is applied until you accept it.",
|
||||
"noReferences": "No reference lyrics are available for this song, so suggestions can’t be generated. Add lyrics from the reference panel first.",
|
||||
"running": "Analyzing lyrics… this usually takes 15–60 seconds.",
|
||||
"adlibRemovalLabel": "Suggest removing adlibs",
|
||||
"adlibRemovalHint": "Allow suggestions that delete backing vocals the references don’t have",
|
||||
"insertionsLabel": "Suggest missing words",
|
||||
"insertionsHint": "Allow suggestions that insert words the transcription missed",
|
||||
"confidenceLabel": "Suggestion filter",
|
||||
"confidenceHint": "How confident the AI must be to show a suggestion",
|
||||
"confidenceHint": "How confident a suggestion must be before it’s shown",
|
||||
"confidence_all": "Show all",
|
||||
"confidence_balanced": "Balanced",
|
||||
"confidence_high": "High confidence",
|
||||
"cancel": "Cancel",
|
||||
"close": "Close",
|
||||
"runButton": "Get suggestions",
|
||||
"noSuggestions": "No issues found — the AI has no suggested changes.",
|
||||
"noSuggestions": "No issues found — there are no suggested changes.",
|
||||
"summary": "{total} suggestions · {pending} pending · {accepted} accepted",
|
||||
"acceptAll": "Accept all",
|
||||
"rejectAll": "Reject all",
|
||||
@@ -1323,7 +1323,7 @@
|
||||
"consensusBadge": "{consensus}/{total} models",
|
||||
"consensusTooltip": "Proposed by: {models}",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Models disagree about these words — accepting one variant rejects the others.",
|
||||
"conflictTooltip": "There are different suggestions for these words — accepting one variant rejects the others.",
|
||||
"cachedBadge": "cached",
|
||||
"cachedTooltip": "Restored from a previous run on the same lyrics — no AI cost was incurred."
|
||||
}
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Aceptar todo",
|
||||
"revertAll": "Revertir todo",
|
||||
"undoAll": "Deshacer todo",
|
||||
"resetAllChanges": "Restablecer todos los cambios a las letras originales corregidas por IA",
|
||||
"resetAllChanges": "Restablecer todos los cambios a las letras corregidas originales",
|
||||
"undo": "Deshacer",
|
||||
"redo": "Rehacer",
|
||||
"findReplace": "Buscar/Reemplazar",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "No se pudo cargar el audio",
|
||||
"audioStillLoading": "El audio aún se está cargando...",
|
||||
"loadingAudio": "Cargando audio...",
|
||||
"autoCorrect": "Sugerencia de IA",
|
||||
"autoCorrectDesc": "Genera sugerencias de corrección de IA comparando la transcripción con las letras de referencia. No se cambia nada hasta que aceptes cada sugerencia."
|
||||
"autoCorrect": "Autocorrección",
|
||||
"autoCorrectDesc": "Genera sugerencias de corrección comparando la transcripción con las letras de referencia. No se cambia nada hasta que aceptas cada sugerencia."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Mostrar consejos",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "otro",
|
||||
"category_repeated_chorus_fix": "estribillo",
|
||||
"collapse": "Contraer sugerencias",
|
||||
"confidenceHint": "Qué tan segura debe estar la IA para mostrar una sugerencia",
|
||||
"confidenceHint": "Qué nivel de confianza debe tener una sugerencia para mostrarse",
|
||||
"confidenceLabel": "Filtro de sugerencias",
|
||||
"confidence_all": "Mostrar todas",
|
||||
"confidence_high": "Alta confianza",
|
||||
"description": "La IA compara las letras transcritas con las fuentes de referencia y sugiere correcciones a nivel de palabra. Tú revisas cada sugerencia — no se aplica nada hasta que la aceptes.",
|
||||
"description": "La autocorrección compara las letras transcritas con las fuentes de referencia y sugiere correcciones a nivel de palabra. Revisas cada sugerencia — no se aplica nada hasta que la aceptas.",
|
||||
"dismiss": "Descartar sugerencias",
|
||||
"expand": "Expandir sugerencias",
|
||||
"insertMarker": "insertar:",
|
||||
"insertionsHint": "Permitir sugerencias que inserten palabras que la transcripción omitió",
|
||||
"insertionsLabel": "Sugerir palabras faltantes",
|
||||
"noReferences": "No hay letras de referencia disponibles para esta canción, por lo que no se pueden generar sugerencias de IA. Agrega letras desde el panel de referencia primero.",
|
||||
"noSuggestions": "No se encontraron problemas — la IA no tiene cambios sugeridos.",
|
||||
"noReferences": "No hay letras de referencia disponibles para esta canción, por lo que no se pueden generar sugerencias. Primero añade letras desde el panel de referencia.",
|
||||
"noSuggestions": "No se encontraron problemas — no hay cambios sugeridos.",
|
||||
"reject": "Rechazar sugerencia",
|
||||
"rejectAll": "Rechazar todas",
|
||||
"rejected": "Rechazada",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analizando letras… esto suele tardar entre 15 y 60 segundos.",
|
||||
"staleHint": "Estas palabras fueron editadas desde que se generaron las sugerencias, por lo que esta sugerencia ya no es válida.",
|
||||
"summary": "{total} sugerencias · {pending} pendientes · {accepted} aceptadas",
|
||||
"title": "Sugerencias de corrección de IA",
|
||||
"title": "Sugerencias de autocorrección",
|
||||
"undo": "Deshacer esta sugerencia",
|
||||
"cancel": "Cancelar",
|
||||
"close": "Cerrar",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Es más lento y cuesta más, pero muestra en qué sugerencias coinciden varios modelos",
|
||||
"compareLabel": "Comparar varios modelos de IA",
|
||||
"conflictBadge": "variante",
|
||||
"conflictTooltip": "Los modelos no coinciden en estas palabras; al aceptar una variante, se rechazan las demás.",
|
||||
"conflictTooltip": "Hay diferentes sugerencias para estas palabras — al aceptar una variante se rechazan las demás.",
|
||||
"consensusBadge": "{consensus}/{total} modelos",
|
||||
"consensusTooltip": "Propuesto por: {models}",
|
||||
"cachedBadge": "en caché",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Hyväksy kaikki",
|
||||
"revertAll": "Palauta kaikki",
|
||||
"undoAll": "Kumoa kaikki",
|
||||
"resetAllChanges": "Palauta kaikki muutokset takaisin alkuperäisiin tekoälyn korjaamiin sanoituksiin",
|
||||
"resetAllChanges": "Peru kaikki muutokset ja palauta alkuperäiset korjatut sanoitukset",
|
||||
"undo": "Kumoa",
|
||||
"redo": "Tee uudelleen",
|
||||
"findReplace": "Etsi/Korvaa",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Ääntä ei voitu ladata",
|
||||
"audioStillLoading": "Ääni latautuu yhä...",
|
||||
"loadingAudio": "Ladataan ääntä...",
|
||||
"autoCorrect": "Tekoälyn ehdotukset",
|
||||
"autoCorrectDesc": "Luo tekoälyn korjausehdotuksia vertaamalla transkriptiota viitesanoituksiin. Mitään ei muuteta ennen kuin hyväksyt jokaisen ehdotuksen."
|
||||
"autoCorrect": "Automaattinen korjaus",
|
||||
"autoCorrectDesc": "Luo korjausehdotuksia vertaamalla litterointia vertailusanoituksiin. Mitään ei muuteta ennen kuin hyväksyt jokaisen ehdotuksen."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Näytä vinkit",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "muu",
|
||||
"category_repeated_chorus_fix": "kertosäe",
|
||||
"collapse": "Pienennä ehdotukset",
|
||||
"confidenceHint": "Kuinka varma tekoälyn on oltava näyttääkseen ehdotuksen",
|
||||
"confidenceHint": "Kuinka varma ehdotuksen pitää olla, jotta se näytetään",
|
||||
"confidenceLabel": "Ehdotusten suodatin",
|
||||
"confidence_all": "Näytä kaikki",
|
||||
"confidence_high": "Korkea varmuus",
|
||||
"description": "Tekoäly vertaa transkriptoituja sanoituksia viitelähteisiin ja ehdottaa sanakohtaisia korjauksia. Tarkistat jokaisen ehdotuksen — mitään ei muuteta ennen kuin hyväksyt sen.",
|
||||
"description": "Automaattinen korjaus vertaa litteroituja sanoituksia vertailulähteisiin ja ehdottaa sanakohtaisia korjauksia. Käyt läpi jokaisen ehdotuksen — mitään ei oteta käyttöön ennen kuin hyväksyt sen.",
|
||||
"dismiss": "Ohita ehdotukset",
|
||||
"expand": "Laajenna ehdotukset",
|
||||
"insertMarker": "lisää:",
|
||||
"insertionsHint": "Salli ehdotukset, jotka lisäävät transkriptiosta puuttuvia sanoja",
|
||||
"insertionsLabel": "Ehdotukset puuttuvien sanojen lisäämisestä",
|
||||
"noReferences": "Tälle kappaleelle ei ole saatavilla viitesanoituksia, joten tekoälyehdotuksia ei voida luoda. Lisää sanoitukset ensin viitepaneelista.",
|
||||
"noSuggestions": "Ongelmia ei löytynyt — tekoäly ei ehdota muutoksia.",
|
||||
"noReferences": "Tälle kappaleelle ei ole saatavilla vertailusanoituksia, joten ehdotuksia ei voida luoda. Lisää sanoitukset ensin vertailupaneelista.",
|
||||
"noSuggestions": "Ongelmia ei löytynyt — ei ehdotettuja muutoksia.",
|
||||
"reject": "Hylkää ehdotus",
|
||||
"rejectAll": "Hylkää kaikki",
|
||||
"rejected": "Hylätty",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analysoidaan sanoituksia… tässä kestää yleensä 15–60 sekuntia.",
|
||||
"staleHint": "Näitä sanoja on muokattu sen jälkeen, kun ehdotukset luotiin, joten tämä ehdotus ei enää päde.",
|
||||
"summary": "{total} ehdotusta · {pending} odottaa · {accepted} hyväksytty",
|
||||
"title": "Tekoälyn korjausehdotukset",
|
||||
"title": "Automaattisen korjauksen ehdotukset",
|
||||
"undo": "Kumoa tämä ehdotus",
|
||||
"cancel": "Peruuta",
|
||||
"close": "Sulje",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Hitaampi ja kalliimpi, mutta näyttää, mistä ehdotuksista useampi malli on samaa mieltä",
|
||||
"compareLabel": "Vertaile useita tekoälymalleja",
|
||||
"conflictBadge": "vaihtoehto",
|
||||
"conflictTooltip": "Mallit ovat eri mieltä näistä sanoista — yhden vaihtoehdon hyväksyminen hylkää muut.",
|
||||
"conflictTooltip": "Näille sanoille on eri ehdotuksia — yhden vaihtoehdon hyväksyminen hylkää muut.",
|
||||
"consensusBadge": "{consensus}/{total} mallia",
|
||||
"consensusTooltip": "Ehdottanut: {models}",
|
||||
"cachedBadge": "välimuistissa",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Tout accepter",
|
||||
"revertAll": "Tout annuler",
|
||||
"undoAll": "Tout défaire",
|
||||
"resetAllChanges": "Réinitialiser toutes les modifications aux paroles originales corrigées par l'IA",
|
||||
"resetAllChanges": "Réinitialiser toutes les modifications pour revenir aux paroles corrigées d'origine",
|
||||
"undo": "Annuler",
|
||||
"redo": "Rétablir",
|
||||
"findReplace": "Chercher/Remplacer",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Impossible de charger l'audio",
|
||||
"audioStillLoading": "L'audio est toujours en cours de chargement...",
|
||||
"loadingAudio": "Chargement de l'audio...",
|
||||
"autoCorrect": "Suggestions de l'IA",
|
||||
"autoCorrectDesc": "Génère des suggestions de correction de l'IA en comparant la transcription aux paroles de référence. Rien n'est modifié tant que tu n'acceptes pas chaque suggestion."
|
||||
"autoCorrect": "Correction automatique",
|
||||
"autoCorrectDesc": "Génère des suggestions de correction en comparant la transcription aux paroles de référence. Rien n'est modifié tant que tu n'as pas accepté chaque suggestion."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Afficher les conseils",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "autre",
|
||||
"category_repeated_chorus_fix": "refrain",
|
||||
"collapse": "Réduire les suggestions",
|
||||
"confidenceHint": "Le niveau de certitude requis par l'IA pour afficher une suggestion",
|
||||
"confidenceHint": "Le niveau de certitude requis pour qu'une suggestion soit affichée",
|
||||
"confidenceLabel": "Filtre de suggestions",
|
||||
"confidence_all": "Tout afficher",
|
||||
"confidence_high": "Certitude élevée",
|
||||
"description": "L'IA compare les paroles transcrites aux sources de référence et suggère des corrections mot par mot. Tu examines chaque suggestion — rien n'est appliqué tant que tu ne l'as pas acceptée.",
|
||||
"description": "La correction automatique compare les paroles transcrites aux sources de référence et suggère des corrections mot par mot. Tu examines chaque suggestion — rien n'est appliqué tant que tu ne l'as pas acceptée.",
|
||||
"dismiss": "Ignorer les suggestions",
|
||||
"expand": "Développer les suggestions",
|
||||
"insertMarker": "insérer :",
|
||||
"insertionsHint": "Autoriser les suggestions qui insèrent les mots manquants dans la transcription",
|
||||
"insertionsLabel": "Suggérer les mots manquants",
|
||||
"noReferences": "Il n'y a pas de paroles de référence pour cette chanson, les suggestions de l'IA ne peuvent donc pas être générées. Ajoute d'abord des paroles depuis le panneau de référence.",
|
||||
"noSuggestions": "Aucun problème trouvé — l'IA n'a aucune modification à suggérer.",
|
||||
"noReferences": "Il n'y a pas de paroles de référence disponibles pour cette chanson, les suggestions ne peuvent donc pas être générées. Ajoute d'abord des paroles depuis le panneau de référence.",
|
||||
"noSuggestions": "Aucun problème trouvé — il n'y a aucune modification suggérée.",
|
||||
"reject": "Rejeter la suggestion",
|
||||
"rejectAll": "Tout rejeter",
|
||||
"rejected": "Rejeté",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analyse des paroles… cela prend généralement 15 à 60 secondes.",
|
||||
"staleHint": "Ces mots ont été modifiés depuis la génération des suggestions, celle-ci ne s'applique donc plus.",
|
||||
"summary": "{total} suggestions · {pending} en attente · {accepted} acceptées",
|
||||
"title": "Suggestions de correction de l'IA",
|
||||
"title": "Suggestions de correction automatique",
|
||||
"undo": "Annuler cette suggestion",
|
||||
"cancel": "Annuler",
|
||||
"close": "Fermer",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Plus lent et plus cher, mais montre les suggestions sur lesquelles plusieurs modèles s'accordent",
|
||||
"compareLabel": "Comparer plusieurs modèles d'IA",
|
||||
"conflictBadge": "variante",
|
||||
"conflictTooltip": "Les modèles ne sont pas d'accord sur ces mots — accepter une variante rejette les autres.",
|
||||
"conflictTooltip": "Plusieurs suggestions sont proposées pour ces mots — en accepter une rejette les autres.",
|
||||
"consensusBadge": "{consensus}/{total} modèles",
|
||||
"consensusTooltip": "Proposé par : {models}",
|
||||
"cachedBadge": "en cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "אשר הכל",
|
||||
"revertAll": "החזר הכל",
|
||||
"undoAll": "בטל הכל",
|
||||
"resetAllChanges": "אפס את כל השינויים חזרה למילים המקוריות שתוקנו על ידי AI",
|
||||
"resetAllChanges": "איפוס כל השינויים בחזרה למילים המתוקנות המקוריות",
|
||||
"undo": "בטל",
|
||||
"redo": "בצע שוב",
|
||||
"findReplace": "חפש/החלף",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "לא הצלחנו לטעון את האודיו",
|
||||
"audioStillLoading": "האודיו עדיין נטען...",
|
||||
"loadingAudio": "טוען אודיו...",
|
||||
"autoCorrect": "הצעות AI",
|
||||
"autoCorrectDesc": "יצירת הצעות תיקון של AI על ידי השוואת התמלול למילות המקור. שום דבר לא ישתנה עד לאישור של כל הצעה."
|
||||
"autoCorrect": "תיקון אוטומטי",
|
||||
"autoCorrectDesc": "הפקת הצעות לתיקון על ידי השוואת התמלול למילות הייחוס. שום שינוי לא יחול עד לאישור כל הצעה."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "הצג טיפים",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "אחר",
|
||||
"category_repeated_chorus_fix": "פזמון",
|
||||
"collapse": "צמצם הצעות",
|
||||
"confidenceHint": "רמת הביטחון הנדרשת מה-AI כדי להציג הצעה",
|
||||
"confidenceHint": "רמת הוודאות הנדרשת לפני שהצעה תוצג",
|
||||
"confidenceLabel": "מסנן הצעות",
|
||||
"confidence_all": "הצג הכל",
|
||||
"confidence_high": "ביטחון גבוה",
|
||||
"description": "ה-AI משווה את המילים שתומללו למקורות ומציע תיקונים ברמת המילה. באפשרותך לעבור על כל הצעה — שום שינוי לא יבוצע עד לאישור.",
|
||||
"description": "התיקון האוטומטי משווה את המילים שתומללו למקורות הייחוס ומציע תיקונים ברמת המילה. באפשרותך לעבור על כל הצעה — שום שינוי לא יחול עד לאישורו.",
|
||||
"dismiss": "הסתר הצעות",
|
||||
"expand": "הרחב הצעות",
|
||||
"insertMarker": "הוסף:",
|
||||
"insertionsHint": "אפשר הצעות שמוסיפות מילים שהתפספסו בתמלול",
|
||||
"insertionsLabel": "הצע מילים חסרות",
|
||||
"noReferences": "אין מילות מקור זמינות לשיר זה, ולכן לא ניתן ליצור הצעות AI. יש להוסיף מילים מחלונית המקורות תחילה.",
|
||||
"noSuggestions": "לא נמצאו בעיות — ל-AI אין שינויים להציע.",
|
||||
"noReferences": "אין מילות ייחוס זמינות לשיר זה, ולכן לא ניתן להפיק הצעות. יש להוסיף תחילה מילים מחלונית הייחוס.",
|
||||
"noSuggestions": "לא נמצאו בעיות — אין שינויים מוצעים.",
|
||||
"reject": "דחה הצעה",
|
||||
"rejectAll": "דחה הכל",
|
||||
"rejected": "נדחה",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "מנתח מילים… פעולה זו אורכת בדרך כלל 15–60 שניות.",
|
||||
"staleHint": "מילים אלו נערכו מאז שההצעות נוצרו, ולכן הצעה זו כבר אינה רלוונטית.",
|
||||
"summary": "{total} הצעות · {pending} ממתינות · {accepted} התקבלו",
|
||||
"title": "הצעות תיקון של AI",
|
||||
"title": "הצעות לתיקון אוטומטי",
|
||||
"undo": "בטל הצעה זו",
|
||||
"cancel": "ביטול",
|
||||
"close": "סגירה",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "איטי ויקר יותר, אבל מראה על אילו הצעות מספר מודלים מסכימים",
|
||||
"compareLabel": "השוואה בין מספר מודלי AI",
|
||||
"conflictBadge": "גרסה",
|
||||
"conflictTooltip": "המודלים לא מסכימים לגבי המילים האלו — אישור גרסה אחת ידחה את האחרות.",
|
||||
"conflictTooltip": "ישנן הצעות שונות למילים אלו — אישור גרסה אחת ידחה את האחרות.",
|
||||
"consensusBadge": "{consensus}/{total} מודלים",
|
||||
"consensusTooltip": "הוצע על ידי: {models}",
|
||||
"cachedBadge": "שמור במטמון",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "सभी एक्सेप्ट करो",
|
||||
"revertAll": "सभी रिवर्ट करो",
|
||||
"undoAll": "सभी अनडू करो",
|
||||
"resetAllChanges": "सभी बदलावों को ओरिजिनल AI-करेक्टेड लिरिक्स पर रीसेट करो",
|
||||
"resetAllChanges": "सभी बदलावों को वापस मूल सही किए गए लिरिक्स पर रीसेट करो",
|
||||
"undo": "अनडू",
|
||||
"redo": "रीडू",
|
||||
"findReplace": "ढूंढो/बदलो",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "ऑडियो लोड नहीं हो सका",
|
||||
"audioStillLoading": "ऑडियो अभी भी लोड हो रहा है...",
|
||||
"loadingAudio": "ऑडियो लोड हो रहा है...",
|
||||
"autoCorrect": "AI सुझाव",
|
||||
"autoCorrectDesc": "ट्रांसक्रिप्शन की तुलना रेफरेंस लिरिक्स से करके AI सुधार के सुझाव पाओ। जब तक तुम हर सुझाव को स्वीकार नहीं करते, तब तक कुछ भी नहीं बदलता।"
|
||||
"autoCorrect": "ऑटो-करेक्ट",
|
||||
"autoCorrectDesc": "ट्रांसक्रिप्शन की तुलना रेफरेंस लिरिक्स से करके सुधार के सुझाव जनरेट करो। जब तक तुम हर सुझाव को स्वीकार नहीं करते, तब तक कुछ नहीं बदलता।"
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "टिप्स दिखाओ",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "अन्य",
|
||||
"category_repeated_chorus_fix": "कोरस",
|
||||
"collapse": "सुझाव छिपाओ",
|
||||
"confidenceHint": "सुझाव दिखाने के लिए AI को कितना आश्वस्त होना चाहिए",
|
||||
"confidenceHint": "दिखाए जाने से पहले किसी सुझाव का कितना सटीक होना ज़रूरी है",
|
||||
"confidenceLabel": "सुझाव फ़िल्टर",
|
||||
"confidence_all": "सभी दिखाओ",
|
||||
"confidence_high": "ज़्यादा आश्वस्त",
|
||||
"description": "AI ट्रांसक्राइब किए गए लिरिक्स की तुलना रेफरेंस से करता है और शब्दों में सुधार के सुझाव देता है। तुम हर सुझाव की समीक्षा कर सकते हो — जब तक तुम उसे स्वीकार नहीं करते, तब तक कुछ भी लागू नहीं होता।",
|
||||
"description": "ऑटो-करेक्ट ट्रांसक्राइब किए गए लिरिक्स की तुलना रेफरेंस सोर्स से करता है और शब्दों को ठीक करने के सुझाव देता है। तुम हर सुझाव का रिव्यू कर सकते हो — जब तक तुम इसे स्वीकार नहीं करते, तब तक कुछ भी लागू नहीं होता।",
|
||||
"dismiss": "सुझाव हटाओ",
|
||||
"expand": "सुझाव पूरे दिखाओ",
|
||||
"insertMarker": "जोड़ो:",
|
||||
"insertionsHint": "उन सुझावों को अनुमति दो जो ट्रांसक्रिप्शन में छूटे हुए शब्दों को जोड़ते हैं",
|
||||
"insertionsLabel": "छूटे हुए शब्दों का सुझाव दो",
|
||||
"noReferences": "इस गाने के लिए कोई रेफरेंस लिरिक्स उपलब्ध नहीं हैं, इसलिए AI सुझाव नहीं दे सकता। पहले रेफरेंस पैनल से लिरिक्स जोड़ो।",
|
||||
"noSuggestions": "कोई समस्या नहीं मिली — AI के पास बदलाव के लिए कोई सुझाव नहीं है।",
|
||||
"noReferences": "इस गाने के लिए कोई रेफरेंस लिरिक्स उपलब्ध नहीं हैं, इसलिए सुझाव जनरेट नहीं किए जा सकते। पहले रेफरेंस पैनल से लिरिक्स जोड़ो।",
|
||||
"noSuggestions": "कोई समस्या नहीं मिली — बदलाव के लिए कोई सुझाव नहीं हैं।",
|
||||
"reject": "सुझाव अस्वीकार करो",
|
||||
"rejectAll": "सभी अस्वीकार करो",
|
||||
"rejected": "अस्वीकार किया गया",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "लिरिक्स का विश्लेषण हो रहा है… इसमें आमतौर पर 15–60 सेकंड लगते हैं।",
|
||||
"staleHint": "सुझाव आने के बाद इन शब्दों को एडिट किया गया था, इसलिए यह सुझाव अब लागू नहीं होता।",
|
||||
"summary": "{total} सुझाव · {pending} लंबित · {accepted} स्वीकार किए गए",
|
||||
"title": "AI सुधार सुझाव",
|
||||
"title": "ऑटो-करेक्ट सुझाव",
|
||||
"undo": "इस सुझाव को पूर्ववत करो",
|
||||
"cancel": "कैंसल",
|
||||
"close": "बंद करो",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "यह धीमा है और इसमें ज़्यादा खर्च आता है, लेकिन यह दिखाता है कि किन सुझावों पर कई मॉडल सहमत हैं",
|
||||
"compareLabel": "कई AI मॉडलों की तुलना करो",
|
||||
"conflictBadge": "विकल्प",
|
||||
"conflictTooltip": "इन शब्दों पर मॉडल असहमत हैं — एक विकल्प को स्वीकार करने से बाकी अस्वीकार हो जाते हैं।",
|
||||
"conflictTooltip": "इन शब्दों के लिए अलग-अलग सुझाव हैं — एक विकल्प को स्वीकार करने पर बाकी अस्वीकार हो जाएंगे।",
|
||||
"consensusBadge": "{consensus}/{total} मॉडल",
|
||||
"consensusTooltip": "इनके द्वारा प्रस्तावित: {models}",
|
||||
"cachedBadge": "कैश्ड",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Prihvati sve",
|
||||
"revertAll": "Vrati sve",
|
||||
"undoAll": "Poništi sve",
|
||||
"resetAllChanges": "Poništi sve promjene natrag na originalni AI-ispravljeni tekst",
|
||||
"resetAllChanges": "Poništi sve promjene i vrati na izvorni ispravljeni tekst",
|
||||
"undo": "Poništi",
|
||||
"redo": "Ponovi",
|
||||
"findReplace": "Pronađi/Zamijeni",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Nije moguće učitati audio",
|
||||
"audioStillLoading": "Audio se još uvijek učitava...",
|
||||
"loadingAudio": "Učitavanje audija...",
|
||||
"autoCorrect": "AI prijedlog",
|
||||
"autoCorrectDesc": "Generiraj AI prijedloge za ispravak uspoređujući transkripciju s referentnim tekstom. Ništa se ne mijenja dok ne prihvatiš svaki prijedlog."
|
||||
"autoCorrect": "Automatski ispravak",
|
||||
"autoCorrectDesc": "Generiraj prijedloge za ispravak usporedbom transkripcije s referentnim tekstom. Ništa se ne mijenja dok ne prihvatiš svaki prijedlog."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Prikaži savjete",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "ostalo",
|
||||
"category_repeated_chorus_fix": "refren",
|
||||
"collapse": "Sažmi prijedloge",
|
||||
"confidenceHint": "Koliko AI mora biti siguran da bi prikazao prijedlog",
|
||||
"confidenceHint": "Koliko prijedlog mora biti pouzdan prije nego što se prikaže",
|
||||
"confidenceLabel": "Filtar prijedloga",
|
||||
"confidence_all": "Prikaži sve",
|
||||
"confidence_high": "Visoka pouzdanost",
|
||||
"description": "AI uspoređuje transkribirani tekst s referentnim izvorima i predlaže ispravke na razini riječi. Pregledavaš svaki prijedlog — ništa se ne primjenjuje dok ga ne prihvatiš.",
|
||||
"description": "Automatski ispravak uspoređuje transkribirani tekst s referentnim izvorima i predlaže ispravke na razini riječi. Pregledavaš svaki prijedlog — ništa se ne primjenjuje dok ga ne prihvatiš.",
|
||||
"dismiss": "Odbaci prijedloge",
|
||||
"expand": "Proširi prijedloge",
|
||||
"insertMarker": "umetni:",
|
||||
"insertionsHint": "Dopusti prijedloge za umetanje riječi koje nedostaju u transkripciji",
|
||||
"insertionsLabel": "Predloži riječi koje nedostaju",
|
||||
"noReferences": "Za ovu pjesmu nema referentnog teksta, pa se AI prijedlozi ne mogu generirati. Prvo dodaj tekst iz referentnog panela.",
|
||||
"noSuggestions": "Nisu pronađeni problemi — AI nema predloženih promjena.",
|
||||
"noReferences": "Za ovu pjesmu nema dostupnog referentnog teksta, pa se prijedlozi ne mogu generirati. Prvo dodaj tekst iz referentnog panela.",
|
||||
"noSuggestions": "Nisu pronađeni problemi — nema predloženih promjena.",
|
||||
"reject": "Odbij prijedlog",
|
||||
"rejectAll": "Odbij sve",
|
||||
"rejected": "Odbijeno",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analiziram tekst… ovo obično traje 15–60 sekundi.",
|
||||
"staleHint": "Ove su riječi uređene nakon generiranja prijedloga, pa ovaj prijedlog više ne vrijedi.",
|
||||
"summary": "{total} prijedloga · {pending} na čekanju · {accepted} prihvaćeno",
|
||||
"title": "AI prijedlozi za ispravak",
|
||||
"title": "Prijedlozi automatskog ispravka",
|
||||
"undo": "Poništi ovaj prijedlog",
|
||||
"cancel": "Odustani",
|
||||
"close": "Zatvori",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Sporije je i košta više, ali pokazuje oko kojih se prijedloga slaže više modela",
|
||||
"compareLabel": "Usporedi više AI modela",
|
||||
"conflictBadge": "varijanta",
|
||||
"conflictTooltip": "Modeli se ne slažu oko ovih riječi — prihvaćanje jedne varijante odbacuje ostale.",
|
||||
"conflictTooltip": "Postoje različiti prijedlozi za ove riječi — prihvaćanjem jedne varijante odbacuješ ostale.",
|
||||
"consensusBadge": "{consensus}/{total} modela",
|
||||
"consensusTooltip": "Predložili: {models}",
|
||||
"cachedBadge": "predmemorirano",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Összes elfogadása",
|
||||
"revertAll": "Összes visszaállítása",
|
||||
"undoAll": "Összes visszavonása",
|
||||
"resetAllChanges": "Minden változtatás visszaállítása az eredeti AI által javított dalszövegre",
|
||||
"resetAllChanges": "Minden módosítás visszaállítása az eredeti, javított dalszövegre",
|
||||
"undo": "Visszavonás",
|
||||
"redo": "Újra",
|
||||
"findReplace": "Keresés/Csere",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Nem sikerült betölteni a hanganyagot",
|
||||
"audioStillLoading": "A hanganyag még töltődik...",
|
||||
"loadingAudio": "Hanganyag betöltése...",
|
||||
"autoCorrect": "MI-javaslat",
|
||||
"autoCorrectDesc": "Generálj MI-javítási javaslatokat az átírás és a referenciadalszöveg összehasonlításával. Semmi sem változik, amíg el nem fogadod az egyes javaslatokat."
|
||||
"autoCorrect": "Automatikus javítás",
|
||||
"autoCorrectDesc": "Javítási javaslatok generálása a leírt és a referencia-dalszöveg összehasonlításával. Semmi sem módosul, amíg el nem fogadod az egyes javaslatokat."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Tippek mutatása",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "egyéb",
|
||||
"category_repeated_chorus_fix": "refrén",
|
||||
"collapse": "Javaslatok összecsukása",
|
||||
"confidenceHint": "Milyen biztosnak kell lennie az MI-nek ahhoz, hogy megjelenítsen egy javaslatot",
|
||||
"confidenceHint": "Mennyire legyen biztos a javaslat, mielőtt megjelenik",
|
||||
"confidenceLabel": "Javaslatszűrő",
|
||||
"confidence_all": "Összes megjelenítése",
|
||||
"confidence_high": "Nagy bizonyosság",
|
||||
"description": "Az MI összehasonlítja az átírt dalszöveget a referenciaforrásokkal, és szószintű javításokat javasol. Minden javaslatot átnézhetsz — semmi sem módosul, amíg el nem fogadod.",
|
||||
"description": "Az automatikus javítás összehasonlítja a leírt dalszöveget a referenciaforrásokkal, és szavankénti javításokat javasol. Minden javaslatot átnézhetsz — semmi sem módosul, amíg el nem fogadod.",
|
||||
"dismiss": "Javaslatok elvetése",
|
||||
"expand": "Javaslatok kibontása",
|
||||
"insertMarker": "beszúrás:",
|
||||
"insertionsHint": "Engedélyezd az olyan javaslatokat, amelyek beszúrják az átírásból kimaradt szavakat",
|
||||
"insertionsLabel": "Hiányzó szavak javaslata",
|
||||
"noReferences": "Ehhez a dalhoz nem érhető el referenciadalszöveg, így az MI nem tud javaslatokat tenni. Először adj hozzá dalszöveget a referenciapanelen.",
|
||||
"noSuggestions": "Nem találtunk problémát — az MI-nek nincsenek módosítási javaslatai.",
|
||||
"noReferences": "Ehhez a dalhoz nincs elérhető referencia-dalszöveg, így nem tudunk javaslatokat generálni. Először adj hozzá dalszöveget a referenciapanelen.",
|
||||
"noSuggestions": "Nem találtunk problémát — nincsenek javasolt módosítások.",
|
||||
"reject": "Javaslat elutasítása",
|
||||
"rejectAll": "Összes elutasítása",
|
||||
"rejected": "Elutasítva",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Dalszöveg elemzése… ez általában 15–60 másodpercet vesz igénybe.",
|
||||
"staleHint": "Ezeket a szavakat módosítottad a javaslatok generálása óta, így ez a javaslat már nem érvényes.",
|
||||
"summary": "{total} javaslat · {pending} függőben · {accepted} elfogadva",
|
||||
"title": "MI-javítási javaslatok",
|
||||
"title": "Automatikus javítási javaslatok",
|
||||
"undo": "Javaslat visszavonása",
|
||||
"cancel": "Mégse",
|
||||
"close": "Bezárás",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Lassabb és többe kerül, de megmutatja, mely javaslatokban ért egyet több modell.",
|
||||
"compareLabel": "Több MI-modell összehasonlítása",
|
||||
"conflictBadge": "változat",
|
||||
"conflictTooltip": "A modellek nem értenek egyet ezeknél a szavaknál — egy változat elfogadásával elutasítod a többit.",
|
||||
"conflictTooltip": "Több különböző javaslat is van ezekre a szavakra — az egyik változat elfogadásával elutasítod a többit.",
|
||||
"consensusBadge": "{consensus}/{total} modell",
|
||||
"consensusTooltip": "Javasolta: {models}",
|
||||
"cachedBadge": "gyorsítótárazott",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Terima Semua",
|
||||
"revertAll": "Kembalikan Semua",
|
||||
"undoAll": "Batalkan Semua",
|
||||
"resetAllChanges": "Atur ulang semua perubahan kembali ke lirik asli yang dikoreksi AI",
|
||||
"resetAllChanges": "Kembalikan semua perubahan ke lirik asli yang telah dikoreksi",
|
||||
"undo": "Urungkan",
|
||||
"redo": "Ulangi",
|
||||
"findReplace": "Cari/Ganti",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Gagal memuat audio",
|
||||
"audioStillLoading": "Audio masih dimuat...",
|
||||
"loadingAudio": "Memuat audio...",
|
||||
"autoCorrect": "Saran AI",
|
||||
"autoCorrectDesc": "Buat saran koreksi AI dengan membandingkan transkripsi dengan lirik referensi. Tidak ada perubahan yang diterapkan sampai Anda menerima setiap saran."
|
||||
"autoCorrect": "Koreksi otomatis",
|
||||
"autoCorrectDesc": "Hasilkan saran koreksi dengan membandingkan transkripsi dengan lirik referensi. Tidak ada perubahan yang diterapkan sampai Anda menyetujui setiap saran."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Tampilkan tips",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "lainnya",
|
||||
"category_repeated_chorus_fix": "korus",
|
||||
"collapse": "Perkecil saran",
|
||||
"confidenceHint": "Tingkat keyakinan AI yang diperlukan untuk menampilkan saran",
|
||||
"confidenceHint": "Tingkat keyakinan minimum sebelum saran ditampilkan",
|
||||
"confidenceLabel": "Filter saran",
|
||||
"confidence_all": "Tampilkan semua",
|
||||
"confidence_high": "Keyakinan tinggi",
|
||||
"description": "AI membandingkan lirik hasil transkripsi dengan sumber referensi dan menyarankan perbaikan per kata. Anda dapat meninjau setiap saran — tidak ada perubahan yang diterapkan sampai Anda menerimanya.",
|
||||
"description": "Koreksi otomatis membandingkan lirik yang ditranskripsi dengan sumber referensi dan menyarankan perbaikan per kata. Anda dapat meninjau setiap saran — tidak ada perubahan yang diterapkan sampai Anda menyetujuinya.",
|
||||
"dismiss": "Abaikan saran",
|
||||
"expand": "Perluas saran",
|
||||
"insertMarker": "sisipkan:",
|
||||
"insertionsHint": "Izinkan saran untuk menyisipkan kata yang terlewat dalam transkripsi",
|
||||
"insertionsLabel": "Sarankan kata yang hilang",
|
||||
"noReferences": "Tidak ada lirik referensi yang tersedia untuk lagu ini, sehingga saran AI tidak dapat dibuat. Tambahkan lirik dari panel referensi terlebih dahulu.",
|
||||
"noSuggestions": "Tidak ada masalah yang ditemukan — AI tidak memiliki saran perubahan.",
|
||||
"noReferences": "Tidak ada lirik referensi yang tersedia untuk lagu ini, sehingga saran tidak dapat dibuat. Tambahkan lirik dari panel referensi terlebih dahulu.",
|
||||
"noSuggestions": "Tidak ada masalah yang ditemukan — tidak ada perubahan yang disarankan.",
|
||||
"reject": "Tolak saran",
|
||||
"rejectAll": "Tolak semua",
|
||||
"rejected": "Ditolak",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Menganalisis lirik… biasanya memakan waktu 15–60 detik.",
|
||||
"staleHint": "Kata-kata ini telah diedit sejak saran dibuat, sehingga saran ini tidak lagi berlaku.",
|
||||
"summary": "{total} saran · {pending} tertunda · {accepted} diterima",
|
||||
"title": "Saran koreksi AI",
|
||||
"title": "Saran koreksi otomatis",
|
||||
"undo": "Batalkan saran ini",
|
||||
"cancel": "Batal",
|
||||
"close": "Tutup",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Lebih lambat dan lebih mahal, tapi menunjukkan saran mana yang disetujui oleh beberapa model",
|
||||
"compareLabel": "Bandingkan beberapa model AI",
|
||||
"conflictBadge": "varian",
|
||||
"conflictTooltip": "Model memberikan hasil berbeda untuk kata-kata ini — menerima satu varian akan menolak varian lainnya.",
|
||||
"conflictTooltip": "Ada beberapa saran berbeda untuk kata-kata ini — menerima satu varian akan mengabaikan varian lainnya.",
|
||||
"consensusBadge": "{consensus}/{total} model",
|
||||
"consensusTooltip": "Diusulkan oleh: {models}",
|
||||
"cachedBadge": "di-cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Accetta tutto",
|
||||
"revertAll": "Ripristina tutto",
|
||||
"undoAll": "Annulla tutto",
|
||||
"resetAllChanges": "Reimposta tutte le modifiche al testo originale corretto dall'IA",
|
||||
"resetAllChanges": "Annulla tutte le modifiche e torna al testo corretto originale",
|
||||
"undo": "Annulla",
|
||||
"redo": "Ripeti",
|
||||
"findReplace": "Trova/Sostituisci",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Impossibile caricare l'audio",
|
||||
"audioStillLoading": "L'audio è ancora in caricamento...",
|
||||
"loadingAudio": "Caricamento audio...",
|
||||
"autoCorrect": "Suggerimenti IA",
|
||||
"autoCorrectDesc": "Genera suggerimenti di correzione dell'IA confrontando la trascrizione con il testo di riferimento. Non verrà modificato nulla finché non accetterai i singoli suggerimenti."
|
||||
"autoCorrect": "Correzione automatica",
|
||||
"autoCorrectDesc": "Genera suggerimenti di correzione confrontando la trascrizione con il testo di riferimento. Non viene modificato nulla finché non accetti ogni suggerimento."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Mostra suggerimenti",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "altro",
|
||||
"category_repeated_chorus_fix": "ritornello",
|
||||
"collapse": "Comprimi suggerimenti",
|
||||
"confidenceHint": "Quanto deve essere sicura l'IA prima di mostrare un suggerimento",
|
||||
"confidenceHint": "Quanto deve essere attendibile un suggerimento prima di essere mostrato",
|
||||
"confidenceLabel": "Filtro suggerimenti",
|
||||
"confidence_all": "Mostra tutti",
|
||||
"confidence_high": "Alta sicurezza",
|
||||
"description": "L'IA confronta il testo trascritto con le fonti di riferimento e suggerisce correzioni parola per parola. Sei tu a controllare ogni suggerimento: non verrà applicato nulla finché non lo accetti.",
|
||||
"description": "La correzione automatica confronta il testo trascritto con le fonti di riferimento e suggerisce correzioni a livello di singola parola. Puoi rivedere ogni suggerimento — non verrà applicato nulla finché non lo accetti.",
|
||||
"dismiss": "Ignora suggerimenti",
|
||||
"expand": "Espandi suggerimenti",
|
||||
"insertMarker": "inserisci:",
|
||||
"insertionsHint": "Consenti suggerimenti che inseriscono le parole omesse nella trascrizione",
|
||||
"insertionsLabel": "Suggerisci parole mancanti",
|
||||
"noReferences": "Non sono disponibili testi di riferimento per questa canzone, quindi non è possibile generare i suggerimenti dell'IA. Aggiungi prima un testo dal pannello dei riferimenti.",
|
||||
"noSuggestions": "Nessun problema trovato: l'IA non ha modifiche da suggerire.",
|
||||
"noReferences": "Non ci sono testi di riferimento disponibili per questa canzone, quindi non è possibile generare suggerimenti. Aggiungi prima il testo dal pannello di riferimento.",
|
||||
"noSuggestions": "Nessun problema trovato — non ci sono modifiche suggerite.",
|
||||
"reject": "Rifiuta suggerimento",
|
||||
"rejectAll": "Rifiuta tutti",
|
||||
"rejected": "Rifiutato",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analisi del testo in corso… di solito richiede dai 15 ai 60 secondi.",
|
||||
"staleHint": "Queste parole sono state modificate dopo la generazione dei suggerimenti, quindi questo suggerimento non è più valido.",
|
||||
"summary": "{total} suggerimenti · {pending} in attesa · {accepted} accettati",
|
||||
"title": "Suggerimenti di correzione dell'IA",
|
||||
"title": "Suggerimenti di correzione automatica",
|
||||
"undo": "Annulla questo suggerimento",
|
||||
"cancel": "Annulla",
|
||||
"close": "Chiudi",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Più lento e costoso, ma mostra su quali suggerimenti concordano più modelli",
|
||||
"compareLabel": "Confronta più modelli IA",
|
||||
"conflictBadge": "variante",
|
||||
"conflictTooltip": "I modelli non concordano su queste parole — accettare una variante esclude le altre.",
|
||||
"conflictTooltip": "Ci sono diversi suggerimenti per queste parole — accettando una variante rifiuterai le altre.",
|
||||
"consensusBadge": "{consensus}/{total} modelli",
|
||||
"consensusTooltip": "Proposto da: {models}",
|
||||
"cachedBadge": "in cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "すべて承認",
|
||||
"revertAll": "すべて元に戻す",
|
||||
"undoAll": "すべて取り消す",
|
||||
"resetAllChanges": "すべての変更を元のAI修正済み歌詞にリセット",
|
||||
"resetAllChanges": "すべての変更をリセットして、元の修正済み歌詞に戻す",
|
||||
"undo": "元に戻す",
|
||||
"redo": "やり直し",
|
||||
"findReplace": "検索/置換",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "音声を読み込めませんでした",
|
||||
"audioStillLoading": "音声はまだ読み込み中です...",
|
||||
"loadingAudio": "音声を読み込み中...",
|
||||
"autoCorrect": "AI提案",
|
||||
"autoCorrectDesc": "文字起こしとリファレンス歌詞を比較して、AIが修正案を生成します。各提案を採用するまで、実際の歌詞が変更されることはありません。"
|
||||
"autoCorrect": "自動補正",
|
||||
"autoCorrectDesc": "文字起こしした歌詞と参照用の歌詞を比較して、修正の提案を生成します。各提案を承認するまで変更は適用されません。"
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "ヒントを表示",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "その他",
|
||||
"category_repeated_chorus_fix": "コーラス",
|
||||
"collapse": "提案を折りたたむ",
|
||||
"confidenceHint": "AIがどの程度の確信度で提案を表示するかの基準です。",
|
||||
"confidenceHint": "提案を表示するために必要な信頼度の基準",
|
||||
"confidenceLabel": "提案フィルター",
|
||||
"confidence_all": "すべて表示",
|
||||
"confidence_high": "高い確信度",
|
||||
"description": "AIが文字起こしされた歌詞とリファレンスを比較し、単語レベルで修正を提案します。すべての提案はご自身で確認でき、採用するまで実際の歌詞には反映されません。",
|
||||
"description": "自動補正は、文字起こしした歌詞を参照用の歌詞と比較し、単語レベルで修正を提案します。すべての提案はご自身で確認でき、承認するまで適用されることはありません。",
|
||||
"dismiss": "提案を非表示にする",
|
||||
"expand": "提案を展開する",
|
||||
"insertMarker": "挿入:",
|
||||
"insertionsHint": "文字起こしで抜けている単語を追加する提案を表示します。",
|
||||
"insertionsLabel": "不足している単語の追加を提案",
|
||||
"noReferences": "この曲にはリファレンス歌詞がないため、AIの提案を生成できません。まずリファレンスパネルから歌詞を追加してください。",
|
||||
"noSuggestions": "問題は見つかりませんでした。AIからの修正提案はありません。",
|
||||
"noReferences": "この曲には参照用の歌詞がないため、提案を生成できません。まずは参照パネルから歌詞を追加してください。",
|
||||
"noSuggestions": "問題は見つかりませんでした。修正の提案はありません。",
|
||||
"reject": "提案を却下する",
|
||||
"rejectAll": "すべて却下する",
|
||||
"rejected": "却下済み",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "歌詞を分析しています… 通常15〜60秒かかります。",
|
||||
"staleHint": "提案の生成後に歌詞が編集されたため、この提案は無効になりました。",
|
||||
"summary": "{total}件の提案 · 保留中 {pending}件 · 採用済み {accepted}件",
|
||||
"title": "AI修正提案",
|
||||
"title": "自動補正の提案",
|
||||
"undo": "この提案を元に戻す",
|
||||
"cancel": "キャンセル",
|
||||
"close": "閉じる",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "時間とコストはかかりますが、複数のモデルで一致した提案を確認できます。",
|
||||
"compareLabel": "複数のAIモデルを比較する",
|
||||
"conflictBadge": "別案",
|
||||
"conflictTooltip": "モデル間でこれらの単語の認識が異なります。1つの案を採用すると、他の案は破棄されます。",
|
||||
"conflictTooltip": "これらの単語には複数の修正候補があります。1つの候補を承認すると、他の候補は破棄されます。",
|
||||
"consensusBadge": "{consensus}/{total} モデル",
|
||||
"consensusTooltip": "提案元: {models}",
|
||||
"cachedBadge": "キャッシュ済み",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "모두 수락",
|
||||
"revertAll": "모두 되돌리기",
|
||||
"undoAll": "모두 실행 취소",
|
||||
"resetAllChanges": "모든 변경 사항을 원래 AI가 수정한 가사로 초기화",
|
||||
"resetAllChanges": "모든 변경 사항을 취소하고 원래 수정된 가사로 되돌립니다",
|
||||
"undo": "실행 취소",
|
||||
"redo": "다시 실행",
|
||||
"findReplace": "찾기/바꾸기",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "오디오를 불러올 수 없습니다",
|
||||
"audioStillLoading": "아직 오디오를 불러오는 중입니다...",
|
||||
"loadingAudio": "오디오를 불러오는 중입니다...",
|
||||
"autoCorrect": "AI 제안",
|
||||
"autoCorrectDesc": "변환된 가사를 참조 가사와 비교하여 AI 수정 제안을 생성합니다. 각 제안을 수락하기 전까지는 아무것도 변경되지 않습니다."
|
||||
"autoCorrect": "자동 수정",
|
||||
"autoCorrectDesc": "추출된 가사를 참조 가사와 비교하여 수정 제안을 생성합니다. 각 제안을 수락하기 전까지는 아무것도 변경되지 않습니다."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "팁 표시",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "기타",
|
||||
"category_repeated_chorus_fix": "코러스",
|
||||
"collapse": "제안 접기",
|
||||
"confidenceHint": "제안을 표시하기 위한 AI의 확신도 기준",
|
||||
"confidenceHint": "제안을 표시할 신뢰도 수준",
|
||||
"confidenceLabel": "제안 필터",
|
||||
"confidence_all": "모두 보기",
|
||||
"confidence_high": "높은 확신도",
|
||||
"description": "AI가 변환된 가사를 참조 소스와 비교하여 단어 수준의 수정을 제안합니다. 모든 제안을 검토할 수 있으며, 수락하기 전까지는 아무것도 적용되지 않습니다.",
|
||||
"description": "자동 수정은 추출된 가사를 참조 소스와 비교하여 단어 단위로 수정 사항을 제안합니다. 모든 제안은 직접 검토하실 수 있으며, 수락하기 전까지는 아무것도 적용되지 않습니다.",
|
||||
"dismiss": "제안 무시",
|
||||
"expand": "제안 펼치기",
|
||||
"insertMarker": "삽입:",
|
||||
"insertionsHint": "변환 과정에서 누락된 단어를 삽입하는 제안 허용",
|
||||
"insertionsLabel": "누락된 단어 제안",
|
||||
"noReferences": "이 노래에 사용할 수 있는 참조 가사가 없어 AI 제안을 생성할 수 없습니다. 먼저 참조 패널에서 가사를 추가해 주세요.",
|
||||
"noSuggestions": "발견된 문제 없음 — AI가 제안할 변경 사항이 없습니다.",
|
||||
"noReferences": "이 노래에 대한 참조 가사가 없어 제안을 생성할 수 없습니다. 먼저 참조 패널에서 가사를 추가해 주세요.",
|
||||
"noSuggestions": "발견된 문제가 없습니다 — 제안할 변경 사항이 없습니다.",
|
||||
"reject": "제안 거부",
|
||||
"rejectAll": "모두 거부",
|
||||
"rejected": "거부됨",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "가사 분석 중… 보통 15~60초 정도 걸립니다.",
|
||||
"staleHint": "제안이 생성된 후 해당 단어가 수정되어 이 제안은 더 이상 유효하지 않습니다.",
|
||||
"summary": "{total}개 제안 · {pending}개 대기 중 · {accepted}개 수락됨",
|
||||
"title": "AI 수정 제안",
|
||||
"title": "자동 수정 제안",
|
||||
"undo": "이 제안 실행 취소",
|
||||
"cancel": "취소",
|
||||
"close": "닫기",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "시간과 비용이 더 들지만, 여러 모델이 공통으로 제안하는 내용을 확인할 수 있어요.",
|
||||
"compareLabel": "여러 AI 모델 비교하기",
|
||||
"conflictBadge": "다른 버전",
|
||||
"conflictTooltip": "이 단어들에 대해 모델마다 제안이 달라요. 하나의 버전을 수락하면 나머지는 반영되지 않아요.",
|
||||
"conflictTooltip": "이 단어들에 대해 여러 제안이 있습니다 — 하나를 수락하면 나머지는 적용되지 않습니다.",
|
||||
"consensusBadge": "{consensus}/{total}개 모델",
|
||||
"consensusTooltip": "제안한 모델: {models}",
|
||||
"cachedBadge": "캐시됨",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Terima Semua",
|
||||
"revertAll": "Kembalikan Semua",
|
||||
"undoAll": "Buat Asal Semua",
|
||||
"resetAllChanges": "Tetapkan semula semua perubahan kembali kepada lirik asal yang dibetulkan AI",
|
||||
"resetAllChanges": "Tetapkan semula semua perubahan kepada lirik asal yang telah dibetulkan",
|
||||
"undo": "Buat Asal",
|
||||
"redo": "Buat Semula",
|
||||
"findReplace": "Cari/Ganti",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Tidak dapat memuatkan audio",
|
||||
"audioStillLoading": "Audio masih dimuatkan...",
|
||||
"loadingAudio": "Memuatkan audio...",
|
||||
"autoCorrect": "Cadangan AI",
|
||||
"autoCorrectDesc": "Jana cadangan pembetulan AI dengan membandingkan transkripsi dengan lirik rujukan. Tiada perubahan akan dibuat sehingga anda menerima setiap cadangan."
|
||||
"autoCorrect": "Pembetulan automatik",
|
||||
"autoCorrectDesc": "Jana cadangan pembetulan dengan membandingkan transkripsi dengan lirik rujukan. Tiada perubahan akan dibuat sehingga anda menerima setiap cadangan."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Tunjukkan petua",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "lain-lain",
|
||||
"category_repeated_chorus_fix": "korus",
|
||||
"collapse": "Tutup cadangan",
|
||||
"confidenceHint": "Tahap keyakinan AI yang diperlukan untuk menunjukkan cadangan",
|
||||
"confidenceHint": "Tahap keyakinan sesuatu cadangan sebelum ia dipaparkan",
|
||||
"confidenceLabel": "Penapis cadangan",
|
||||
"confidence_all": "Tunjukkan semua",
|
||||
"confidence_high": "Keyakinan tinggi",
|
||||
"description": "AI membandingkan lirik yang ditranskripsi dengan sumber rujukan dan mencadangkan pembetulan pada peringkat perkataan. Anda menyemak setiap cadangan — tiada perubahan akan dibuat sehingga anda menerimanya.",
|
||||
"description": "Pembetulan automatik membandingkan lirik yang ditranskrip dengan sumber rujukan dan mencadangkan pembetulan pada peringkat perkataan. Anda boleh menyemak setiap cadangan — tiada perubahan akan dibuat sehingga anda menerimanya.",
|
||||
"dismiss": "Abaikan cadangan",
|
||||
"expand": "Kembangkan cadangan",
|
||||
"insertMarker": "masukkan:",
|
||||
"insertionsHint": "Benarkan cadangan yang memasukkan perkataan yang tertinggal dalam transkripsi",
|
||||
"insertionsLabel": "Cadangkan perkataan yang tertinggal",
|
||||
"noReferences": "Tiada lirik rujukan tersedia untuk lagu ini, jadi cadangan AI tidak dapat dijana. Tambahkan lirik dari panel rujukan terlebih dahulu.",
|
||||
"noSuggestions": "Tiada isu ditemui — AI tiada cadangan perubahan.",
|
||||
"noReferences": "Tiada lirik rujukan tersedia untuk lagu ini, jadi cadangan tidak dapat dijana. Tambahkan lirik dari panel rujukan terlebih dahulu.",
|
||||
"noSuggestions": "Tiada isu ditemui — tiada perubahan yang dicadangkan.",
|
||||
"reject": "Tolak cadangan",
|
||||
"rejectAll": "Tolak semua",
|
||||
"rejected": "Ditolak",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Menganalisis lirik… ini biasanya mengambil masa 15–60 saat.",
|
||||
"staleHint": "Perkataan ini telah diedit sejak cadangan dijana, jadi cadangan ini tidak lagi terpakai.",
|
||||
"summary": "{total} cadangan · {pending} belum selesai · {accepted} diterima",
|
||||
"title": "Cadangan pembetulan AI",
|
||||
"title": "Cadangan pembetulan automatik",
|
||||
"undo": "Buat asal cadangan ini",
|
||||
"cancel": "Batal",
|
||||
"close": "Tutup",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Lebih perlahan dan kos lebih tinggi, tetapi menunjukkan cadangan yang dipersetujui oleh berbilang model",
|
||||
"compareLabel": "Bandingkan berbilang model AI",
|
||||
"conflictBadge": "varian",
|
||||
"conflictTooltip": "Model tidak bersetuju tentang perkataan ini — menerima satu varian akan menolak varian yang lain.",
|
||||
"conflictTooltip": "Terdapat cadangan yang berbeza untuk perkataan ini — menerima satu varian akan menolak yang lain.",
|
||||
"consensusBadge": "{consensus}/{total} model",
|
||||
"consensusTooltip": "Dicadangkan oleh: {models}",
|
||||
"cachedBadge": "dicache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Godta alle",
|
||||
"revertAll": "Tilbakestill alle",
|
||||
"undoAll": "Angre alle",
|
||||
"resetAllChanges": "Tilbakestill alle endringer til den originale AI-korrigerte teksten",
|
||||
"resetAllChanges": "Tilbakestill alle endringer til den opprinnelige korrigerte teksten",
|
||||
"undo": "Angre",
|
||||
"redo": "Gjør om",
|
||||
"findReplace": "Finn/Erstatt",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Kunne ikke laste inn lyd",
|
||||
"audioStillLoading": "Lyden laster fortsatt inn...",
|
||||
"loadingAudio": "Laster inn lyd...",
|
||||
"autoCorrect": "AI-forslag",
|
||||
"autoCorrectDesc": "Generer AI-korrigeringsforslag ved å sammenligne transkripsjonen med referanseteksten. Ingenting endres før du godtar hvert forslag."
|
||||
"autoCorrect": "Autokorrektur",
|
||||
"autoCorrectDesc": "Generer forslag til rettelser ved å sammenligne transkripsjonen med referanseteksten. Ingenting endres før du godtar hvert forslag."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Vis tips",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "annet",
|
||||
"category_repeated_chorus_fix": "refreng",
|
||||
"collapse": "Skjul forslag",
|
||||
"confidenceHint": "Hvor sikker AI-en må være for å vise et forslag",
|
||||
"confidenceHint": "Hvor sikkert et forslag må være før det vises",
|
||||
"confidenceLabel": "Forslagsfilter",
|
||||
"confidence_all": "Vis alle",
|
||||
"confidence_high": "Høy sikkerhet",
|
||||
"description": "AI-en sammenligner den transkriberte teksten med referansekildene og foreslår rettelser på ordnivå. Du går gjennom hvert forslag — ingenting blir endret før du godtar det.",
|
||||
"description": "Autokorrektur sammenligner den transkriberte teksten med referansekildene og foreslår rettelser på ordnivå. Du går gjennom hvert forslag — ingenting tas i bruk før du godtar det.",
|
||||
"dismiss": "Lukk forslag",
|
||||
"expand": "Utvid forslag",
|
||||
"insertMarker": "sett inn:",
|
||||
"insertionsHint": "Tillat forslag som setter inn ord som mangler i transkripsjonen",
|
||||
"insertionsLabel": "Foreslå manglende ord",
|
||||
"noReferences": "Ingen referansetekster er tilgjengelige for denne sangen, så AI-forslag kan ikke genereres. Legg til tekst fra referansepanelet først.",
|
||||
"noSuggestions": "Ingen problemer funnet — AI-en har ingen foreslåtte endringer.",
|
||||
"noReferences": "Ingen referansetekst er tilgjengelig for denne sangen, så det kan ikke genereres forslag. Legg til tekst fra referansepanelet først.",
|
||||
"noSuggestions": "Ingen problemer funnet — det er ingen foreslåtte endringer.",
|
||||
"reject": "Avvis forslag",
|
||||
"rejectAll": "Avvis alle",
|
||||
"rejected": "Avvist",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analyserer tekst… dette tar vanligvis 15–60 sekunder.",
|
||||
"staleHint": "Disse ordene ble redigert etter at forslagene ble generert, så dette forslaget gjelder ikke lenger.",
|
||||
"summary": "{total} forslag · {pending} ventende · {accepted} godtatt",
|
||||
"title": "AI-korrigeringsforslag",
|
||||
"title": "Forslag til autokorrektur",
|
||||
"undo": "Angre dette forslaget",
|
||||
"cancel": "Avbryt",
|
||||
"close": "Lukk",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Tar lengre tid og koster mer, men viser hvilke forslag flere modeller er enige om",
|
||||
"compareLabel": "Sammenlign flere AI-modeller",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Modellene er uenige om disse ordene — hvis du godtar én variant, avvises de andre.",
|
||||
"conflictTooltip": "Det finnes forskjellige forslag for disse ordene — hvis du godtar én variant, avvises de andre.",
|
||||
"consensusBadge": "{consensus}/{total} modeller",
|
||||
"consensusTooltip": "Foreslått av: {models}",
|
||||
"cachedBadge": "bufret",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Accepteer alles",
|
||||
"revertAll": "Draai alles terug",
|
||||
"undoAll": "Maak alles ongedaan",
|
||||
"resetAllChanges": "Reset alle wijzigingen terug naar de originele AI-gecorrigeerde songteksten",
|
||||
"resetAllChanges": "Zet alle wijzigingen terug naar de oorspronkelijke gecorrigeerde songtekst",
|
||||
"undo": "Ongedaan maken",
|
||||
"redo": "Opnieuw uitvoeren",
|
||||
"findReplace": "Zoeken/Vervangen",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Kon audio niet laden",
|
||||
"audioStillLoading": "Audio is nog aan het laden...",
|
||||
"loadingAudio": "Audio laden...",
|
||||
"autoCorrect": "AI-suggesties",
|
||||
"autoCorrectDesc": "Genereer AI-correctiesuggesties door de transcriptie te vergelijken met de referentiesongtekst. Er wordt niets gewijzigd totdat je elke suggestie accepteert."
|
||||
"autoCorrect": "Autocorrectie",
|
||||
"autoCorrectDesc": "Genereer correctiesuggesties door de transcriptie te vergelijken met de referentiesongtekst. Er wordt niets gewijzigd totdat je elke suggestie accepteert."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Toon tips",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "overig",
|
||||
"category_repeated_chorus_fix": "refrein",
|
||||
"collapse": "Suggesties inklappen",
|
||||
"confidenceHint": "Hoe zeker de AI moet zijn om een suggestie te tonen",
|
||||
"confidenceHint": "Hoe zeker een suggestie moet zijn voordat deze wordt getoond",
|
||||
"confidenceLabel": "Suggestiefilter",
|
||||
"confidence_all": "Toon alles",
|
||||
"confidence_high": "Hoge zekerheid",
|
||||
"description": "De AI vergelijkt de getranscribeerde songtekst met de referentiebronnen en stelt correcties op woordniveau voor. Je beoordeelt elke suggestie — er wordt niets toegepast totdat je deze accepteert.",
|
||||
"description": "Autocorrectie vergelijkt de getranscribeerde songtekst met de referentiebronnen en stelt correcties op woordniveau voor. Je beoordeelt elke suggestie — er wordt niets toegepast totdat je deze accepteert.",
|
||||
"dismiss": "Suggesties negeren",
|
||||
"expand": "Suggesties uitklappen",
|
||||
"insertMarker": "invoegen:",
|
||||
"insertionsHint": "Sta suggesties toe die woorden invoegen die in de transcriptie ontbreken",
|
||||
"insertionsLabel": "Stel ontbrekende woorden voor",
|
||||
"noReferences": "Er is geen referentiesongtekst beschikbaar voor dit nummer, dus er kunnen geen AI-suggesties worden gegenereerd. Voeg eerst een songtekst toe via het referentiepaneel.",
|
||||
"noSuggestions": "Geen problemen gevonden — de AI heeft geen wijzigingen voorgesteld.",
|
||||
"noReferences": "Er is geen referentiesongtekst beschikbaar voor dit nummer, dus er kunnen geen suggesties worden gegenereerd. Voeg eerst een songtekst toe via het referentiepaneel.",
|
||||
"noSuggestions": "Geen problemen gevonden — er zijn geen voorgestelde wijzigingen.",
|
||||
"reject": "Wijs suggestie af",
|
||||
"rejectAll": "Wijs alles af",
|
||||
"rejected": "Afgewezen",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Songtekst analyseren… dit duurt meestal 15–60 seconden.",
|
||||
"staleHint": "Deze woorden zijn bewerkt nadat de suggesties zijn gegenereerd, dus deze suggestie is niet meer van toepassing.",
|
||||
"summary": "{total} suggesties · {pending} in afwachting · {accepted} geaccepteerd",
|
||||
"title": "AI-correctiesuggesties",
|
||||
"title": "Suggesties voor autocorrectie",
|
||||
"undo": "Maak deze suggestie ongedaan",
|
||||
"cancel": "Annuleren",
|
||||
"close": "Sluiten",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Langzamer en kost meer, maar laat zien over welke suggesties meerdere modellen het eens zijn",
|
||||
"compareLabel": "Vergelijk meerdere AI-modellen",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "De modellen zijn het niet eens over deze woorden — als je één variant accepteert, wijs je de andere af.",
|
||||
"conflictTooltip": "Er zijn verschillende suggesties voor deze woorden — als je één variant accepteert, worden de andere afgewezen.",
|
||||
"consensusBadge": "{consensus}/{total} modellen",
|
||||
"consensusTooltip": "Voorgesteld door: {models}",
|
||||
"cachedBadge": "gecached",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Akceptuj wszystko",
|
||||
"revertAll": "Cofnij wszystko",
|
||||
"undoAll": "Cofnij wszystko",
|
||||
"resetAllChanges": "Zresetuj wszystkie zmiany z powrotem do oryginalnego tekstu poprawionego przez AI",
|
||||
"resetAllChanges": "Cofnij wszystkie zmiany i przywróć oryginalny, poprawiony tekst",
|
||||
"undo": "Cofnij",
|
||||
"redo": "Ponów",
|
||||
"findReplace": "Znajdź/Zamień",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Nie udało się załadować audio",
|
||||
"audioStillLoading": "Audio wciąż się ładuje...",
|
||||
"loadingAudio": "Ładowanie audio...",
|
||||
"autoCorrect": "Sugestie AI",
|
||||
"autoCorrectDesc": "Generuj sugestie poprawek AI, porównując transkrypcję z tekstem referencyjnym. Nic nie zostanie zmienione, dopóki nie zaakceptujesz każdej sugestii."
|
||||
"autoCorrect": "Autokorekta",
|
||||
"autoCorrectDesc": "Wygeneruj sugestie poprawek, porównując transkrypcję z tekstem referencyjnym. Żadna zmiana nie zostanie wprowadzona, dopóki jej nie zaakceptujesz."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Pokaż wskazówki",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "inne",
|
||||
"category_repeated_chorus_fix": "refren",
|
||||
"collapse": "Zwiń sugestie",
|
||||
"confidenceHint": "Jak pewne musi być AI, aby pokazać sugestię",
|
||||
"confidenceHint": "Wymagany poziom pewności, aby wyświetlić sugestię",
|
||||
"confidenceLabel": "Filtr sugestii",
|
||||
"confidence_all": "Pokaż wszystkie",
|
||||
"confidence_high": "Wysoka pewność",
|
||||
"description": "AI porównuje transkrybowany tekst ze źródłami referencyjnymi i sugeruje poprawki na poziomie słów. Przeglądasz każdą sugestię — nic nie zostanie zmienione, dopóki jej nie zaakceptujesz.",
|
||||
"description": "Autokorekta porównuje tekst z transkrypcji ze źródłami referencyjnymi i sugeruje poprawki poszczególnych słów. Samodzielnie przeglądasz każdą sugestię — żadna zmiana nie zostanie wprowadzona, dopóki jej nie zaakceptujesz.",
|
||||
"dismiss": "Odrzuć sugestie",
|
||||
"expand": "Rozwiń sugestie",
|
||||
"insertMarker": "wstaw:",
|
||||
"insertionsHint": "Zezwalaj na sugestie wstawiające słowa, które pominięto w transkrypcji",
|
||||
"insertionsLabel": "Sugeruj brakujące słowa",
|
||||
"noReferences": "Dla tej piosenki nie ma dostępnego tekstu referencyjnego, więc nie można wygenerować sugestii AI. Najpierw dodaj tekst w panelu referencyjnym.",
|
||||
"noSuggestions": "Nie znaleziono problemów — AI nie ma sugerowanych zmian.",
|
||||
"noReferences": "Dla tego utworu brakuje tekstów referencyjnych, dlatego nie można wygenerować sugestii. Najpierw dodaj tekst w panelu referencyjnym.",
|
||||
"noSuggestions": "Nie znaleziono problemów — brak sugerowanych zmian.",
|
||||
"reject": "Odrzuć sugestię",
|
||||
"rejectAll": "Odrzuć wszystkie",
|
||||
"rejected": "Odrzucone",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analizowanie tekstu… zazwyczaj zajmuje to 15–60 sekund.",
|
||||
"staleHint": "Te słowa zostały edytowane od czasu wygenerowania sugestii, więc ta sugestia nie ma już zastosowania.",
|
||||
"summary": "{total} sugestii · {pending} oczekujących · {accepted} zaakceptowanych",
|
||||
"title": "Sugestie poprawek AI",
|
||||
"title": "Sugestie autokorekty",
|
||||
"undo": "Cofnij tę sugestię",
|
||||
"cancel": "Anuluj",
|
||||
"close": "Zamknij",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Działa wolniej i kosztuje więcej, ale pokazuje, co do których sugestii zgadza się kilka modeli",
|
||||
"compareLabel": "Porównaj kilka modeli AI",
|
||||
"conflictBadge": "wariant",
|
||||
"conflictTooltip": "Modele nie są zgodne co do tych słów — zaakceptowanie jednego wariantu odrzuca pozostałe.",
|
||||
"conflictTooltip": "Dla tych słów dostępne są różne sugestie — zaakceptowanie jednego wariantu odrzuca pozostałe.",
|
||||
"consensusBadge": "{consensus}/{total} modeli",
|
||||
"consensusTooltip": "Zaproponowane przez: {models}",
|
||||
"cachedBadge": "zbuforowane",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Aceitar Tudo",
|
||||
"revertAll": "Reverter Tudo",
|
||||
"undoAll": "Desfazer Tudo",
|
||||
"resetAllChanges": "Redefinir todas as alterações de volta para as letras originais corrigidas por IA",
|
||||
"resetAllChanges": "Redefinir todas as alterações para a letra original corrigida",
|
||||
"undo": "Desfazer",
|
||||
"redo": "Refazer",
|
||||
"findReplace": "Localizar/Substituir",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Não foi possível carregar o áudio",
|
||||
"audioStillLoading": "O áudio ainda está carregando...",
|
||||
"loadingAudio": "Carregando áudio...",
|
||||
"autoCorrect": "Sugestões da IA",
|
||||
"autoCorrectDesc": "Gere sugestões de correção da IA comparando a transcrição com a letra de referência. Nada é alterado até que você aceite cada sugestão."
|
||||
"autoCorrect": "Autocorreção",
|
||||
"autoCorrectDesc": "Gere sugestões de correção comparando a transcrição com a letra de referência. Nada é alterado até que você aceite cada sugestão."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Mostrar dicas",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "outro",
|
||||
"category_repeated_chorus_fix": "refrão",
|
||||
"collapse": "Recolher sugestões",
|
||||
"confidenceHint": "O nível de confiança que a IA deve ter para mostrar uma sugestão",
|
||||
"confidenceHint": "Nível de confiança necessário para que uma sugestão seja exibida",
|
||||
"confidenceLabel": "Filtro de sugestões",
|
||||
"confidence_all": "Mostrar todas",
|
||||
"confidence_high": "Alta confiança",
|
||||
"description": "A IA compara a letra transcrita com as fontes de referência e sugere correções palavra por palavra. Você revisa cada sugestão — nada é aplicado até que você aceite.",
|
||||
"description": "A autocorreção compara a letra transcrita com as fontes de referência e sugere correções palavra por palavra. Você revisa cada sugestão — nada é aplicado até que você aceite.",
|
||||
"dismiss": "Descartar sugestões",
|
||||
"expand": "Expandir sugestões",
|
||||
"insertMarker": "inserir:",
|
||||
"insertionsHint": "Permitir sugestões que insiram palavras que faltaram na transcrição",
|
||||
"insertionsLabel": "Sugerir palavras ausentes",
|
||||
"noReferences": "Não há letras de referência disponíveis para esta música, portanto, não é possível gerar sugestões da IA. Adicione a letra no painel de referência primeiro.",
|
||||
"noSuggestions": "Nenhum problema encontrado — a IA não tem alterações sugeridas.",
|
||||
"noReferences": "Não há letras de referência disponíveis para esta música, então não é possível gerar sugestões. Primeiro, adicione a letra no painel de referência.",
|
||||
"noSuggestions": "Nenhum problema encontrado — não há alterações sugeridas.",
|
||||
"reject": "Rejeitar sugestão",
|
||||
"rejectAll": "Rejeitar todas",
|
||||
"rejected": "Rejeitada",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analisando a letra… isso geralmente leva de 15 a 60 segundos.",
|
||||
"staleHint": "Estas palavras foram editadas desde que as sugestões foram geradas, então esta sugestão não se aplica mais.",
|
||||
"summary": "{total} sugestões · {pending} pendentes · {accepted} aceitas",
|
||||
"title": "Sugestões de correção da IA",
|
||||
"title": "Sugestões de autocorreção",
|
||||
"undo": "Desfazer esta sugestão",
|
||||
"cancel": "Cancelar",
|
||||
"close": "Fechar",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Mais lento e mais caro, mas mostra em quais sugestões vários modelos concordam",
|
||||
"compareLabel": "Comparar vários modelos de IA",
|
||||
"conflictBadge": "variante",
|
||||
"conflictTooltip": "Os modelos discordam sobre estas palavras — aceitar uma variante rejeita as outras.",
|
||||
"conflictTooltip": "Há sugestões diferentes para estas palavras — aceitar uma variante rejeita as outras.",
|
||||
"consensusBadge": "{consensus}/{total} modelos",
|
||||
"consensusTooltip": "Proposto por: {models}",
|
||||
"cachedBadge": "em cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Acceptă tot",
|
||||
"revertAll": "Revino la tot",
|
||||
"undoAll": "Anulează tot",
|
||||
"resetAllChanges": "Resetează toate modificările înapoi la versurile originale corectate de AI",
|
||||
"resetAllChanges": "Resetează toate modificările la versurile corectate inițial",
|
||||
"undo": "Anulează",
|
||||
"redo": "Refă",
|
||||
"findReplace": "Caută/Înlocuiește",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Nu s-a putut încărca fișierul audio",
|
||||
"audioStillLoading": "Fișierul audio încă se încarcă...",
|
||||
"loadingAudio": "Se încarcă fișierul audio...",
|
||||
"autoCorrect": "Sugestii AI",
|
||||
"autoCorrectDesc": "Generează sugestii de corectare AI comparând transcrierea cu versurile de referință. Nimic nu se modifică până nu accepți fiecare sugestie."
|
||||
"autoCorrect": "Autocorectare",
|
||||
"autoCorrectDesc": "Generează sugestii de corectare comparând transcrierea cu versurile de referință. Nimic nu este modificat până când nu accepți fiecare sugestie."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Arată sfaturi",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "altele",
|
||||
"category_repeated_chorus_fix": "refren",
|
||||
"collapse": "Restrânge sugestiile",
|
||||
"confidenceHint": "Cât de sigur trebuie să fie AI-ul pentru a afișa o sugestie",
|
||||
"confidenceHint": "Cât de sigură trebuie să fie o sugestie înainte de a fi afișată",
|
||||
"confidenceLabel": "Filtru de sugestii",
|
||||
"confidence_all": "Afișează toate",
|
||||
"confidence_high": "Încredere ridicată",
|
||||
"description": "AI-ul compară versurile transcrise cu sursele de referință și sugerează corecturi la nivel de cuvânt. Tu revizuiești fiecare sugestie — nimic nu se aplică până nu o accepți.",
|
||||
"description": "Autocorectarea compară versurile transcrise cu sursele de referință și sugerează corecturi la nivel de cuvânt. Tu revizuiești fiecare sugestie — nimic nu este aplicat până nu o accepți.",
|
||||
"dismiss": "Ignoră sugestiile",
|
||||
"expand": "Extinde sugestiile",
|
||||
"insertMarker": "inserează:",
|
||||
"insertionsHint": "Permite sugestii care inserează cuvinte pe care transcrierea le-a omis",
|
||||
"insertionsLabel": "Sugerează cuvintele lipsă",
|
||||
"noReferences": "Nu sunt disponibile versuri de referință pentru această melodie, așa că sugestiile AI nu pot fi generate. Adaugă mai întâi versuri din panoul de referință.",
|
||||
"noSuggestions": "Nu s-au găsit probleme — AI-ul nu are modificări sugerate.",
|
||||
"noReferences": "Nu există versuri de referință disponibile pentru această melodie, așa că nu pot fi generate sugestii. Adaugă mai întâi versuri din panoul de referință.",
|
||||
"noSuggestions": "Nu s-au găsit probleme — nu există modificări sugerate.",
|
||||
"reject": "Respinge sugestia",
|
||||
"rejectAll": "Respinge toate",
|
||||
"rejected": "Respins",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Se analizează versurile… de obicei durează 15–60 de secunde.",
|
||||
"staleHint": "Aceste cuvinte au fost editate după generarea sugestiilor, așa că această sugestie nu se mai aplică.",
|
||||
"summary": "{total} sugestii · {pending} în așteptare · {accepted} acceptate",
|
||||
"title": "Sugestii de corectare AI",
|
||||
"title": "Sugestii de autocorectare",
|
||||
"undo": "Anulează această sugestie",
|
||||
"cancel": "Anulează",
|
||||
"close": "Închide",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Mai lent și costă mai mult, dar îți arată cu ce sugestii sunt de acord mai multe modele",
|
||||
"compareLabel": "Compară mai multe modele AI",
|
||||
"conflictBadge": "variantă",
|
||||
"conflictTooltip": "Modelele au variante diferite pentru aceste cuvinte — dacă accepți o variantă, le respingi pe celelalte.",
|
||||
"conflictTooltip": "Există sugestii diferite pentru aceste cuvinte — acceptarea unei variante le respinge pe celelalte.",
|
||||
"consensusBadge": "{consensus}/{total} modele",
|
||||
"consensusTooltip": "Propus de: {models}",
|
||||
"cachedBadge": "din cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Принять все",
|
||||
"revertAll": "Отменить все",
|
||||
"undoAll": "Отменить все действия",
|
||||
"resetAllChanges": "Сбросить все изменения до оригинального текста, исправленного ИИ",
|
||||
"resetAllChanges": "Сбросить все изменения и вернуть исходный исправленный текст",
|
||||
"undo": "Отменить",
|
||||
"redo": "Повторить",
|
||||
"findReplace": "Найти/Заменить",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Не удалось загрузить аудио",
|
||||
"audioStillLoading": "Аудио всё ещё загружается...",
|
||||
"loadingAudio": "Загрузка аудио...",
|
||||
"autoCorrect": "Правки от ИИ",
|
||||
"autoCorrectDesc": "Получи от ИИ предложенные правки на основе сравнения распознанного текста с эталонным. Ничего не изменится, пока ты не примешь каждую из них."
|
||||
"autoCorrect": "Автоисправление",
|
||||
"autoCorrectDesc": "Сгенерируй варианты исправлений, сравнив распознанный текст с эталонным. Ничего не изменится, пока ты не примешь каждый вариант."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Показать советы",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "другое",
|
||||
"category_repeated_chorus_fix": "припев",
|
||||
"collapse": "Свернуть правки",
|
||||
"confidenceHint": "Насколько ИИ должен быть уверен, чтобы предложить правку",
|
||||
"confidenceHint": "Насколько точным должен быть предложенный вариант, чтобы он отобразился",
|
||||
"confidenceLabel": "Фильтр правок",
|
||||
"confidence_all": "Показать все",
|
||||
"confidence_high": "Высокая уверенность",
|
||||
"description": "ИИ сравнивает распознанный текст с эталонными источниками и предлагает исправления отдельных слов. Ты проверяешь каждую правку — ничего не изменится, пока ты её не примешь.",
|
||||
"description": "Автоисправление сравнивает распознанный текст с эталонными источниками и предлагает исправления отдельных слов. Ты проверяешь каждый вариант — ничего не изменится, пока ты его не примешь.",
|
||||
"dismiss": "Скрыть правки",
|
||||
"expand": "Развернуть правки",
|
||||
"insertMarker": "вставить:",
|
||||
"insertionsHint": "Предлагать добавление слов, пропущенных при распознавании",
|
||||
"insertionsLabel": "Предлагать пропущенные слова",
|
||||
"noReferences": "Для этой песни нет эталонного текста, поэтому ИИ не может предложить исправления. Сначала добавь текст на панели эталонов.",
|
||||
"noSuggestions": "Проблем не найдено — ИИ не предлагает никаких правок.",
|
||||
"noReferences": "Для этой песни нет эталонного текста, поэтому мы не можем сгенерировать исправления. Сначала добавь текст на панели эталонов.",
|
||||
"noSuggestions": "Ошибок не найдено — нет предложенных изменений.",
|
||||
"reject": "Отклонить правку",
|
||||
"rejectAll": "Отклонить все",
|
||||
"rejected": "Отклонено",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Анализ текста… обычно это занимает 15–60 секунд.",
|
||||
"staleHint": "Эти слова были изменены после запуска проверки, поэтому данная правка больше не актуальна.",
|
||||
"summary": "{total} правок · {pending} в ожидании · {accepted} принято",
|
||||
"title": "Исправления от ИИ",
|
||||
"title": "Варианты автоисправления",
|
||||
"undo": "Отменить эту правку",
|
||||
"cancel": "Отмена",
|
||||
"close": "Закрыть",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Медленнее и стоит дороже, но показывает, в каких вариантах сходятся несколько моделей",
|
||||
"compareLabel": "Сравнить несколько ИИ-моделей",
|
||||
"conflictBadge": "вариант",
|
||||
"conflictTooltip": "Модели расходятся во мнении насчет этих слов — приняв один вариант, ты отклонишь остальные.",
|
||||
"conflictTooltip": "Для этих слов есть разные варианты — если принять один, остальные будут отклонены.",
|
||||
"consensusBadge": "{consensus}/{total} моделей",
|
||||
"consensusTooltip": "Предложено: {models}",
|
||||
"cachedBadge": "кэшировано",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Prijať všetky",
|
||||
"revertAll": "Vrátiť všetky",
|
||||
"undoAll": "Vrátiť späť všetky",
|
||||
"resetAllChanges": "Obnoviť všetky zmeny späť na pôvodný text opravený AI",
|
||||
"resetAllChanges": "Vrátiť všetky zmeny späť na pôvodný opravený text",
|
||||
"undo": "Späť",
|
||||
"redo": "Znova",
|
||||
"findReplace": "Nájsť/Nahradiť",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Nepodarilo sa načítať audio",
|
||||
"audioStillLoading": "Audio sa stále načítava...",
|
||||
"loadingAudio": "Načítava sa audio...",
|
||||
"autoCorrect": "Návrhy AI",
|
||||
"autoCorrectDesc": "Vygeneruj si návrhy opráv od AI porovnaním prepisu s referenčným textom. Nič sa nezmení, kým jednotlivé návrhy neprijmeš."
|
||||
"autoCorrect": "Automatická oprava",
|
||||
"autoCorrectDesc": "Vygeneruj návrhy na opravu porovnaním prepisu s referenčným textom. Nič sa nezmení, kým neschváliš každý návrh."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Zobraziť tipy",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "iné",
|
||||
"category_repeated_chorus_fix": "refrén",
|
||||
"collapse": "Zbaliť návrhy",
|
||||
"confidenceHint": "Nakoľko si musí byť AI istá, aby zobrazila návrh",
|
||||
"confidenceHint": "Aká vysoká musí byť miera istoty návrhu, aby sa zobrazil",
|
||||
"confidenceLabel": "Filter návrhov",
|
||||
"confidence_all": "Zobraziť všetko",
|
||||
"confidence_high": "Vysoká istota",
|
||||
"description": "AI porovná prepísaný text s referenčnými zdrojmi a navrhne opravy na úrovni jednotlivých slov. Každý návrh si môžeš skontrolovať — nič sa nezmení, kým ho neprijmeš.",
|
||||
"description": "Automatická oprava porovnáva prepísaný text s referenčnými zdrojmi a navrhuje opravy na úrovni slov. Každý návrh skontroluješ — nič sa neaplikuje, kým ho neschváliš.",
|
||||
"dismiss": "Zahodiť návrhy",
|
||||
"expand": "Rozbaliť návrhy",
|
||||
"insertMarker": "vložiť:",
|
||||
"insertionsHint": "Povoliť návrhy na vloženie slov, ktoré v prepise chýbajú",
|
||||
"insertionsLabel": "Navrhnúť chýbajúce slová",
|
||||
"noReferences": "Pre túto pieseň nie sú k dispozícii žiadne referenčné texty, takže nie je možné vygenerovať návrhy AI. Najprv pridaj text z referenčného panela.",
|
||||
"noSuggestions": "Nenašli sa žiadne problémy — AI nenavrhuje žiadne zmeny.",
|
||||
"noReferences": "Pre túto pieseň nie je k dispozícii žiadny referenčný text, takže návrhy nie je možné vygenerovať. Najprv pridaj text z referenčného panela.",
|
||||
"noSuggestions": "Nenašli sa žiadne problémy — nie sú tu žiadne navrhované zmeny.",
|
||||
"reject": "Odmietnuť návrh",
|
||||
"rejectAll": "Odmietnuť všetko",
|
||||
"rejected": "Odmietnuté",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analyzuje sa text… zvyčajne to trvá 15 – 60 sekúnd.",
|
||||
"staleHint": "Tieto slová boli od vygenerovania návrhov upravené, takže tento návrh už neplatí.",
|
||||
"summary": "{total} návrhov · {pending} čakajúcich · {accepted} prijatých",
|
||||
"title": "Návrhy opráv AI",
|
||||
"title": "Návrhy automatickej opravy",
|
||||
"undo": "Vrátiť späť tento návrh",
|
||||
"cancel": "Zrušiť",
|
||||
"close": "Zavrieť",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Je to pomalšie a stojí to viac, ale ukáže, na ktorých návrhoch sa zhoduje viacero modelov",
|
||||
"compareLabel": "Porovnať viacero AI modelov",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Modely sa nezhodujú na týchto slovách — prijatím jedného variantu odmietneš ostatné.",
|
||||
"conflictTooltip": "Pre tieto slová existujú rôzne návrhy — prijatím jedného variantu odmietneš ostatné.",
|
||||
"consensusBadge": "{consensus}/{total} modelov",
|
||||
"consensusTooltip": "Navrhli: {models}",
|
||||
"cachedBadge": "z cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Acceptera alla",
|
||||
"revertAll": "Återställ alla",
|
||||
"undoAll": "Ångra alla",
|
||||
"resetAllChanges": "Återställ alla ändringar tillbaka till den ursprungliga AI-korrigerade texten",
|
||||
"resetAllChanges": "Återställ alla ändringar till den ursprungliga korrigerade låttexten",
|
||||
"undo": "Ångra",
|
||||
"redo": "Gör om",
|
||||
"findReplace": "Sök/Ersätt",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Kunde inte ladda ljudet",
|
||||
"audioStillLoading": "Ljudet laddas fortfarande...",
|
||||
"loadingAudio": "Laddar ljud...",
|
||||
"autoCorrect": "AI-förslag",
|
||||
"autoCorrectDesc": "Generera AI-korrigeringsförslag genom att jämföra transkriberingen med referenstexten. Inget ändras förrän du accepterar varje förslag."
|
||||
"autoCorrect": "Autokorrigering",
|
||||
"autoCorrectDesc": "Generera korrigeringsförslag genom att jämföra transkriptionen med referenstexten. Ingenting ändras förrän du accepterar varje förslag."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Visa tips",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "annat",
|
||||
"category_repeated_chorus_fix": "refräng",
|
||||
"collapse": "Fäll ihop förslag",
|
||||
"confidenceHint": "Hur säker AI:n måste vara för att visa ett förslag",
|
||||
"confidenceHint": "Hur säkert ett förslag måste vara innan det visas",
|
||||
"confidenceLabel": "Förslagsfilter",
|
||||
"confidence_all": "Visa alla",
|
||||
"confidence_high": "Hög säkerhet",
|
||||
"description": "AI:n jämför den transkriberade låttexten med referenskällorna och föreslår korrigeringar på ordnivå. Du granskar varje förslag — inget tillämpas förrän du accepterar det.",
|
||||
"description": "Autokorrigering jämför den transkriberade låttexten med referenskällorna och föreslår rättningar på ordnivå. Du granskar varje förslag — ingenting tillämpas förrän du accepterar det.",
|
||||
"dismiss": "Avfärda förslag",
|
||||
"expand": "Fäll ut förslag",
|
||||
"insertMarker": "infoga:",
|
||||
"insertionsHint": "Tillåt förslag som infogar ord som transkriberingen missade",
|
||||
"insertionsLabel": "Föreslå saknade ord",
|
||||
"noReferences": "Inga referenstexter är tillgängliga för den här låten, så AI-förslag kan inte genereras. Lägg till låttext från referenspanelen först.",
|
||||
"noSuggestions": "Inga problem hittades — AI:n har inga föreslagna ändringar.",
|
||||
"noReferences": "Det finns inga referenstexter för den här låten, så det går inte att generera några förslag. Lägg till en text från referenspanelen först.",
|
||||
"noSuggestions": "Inga problem hittades — det finns inga föreslagna ändringar.",
|
||||
"reject": "Avvisa förslag",
|
||||
"rejectAll": "Avvisa alla",
|
||||
"rejected": "Avvisat",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Analyserar låttext… detta tar vanligtvis 15–60 sekunder.",
|
||||
"staleHint": "Dessa ord har redigerats sedan förslagen genererades, så detta förslag gäller inte längre.",
|
||||
"summary": "{total} förslag · {pending} väntande · {accepted} accepterade",
|
||||
"title": "AI-korrigeringsförslag",
|
||||
"title": "Autokorrigeringsförslag",
|
||||
"undo": "Ångra detta förslag",
|
||||
"cancel": "Avbryt",
|
||||
"close": "Stäng",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Långsammare och kostar mer, men visar vilka förslag flera modeller är överens om",
|
||||
"compareLabel": "Jämför flera AI-modeller",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Modellerna är oense om dessa ord — om du accepterar en variant väljs de andra bort.",
|
||||
"conflictTooltip": "Det finns olika förslag för dessa ord — om du accepterar en variant avvisas de andra.",
|
||||
"consensusBadge": "{consensus}/{total} modeller",
|
||||
"consensusTooltip": "Föreslaget av: {models}",
|
||||
"cachedBadge": "cachad",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "ยอมรับทั้งหมด",
|
||||
"revertAll": "ย้อนกลับทั้งหมด",
|
||||
"undoAll": "เลิกทำทั้งหมด",
|
||||
"resetAllChanges": "รีเซ็ตการเปลี่ยนแปลงทั้งหมดกลับไปเป็นเนื้อเพลงที่แก้ไขโดย AI ต้นฉบับ",
|
||||
"resetAllChanges": "รีเซ็ตการเปลี่ยนแปลงทั้งหมดกลับไปเป็นเนื้อเพลงที่แก้ไขไว้ตอนแรก",
|
||||
"undo": "เลิกทำ",
|
||||
"redo": "ทำซ้ำ",
|
||||
"findReplace": "ค้นหา/แทนที่",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "ไม่สามารถโหลดเสียงได้",
|
||||
"audioStillLoading": "ไฟล์เสียงยังโหลดอยู่...",
|
||||
"loadingAudio": "กำลังโหลดไฟล์เสียง...",
|
||||
"autoCorrect": "AI แนะนำ",
|
||||
"autoCorrectDesc": "สร้างคำแนะนำการแก้ไขจาก AI โดยเปรียบเทียบการถอดความกับเนื้อเพลงอ้างอิง จะไม่มีการเปลี่ยนแปลงใดๆ จนกว่าคุณจะยอมรับแต่ละคำแนะนำ"
|
||||
"autoCorrect": "แก้ไขอัตโนมัติ",
|
||||
"autoCorrectDesc": "สร้างคำแนะนำการแก้ไขโดยเปรียบเทียบการถอดเสียงกับเนื้อเพลงอ้างอิง จะไม่มีการเปลี่ยนแปลงใดๆ จนกว่าคุณจะยอมรับแต่ละคำแนะนำ"
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "แสดงเคล็ดลับ",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "อื่นๆ",
|
||||
"category_repeated_chorus_fix": "คอรัส",
|
||||
"collapse": "ย่อคำแนะนำ",
|
||||
"confidenceHint": "ระดับความมั่นใจของ AI ในการแสดงคำแนะนำ",
|
||||
"confidenceHint": "ระดับความมั่นใจขั้นต่ำก่อนแสดงคำแนะนำ",
|
||||
"confidenceLabel": "ตัวกรองคำแนะนำ",
|
||||
"confidence_all": "แสดงทั้งหมด",
|
||||
"confidence_high": "ความมั่นใจสูง",
|
||||
"description": "AI จะเปรียบเทียบเนื้อเพลงที่ถอดความกับแหล่งอ้างอิงและแนะนำการแก้ไขในระดับคำ คุณสามารถตรวจสอบทุกคำแนะนำได้ — จะไม่มีการเปลี่ยนแปลงใดๆ จนกว่าคุณจะกดยอมรับ",
|
||||
"description": "ระบบแก้ไขอัตโนมัติจะเปรียบเทียบเนื้อเพลงที่ถอดเสียงกับแหล่งอ้างอิงและแนะนำการแก้ไขในระดับคำ คุณสามารถตรวจสอบทุกคำแนะนำ — จะไม่มีการเปลี่ยนแปลงใดๆ จนกว่าคุณจะกดยอมรับ",
|
||||
"dismiss": "ปิดคำแนะนำ",
|
||||
"expand": "ขยายคำแนะนำ",
|
||||
"insertMarker": "แทรก:",
|
||||
"insertionsHint": "อนุญาตให้แนะนำการแทรกคำที่ตกหล่นจากการถอดความ",
|
||||
"insertionsLabel": "แนะนำคำที่หายไป",
|
||||
"noReferences": "ไม่มีเนื้อเพลงอ้างอิงสำหรับเพลงนี้ จึงไม่สามารถสร้างคำแนะนำจาก AI ได้ กรุณาเพิ่มเนื้อเพลงจากแผงอ้างอิงก่อน",
|
||||
"noSuggestions": "ไม่พบปัญหา — AI ไม่มีคำแนะนำในการแก้ไข",
|
||||
"noReferences": "ไม่มีเนื้อเพลงอ้างอิงสำหรับเพลงนี้ จึงไม่สามารถสร้างคำแนะนำได้ โปรดเพิ่มเนื้อเพลงจากแผงอ้างอิงก่อน",
|
||||
"noSuggestions": "ไม่พบปัญหา — ไม่มีคำแนะนำในการแก้ไข",
|
||||
"reject": "ปฏิเสธคำแนะนำ",
|
||||
"rejectAll": "ปฏิเสธทั้งหมด",
|
||||
"rejected": "ปฏิเสธแล้ว",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "กำลังวิเคราะห์เนื้อเพลง… โดยปกติจะใช้เวลา 15–60 วินาที",
|
||||
"staleHint": "คำเหล่านี้ถูกแก้ไขหลังจากที่สร้างคำแนะนำแล้ว คำแนะนำนี้จึงไม่สามารถใช้ได้อีกต่อไป",
|
||||
"summary": "{total} คำแนะนำ · รอดำเนินการ {pending} รายการ · ยอมรับแล้ว {accepted} รายการ",
|
||||
"title": "คำแนะนำการแก้ไขจาก AI",
|
||||
"title": "คำแนะนำการแก้ไขอัตโนมัติ",
|
||||
"undo": "เลิกทำคำแนะนำนี้",
|
||||
"cancel": "ยกเลิก",
|
||||
"close": "ปิด",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "ช้ากว่าและมีค่าใช้จ่ายสูงกว่า แต่จะแสดงคำแนะนำที่โมเดลหลายตัวเห็นตรงกัน",
|
||||
"compareLabel": "เปรียบเทียบโมเดล AI หลายตัว",
|
||||
"conflictBadge": "ตัวเลือก",
|
||||
"conflictTooltip": "โมเดลให้ผลลัพธ์คำเหล่านี้ไม่ตรงกัน — การเลือกตัวเลือกหนึ่งจะเป็นการปฏิเสธตัวเลือกอื่นๆ",
|
||||
"conflictTooltip": "มีคำแนะนำหลายแบบสำหรับคำเหล่านี้ — การเลือกแบบหนึ่งจะเป็นการปฏิเสธแบบอื่นๆ",
|
||||
"consensusBadge": "{consensus}/{total} โมเดล",
|
||||
"consensusTooltip": "เสนอโดย: {models}",
|
||||
"cachedBadge": "แคชไว้",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Tanggapin Lahat",
|
||||
"revertAll": "I-revert Lahat",
|
||||
"undoAll": "I-undo Lahat",
|
||||
"resetAllChanges": "I-reset ang lahat ng pagbabago pabalik sa orihinal na AI-corrected lyrics",
|
||||
"resetAllChanges": "I-reset ang lahat ng pagbabago pabalik sa orihinal na naiwastong lyrics",
|
||||
"undo": "I-undo",
|
||||
"redo": "I-redo",
|
||||
"findReplace": "Hanapin/Palitan",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Hindi ma-load ang audio",
|
||||
"audioStillLoading": "Naglo-load pa rin ang audio...",
|
||||
"loadingAudio": "Naglo-load ang audio...",
|
||||
"autoCorrect": "AI Suggest",
|
||||
"autoCorrectDesc": "Bumuo ng mga mungkahi sa pagwawasto ng AI sa pamamagitan ng pagkukumpara ng transcription sa reference lyrics. Walang mababago hangga't hindi mo tinatanggap ang bawat mungkahi."
|
||||
"autoCorrect": "Auto-correct",
|
||||
"autoCorrectDesc": "Bumuo ng mga mungkahi sa pagwawasto sa pamamagitan ng paghahambing ng transcription sa reference lyrics. Walang mababago hangga't hindi mo tinatanggap ang bawat mungkahi."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Ipakita ang mga tip",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "iba pa",
|
||||
"category_repeated_chorus_fix": "chorus",
|
||||
"collapse": "I-collapse ang mga mungkahi",
|
||||
"confidenceHint": "Gaano dapat kasigurado ang AI para magpakita ng mungkahi",
|
||||
"confidenceHint": "Kung gaano katiyak ang isang mungkahi bago ito ipakita",
|
||||
"confidenceLabel": "Filter ng mungkahi",
|
||||
"confidence_all": "Ipakita lahat",
|
||||
"confidence_high": "Mataas na kasiguraduhan",
|
||||
"description": "Ikinukumpara ng AI ang mga na-transcribe na lyrics sa mga reference source at nagmumungkahi ng mga pag-aayos sa bawat salita. Susuriin mo ang bawat mungkahi — walang mababago hangga't hindi mo ito tinatanggap.",
|
||||
"description": "Inihahambing ng auto-correct ang mga na-transcribe na lyrics sa mga reference source at nagmumungkahi ng mga pag-aayos sa bawat salita. Susuriin mo ang bawat mungkahi — walang ilalapat hangga't hindi mo ito tinatanggap.",
|
||||
"dismiss": "I-dismiss ang mga mungkahi",
|
||||
"expand": "I-expand ang mga mungkahi",
|
||||
"insertMarker": "idagdag:",
|
||||
"insertionsHint": "Payagan ang mga mungkahi na magdagdag ng mga salitang nakaligtaan sa transcription",
|
||||
"insertionsLabel": "I-suggest ang mga nawawalang salita",
|
||||
"noReferences": "Walang available na reference lyrics para sa kantang ito, kaya hindi makakagawa ng mga AI suggestion. Magdagdag muna ng lyrics mula sa reference panel.",
|
||||
"noSuggestions": "Walang nakitang isyu — walang iminumungkahing pagbabago ang AI.",
|
||||
"noReferences": "Walang available na reference lyrics para sa kantang ito, kaya hindi makabuo ng mga mungkahi. Magdagdag muna ng lyrics mula sa reference panel.",
|
||||
"noSuggestions": "Walang nakitang isyu — walang mga iminungkahing pagbabago.",
|
||||
"reject": "Tanggihan ang mungkahi",
|
||||
"rejectAll": "Tanggihan lahat",
|
||||
"rejected": "Tinanggihan",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Sinusuri ang lyrics… karaniwang tumatagal ito ng 15–60 segundo.",
|
||||
"staleHint": "Na-edit na ang mga salitang ito simula nang ginawa ang mga mungkahi, kaya hindi na naaangkop ang mungkahing ito.",
|
||||
"summary": "{total} mungkahi · {pending} nakabinbin · {accepted} tinanggap",
|
||||
"title": "Mga mungkahi sa pagwawasto ng AI",
|
||||
"title": "Mga mungkahi ng auto-correct",
|
||||
"undo": "I-undo ang mungkahing ito",
|
||||
"cancel": "Kanselahin",
|
||||
"close": "Isara",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Mas mabagal at mas mahal, pero ipinapakita nito kung aling mga mungkahi ang sinasang-ayunan ng iba't ibang modelo",
|
||||
"compareLabel": "Paghambingin ang iba't ibang AI model",
|
||||
"conflictBadge": "variant",
|
||||
"conflictTooltip": "Hindi magkakatugma ang mga modelo sa mga salitang ito — kapag tinanggap ang isang variant, tatanggihan ang iba.",
|
||||
"conflictTooltip": "May iba't ibang mungkahi para sa mga salitang ito — ang pagtanggap sa isang variant ay magbabalewala sa iba.",
|
||||
"consensusBadge": "{consensus}/{total} modelo",
|
||||
"consensusTooltip": "Iminungkahi ng: {models}",
|
||||
"cachedBadge": "naka-cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Tümünü Kabul Et",
|
||||
"revertAll": "Tümünü Geri Al",
|
||||
"undoAll": "Tümünü Geri Al",
|
||||
"resetAllChanges": "Tüm değişiklikleri orijinal yapay zeka düzeltmeli sözlere sıfırla",
|
||||
"resetAllChanges": "Tüm değişiklikleri sıfırla ve orijinal düzeltilmiş şarkı sözlerine geri dön",
|
||||
"undo": "Geri Al",
|
||||
"redo": "Yinele",
|
||||
"findReplace": "Bul/Değiştir",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Ses yüklenemedi",
|
||||
"audioStillLoading": "Ses hâlâ yükleniyor...",
|
||||
"loadingAudio": "Ses yükleniyor...",
|
||||
"autoCorrect": "Yapay Zeka Önerisi",
|
||||
"autoCorrectDesc": "Deşifreyi referans şarkı sözleriyle karşılaştırarak yapay zeka düzeltme önerileri oluştur. Sen her bir öneriyi kabul edene kadar hiçbir şey değiştirilmez."
|
||||
"autoCorrect": "Otomatik düzeltme",
|
||||
"autoCorrectDesc": "Yazıya dökülen metni referans şarkı sözleriyle karşılaştırarak düzeltme önerileri oluştur. Sen her bir öneriyi onaylamadan hiçbir şey değiştirilmez."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "İpuçlarını göster",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "diğer",
|
||||
"category_repeated_chorus_fix": "nakarat",
|
||||
"collapse": "Önerileri daralt",
|
||||
"confidenceHint": "Yapay zekanın bir öneri göstermesi için ne kadar emin olması gerektiği",
|
||||
"confidenceHint": "Bir önerinin gösterilmesi için gereken güvenilirlik seviyesi",
|
||||
"confidenceLabel": "Öneri filtresi",
|
||||
"confidence_all": "Tümünü göster",
|
||||
"confidence_high": "Yüksek güven",
|
||||
"description": "Yapay zeka, deşifre edilen şarkı sözlerini referans kaynaklarla karşılaştırır ve kelime düzeyinde düzeltmeler önerir. Her öneriyi sen incelersin — sen kabul edene kadar hiçbir şey uygulanmaz.",
|
||||
"description": "Otomatik düzeltme, yazıya dökülen şarkı sözlerini referans kaynaklarla karşılaştırır ve kelime bazında düzeltmeler önerir. Tüm önerileri sen incelersin — sen onaylamadan hiçbir değişiklik uygulanmaz.",
|
||||
"dismiss": "Önerileri yoksay",
|
||||
"expand": "Önerileri genişlet",
|
||||
"insertMarker": "ekle:",
|
||||
"insertionsHint": "Deşifrede eksik olan kelimeleri ekleyen önerilere izin ver",
|
||||
"insertionsLabel": "Eksik kelimeleri öner",
|
||||
"noReferences": "Bu şarkı için referans şarkı sözü bulunmuyor, bu yüzden yapay zeka önerileri oluşturulamıyor. Önce referans panelinden şarkı sözlerini ekle.",
|
||||
"noSuggestions": "Hiçbir sorun bulunamadı — yapay zekanın önerdiği bir değişiklik yok.",
|
||||
"noReferences": "Bu şarkı için referans şarkı sözleri bulunmuyor, bu yüzden öneriler oluşturulamıyor. Önce referans panelinden şarkı sözlerini ekle.",
|
||||
"noSuggestions": "Sorun bulunamadı — önerilen bir değişiklik yok.",
|
||||
"reject": "Öneriyi reddet",
|
||||
"rejectAll": "Tümünü reddet",
|
||||
"rejected": "Reddedildi",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Şarkı sözleri analiz ediliyor… bu genellikle 15–60 saniye sürer.",
|
||||
"staleHint": "Bu kelimeler öneriler oluşturulduktan sonra düzenlendi, bu yüzden bu öneri artık geçerli değil.",
|
||||
"summary": "{total} öneri · {pending} bekleyen · {accepted} kabul edilen",
|
||||
"title": "Yapay zeka düzeltme önerileri",
|
||||
"title": "Otomatik düzeltme önerileri",
|
||||
"undo": "Bu öneriyi geri al",
|
||||
"cancel": "İptal",
|
||||
"close": "Kapat",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Daha yavaş ve daha maliyetli, ancak birden fazla modelin hangi önerilerde uzlaştığını gösterir",
|
||||
"compareLabel": "Birden fazla yapay zeka modelini karşılaştır",
|
||||
"conflictBadge": "varyant",
|
||||
"conflictTooltip": "Modeller bu kelimeler konusunda uyuşmuyor — bir varyantı kabul ettiğinde diğerlerini reddetmiş olursun.",
|
||||
"conflictTooltip": "Bu kelimeler için farklı öneriler var — bir seçeneği kabul ettiğinde diğerleri reddedilir.",
|
||||
"consensusBadge": "{consensus}/{total} model",
|
||||
"consensusTooltip": "Öneren: {models}",
|
||||
"cachedBadge": "önbellekten",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Прийняти всі",
|
||||
"revertAll": "Повернути всі",
|
||||
"undoAll": "Скасувати всі",
|
||||
"resetAllChanges": "Скинути всі зміни до оригінального тексту, виправленого ШІ",
|
||||
"resetAllChanges": "Скинути всі зміни до початкового виправленого тексту",
|
||||
"undo": "Скасувати",
|
||||
"redo": "Повторити",
|
||||
"findReplace": "Знайти/Замінити",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Не вдалося завантажити аудіо",
|
||||
"audioStillLoading": "Аудіо ще завантажується...",
|
||||
"loadingAudio": "Завантаження аудіо...",
|
||||
"autoCorrect": "Пропозиції ШІ",
|
||||
"autoCorrectDesc": "Генеруй пропозиції виправлень від ШІ, порівнюючи транскрипцію з еталонним текстом. Нічого не зміниться, поки ти не приймеш кожну пропозицію."
|
||||
"autoCorrect": "Автовиправлення",
|
||||
"autoCorrectDesc": "Згенеруй пропозиції виправлень, порівнявши транскрипцію з еталонним текстом. Нічого не зміниться, поки ти не приймеш кожну пропозицію."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Показати підказки",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "інше",
|
||||
"category_repeated_chorus_fix": "приспів",
|
||||
"collapse": "Згорнути пропозиції",
|
||||
"confidenceHint": "Наскільки впевненим має бути ШІ, щоб показати пропозицію",
|
||||
"confidenceHint": "Наскільки надійною має бути пропозиція, перш ніж її буде показано",
|
||||
"confidenceLabel": "Фільтр пропозицій",
|
||||
"confidence_all": "Показати всі",
|
||||
"confidence_high": "Висока впевненість",
|
||||
"description": "ШІ порівнює транскрибований текст із еталонними джерелами та пропонує виправлення на рівні слів. Ти перевіряєш кожну пропозицію — нічого не застосовується, поки ти її не приймеш.",
|
||||
"description": "Автовиправлення порівнює транскрибований текст із еталонними джерелами та пропонує виправлення окремих слів. Ти перевіряєш кожну пропозицію — нічого не застосується, поки ти її не приймеш.",
|
||||
"dismiss": "Приховати пропозиції",
|
||||
"expand": "Розгорнути пропозиції",
|
||||
"insertMarker": "вставити:",
|
||||
"insertionsHint": "Дозволити пропозиції, що додають слова, пропущені в транскрипції",
|
||||
"insertionsLabel": "Пропонувати пропущені слова",
|
||||
"noReferences": "Для цієї пісні немає еталонних текстів, тому пропозиції ШІ неможливо згенерувати. Спочатку додай текст на панелі еталонних текстів.",
|
||||
"noSuggestions": "Проблем не знайдено — ШІ не має пропозицій щодо змін.",
|
||||
"noReferences": "Для цієї пісні немає еталонного тексту, тому згенерувати пропозиції неможливо. Спочатку додай текст із панелі еталонів.",
|
||||
"noSuggestions": "Проблем не знайдено — немає запропонованих змін.",
|
||||
"reject": "Відхилити пропозицію",
|
||||
"rejectAll": "Відхилити всі",
|
||||
"rejected": "Відхилено",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Аналізуємо текст… зазвичай це займає 15–60 секунд.",
|
||||
"staleHint": "Ці слова були відредаговані після генерації пропозицій, тому ця пропозиція більше не актуальна.",
|
||||
"summary": "{total} пропозицій · {pending} очікують · {accepted} прийнято",
|
||||
"title": "Пропозиції виправлень від ШІ",
|
||||
"title": "Пропозиції автовиправлення",
|
||||
"undo": "Скасувати цю пропозицію",
|
||||
"cancel": "Скасувати",
|
||||
"close": "Закрити",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Повільніше та дорожче, але показує, з якими пропозиціями погоджуються кілька моделей",
|
||||
"compareLabel": "Порівняти кілька моделей ШІ",
|
||||
"conflictBadge": "варіант",
|
||||
"conflictTooltip": "Моделі розійшлися щодо цих слів — обравши один варіант, ти відхилиш інші.",
|
||||
"conflictTooltip": "Для цих слів є різні пропозиції — якщо прийняти один варіант, інші будуть відхилені.",
|
||||
"consensusBadge": "{consensus}/{total} моделей",
|
||||
"consensusTooltip": "Запропоновано: {models}",
|
||||
"cachedBadge": "кешовано",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "Chấp nhận tất cả",
|
||||
"revertAll": "Hoàn nguyên tất cả",
|
||||
"undoAll": "Hoàn tác tất cả",
|
||||
"resetAllChanges": "Khôi phục tất cả thay đổi về lời bài hát gốc do AI tạo",
|
||||
"resetAllChanges": "Khôi phục tất cả thay đổi về lời bài hát đã sửa ban đầu",
|
||||
"undo": "Hoàn tác",
|
||||
"redo": "Làm lại",
|
||||
"findReplace": "Tìm/Thay thế",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "Không thể tải âm thanh",
|
||||
"audioStillLoading": "Âm thanh vẫn đang tải...",
|
||||
"loadingAudio": "Đang tải âm thanh...",
|
||||
"autoCorrect": "Đề xuất từ AI",
|
||||
"autoCorrectDesc": "Tạo các đề xuất sửa lỗi từ AI bằng cách so sánh bản chép lời với lời bài hát tham khảo. Sẽ không có thay đổi nào được áp dụng cho đến khi bạn chấp nhận từng đề xuất."
|
||||
"autoCorrect": "Tự động sửa lỗi",
|
||||
"autoCorrectDesc": "Tạo các đề xuất sửa lỗi bằng cách so sánh bản chép lời với lời bài hát tham khảo. Sẽ không có thay đổi nào được áp dụng cho đến khi bạn chấp nhận từng đề xuất."
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "Hiện mẹo",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "khác",
|
||||
"category_repeated_chorus_fix": "điệp khúc",
|
||||
"collapse": "Thu gọn đề xuất",
|
||||
"confidenceHint": "Mức độ chắc chắn của AI để hiển thị đề xuất",
|
||||
"confidenceHint": "Mức độ tin cậy cần thiết để hiển thị đề xuất",
|
||||
"confidenceLabel": "Bộ lọc đề xuất",
|
||||
"confidence_all": "Hiển thị tất cả",
|
||||
"confidence_high": "Độ chắc chắn cao",
|
||||
"description": "AI sẽ so sánh bản chép lời với các nguồn tham khảo và đề xuất sửa lỗi cho từng từ. Bạn có thể xem xét từng đề xuất — sẽ không có thay đổi nào được áp dụng cho đến khi bạn chấp nhận.",
|
||||
"description": "Tính năng Tự động sửa lỗi sẽ so sánh bản chép lời với các nguồn tham khảo và đề xuất sửa lỗi cho từng từ. Bạn có thể xem xét từng đề xuất — sẽ không có thay đổi nào được áp dụng cho đến khi bạn chấp nhận.",
|
||||
"dismiss": "Bỏ qua đề xuất",
|
||||
"expand": "Mở rộng đề xuất",
|
||||
"insertMarker": "chèn:",
|
||||
"insertionsHint": "Cho phép các đề xuất chèn những từ mà bản chép lời bỏ sót",
|
||||
"insertionsLabel": "Đề xuất các từ còn thiếu",
|
||||
"noReferences": "Bài hát này chưa có lời tham khảo nên AI không thể tạo đề xuất. Bạn hãy thêm lời bài hát từ bảng tham khảo trước nhé.",
|
||||
"noSuggestions": "Không tìm thấy vấn đề nào — AI không có đề xuất thay đổi nào.",
|
||||
"noReferences": "Không có lời bài hát tham khảo nào cho bài hát này, vì vậy không thể tạo đề xuất. Hãy thêm lời bài hát từ bảng tham khảo trước.",
|
||||
"noSuggestions": "Không tìm thấy vấn đề nào — không có thay đổi nào được đề xuất.",
|
||||
"reject": "Từ chối đề xuất",
|
||||
"rejectAll": "Từ chối tất cả",
|
||||
"rejected": "Đã từ chối",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "Đang phân tích lời bài hát… quá trình này thường mất 15–60 giây.",
|
||||
"staleHint": "Những từ này đã được chỉnh sửa sau khi đề xuất được tạo, nên đề xuất này không còn phù hợp nữa.",
|
||||
"summary": "{total} đề xuất · {pending} đang chờ · {accepted} đã chấp nhận",
|
||||
"title": "Đề xuất sửa lỗi của AI",
|
||||
"title": "Đề xuất tự động sửa lỗi",
|
||||
"undo": "Hoàn tác đề xuất này",
|
||||
"cancel": "Hủy",
|
||||
"close": "Đóng",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "Chậm hơn và tốn phí hơn, nhưng cho thấy các gợi ý được nhiều mô hình cùng đồng thuận",
|
||||
"compareLabel": "So sánh nhiều mô hình AI",
|
||||
"conflictBadge": "biến thể",
|
||||
"conflictTooltip": "Các mô hình không đồng nhất về những từ này — chấp nhận một biến thể sẽ loại bỏ các biến thể khác.",
|
||||
"conflictTooltip": "Có nhiều đề xuất khác nhau cho những từ này — khi chấp nhận một phương án, các phương án khác sẽ bị loại bỏ.",
|
||||
"consensusBadge": "{consensus}/{total} mô hình",
|
||||
"consensusTooltip": "Đề xuất bởi: {models}",
|
||||
"cachedBadge": "đã lưu cache",
|
||||
|
||||
@@ -807,7 +807,7 @@
|
||||
"acceptAll": "全部接受",
|
||||
"revertAll": "全部还原",
|
||||
"undoAll": "全部撤销",
|
||||
"resetAllChanges": "将所有更改重置回原始的 AI 校对歌词",
|
||||
"resetAllChanges": "将所有更改重置为最初的纠错歌词",
|
||||
"undo": "撤销",
|
||||
"redo": "重做",
|
||||
"findReplace": "查找/替换",
|
||||
@@ -821,8 +821,8 @@
|
||||
"audioLoadError": "无法加载音频",
|
||||
"audioStillLoading": "音频仍在加载中...",
|
||||
"loadingAudio": "正在加载音频...",
|
||||
"autoCorrect": "AI 建议",
|
||||
"autoCorrectDesc": "通过将转录歌词与参考歌词进行对比来生成 AI 纠错建议。在您接受每条建议之前,不会更改任何内容。"
|
||||
"autoCorrect": "自动纠错",
|
||||
"autoCorrectDesc": "通过将转录内容与参考歌词进行比较来生成纠错建议。在您接受每条建议之前,不会进行任何更改。"
|
||||
},
|
||||
"guidance": {
|
||||
"showTips": "显示提示",
|
||||
@@ -1294,18 +1294,18 @@
|
||||
"category_other": "其他",
|
||||
"category_repeated_chorus_fix": "副歌",
|
||||
"collapse": "折叠建议",
|
||||
"confidenceHint": "AI 显示建议所需的置信度",
|
||||
"confidenceHint": "显示建议所需的置信度",
|
||||
"confidenceLabel": "建议筛选器",
|
||||
"confidence_all": "显示全部",
|
||||
"confidence_high": "高置信度",
|
||||
"description": "AI 会将转录的歌词与参考来源进行对比,并提供词级别的修改建议。您可以逐一审核所有建议——在您接受之前,不会应用任何更改。",
|
||||
"description": "自动纠错功能会将转录歌词与参考来源进行比较,并提供词级别的修改建议。您可以审查每条建议——在您接受之前,不会应用任何更改。",
|
||||
"dismiss": "忽略建议",
|
||||
"expand": "展开建议",
|
||||
"insertMarker": "插入:",
|
||||
"insertionsHint": "允许建议插入转录漏掉的词",
|
||||
"insertionsLabel": "建议补充漏词",
|
||||
"noReferences": "这首歌没有可用的参考歌词,因此无法生成 AI 建议。请先从参考面板添加歌词。",
|
||||
"noSuggestions": "未发现问题——AI 没有修改建议。",
|
||||
"noReferences": "此歌曲没有可用的参考歌词,因此无法生成建议。请先从参考面板添加歌词。",
|
||||
"noSuggestions": "未发现问题——没有修改建议。",
|
||||
"reject": "拒绝建议",
|
||||
"rejectAll": "全部拒绝",
|
||||
"rejected": "已拒绝",
|
||||
@@ -1313,7 +1313,7 @@
|
||||
"running": "正在分析歌词……这通常需要 15–60 秒。",
|
||||
"staleHint": "这些词在生成建议后已被编辑,因此该建议不再适用。",
|
||||
"summary": "{total} 条建议 · {pending} 条待处理 · {accepted} 条已接受",
|
||||
"title": "AI 纠错建议",
|
||||
"title": "自动纠错建议",
|
||||
"undo": "撤销此建议",
|
||||
"cancel": "取消",
|
||||
"close": "关闭",
|
||||
@@ -1321,7 +1321,7 @@
|
||||
"compareHint": "速度较慢且成本较高,但会显示多个模型一致认可的建议",
|
||||
"compareLabel": "比较多个 AI 模型",
|
||||
"conflictBadge": "选项",
|
||||
"conflictTooltip": "各模型对这些词存在分歧 —— 接受其中一个选项将放弃其他选项。",
|
||||
"conflictTooltip": "这些词有不同的修改建议——接受其中一个会自动拒绝其他建议。",
|
||||
"consensusBadge": "{consensus}/{total} 个模型",
|
||||
"consensusTooltip": "推荐模型:{models}",
|
||||
"cachedBadge": "已缓存",
|
||||
|
||||
+1
-1
@@ -1,6 +1,6 @@
|
||||
[tool.poetry]
|
||||
name = "karaoke-gen"
|
||||
version = "0.179.0"
|
||||
version = "0.180.0"
|
||||
description = "Generate karaoke videos with synchronized lyrics. Handles the entire process from downloading audio and lyrics to creating the final video with title screens."
|
||||
authors = ["Andrew Beveridge <andrew@beveridge.uk>"]
|
||||
license = "MIT"
|
||||
|
||||
@@ -0,0 +1,323 @@
|
||||
#!/usr/bin/env python3
|
||||
"""
|
||||
Measure the real token usage and cost of the AI lyrics auto-correct feature
|
||||
against production jobs.
|
||||
|
||||
For each job it loads the stored corrections (the same transcription +
|
||||
reference lyrics the review UI sends to ``POST /{job_id}/auto-correct``),
|
||||
rebuilds the exact prompts, calls each configured model once, and reports
|
||||
input/output/thinking tokens and estimated USD per model — plus a per-track
|
||||
and aggregate summary. This is the immediate ground-truth read that the
|
||||
production usage logging (``auto-correct usage`` / ``auto-correct cost`` lines)
|
||||
accumulates over time.
|
||||
|
||||
Usage:
|
||||
# Auto-pick the N most recent complete jobs that have corrections
|
||||
python scripts/measure_autocorrect_cost.py --limit 3
|
||||
|
||||
# Measure specific jobs
|
||||
python scripts/measure_autocorrect_cost.py --jobs abc123 def456
|
||||
|
||||
# Single-model instead of the multi-model compare config
|
||||
python scripts/measure_autocorrect_cost.py --limit 3 --models claude-fable-5
|
||||
|
||||
Requires:
|
||||
- ANTHROPIC_API_KEY in the environment (for claude-* models)
|
||||
- GCP creds for Vertex/Firestore/GCS:
|
||||
gcloud auth application-default login (or GOOGLE_APPLICATION_CREDENTIALS)
|
||||
- GOOGLE_CLOUD_PROJECT=nomadkaraoke (set automatically below if unset)
|
||||
|
||||
NOTE: every run makes real, billed model calls. Gemini USD figures use the
|
||||
ESTIMATED rates in backend/services/auto_correct/pricing.py — token counts are
|
||||
exact; verify the Gemini rate before trusting its dollar value.
|
||||
"""
|
||||
from __future__ import annotations
|
||||
|
||||
import argparse
|
||||
import json
|
||||
import logging
|
||||
import os
|
||||
import sys
|
||||
import time
|
||||
from typing import Optional
|
||||
|
||||
os.environ.setdefault("GOOGLE_CLOUD_PROJECT", "nomadkaraoke")
|
||||
|
||||
# Running "python scripts/foo.py" puts scripts/ on sys.path, not the repo root,
|
||||
# so the backend.* imports below would fail. Add the repo root explicitly.
|
||||
sys.path.insert(0, os.path.dirname(os.path.dirname(os.path.abspath(__file__))))
|
||||
|
||||
from google.cloud import firestore, storage # noqa: E402
|
||||
|
||||
logging.basicConfig(level=logging.WARNING, format="%(levelname)s %(message)s")
|
||||
logger = logging.getLogger("measure_autocorrect_cost")
|
||||
logger.setLevel(logging.INFO)
|
||||
|
||||
PROJECT = "nomadkaraoke"
|
||||
GCS_BUCKET = "karaoke-gen-storage-nomadkaraoke"
|
||||
DEFAULT_MODELS = ["claude-fable-5", "gemini-3.1-pro-preview"]
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Job + corrections loading
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
def _corrections_path(job_data: dict, job_id: str) -> Optional[str]:
|
||||
"""Mirror review.py: prefer the latest review edits, then the original."""
|
||||
lyrics = (job_data.get("file_urls") or {}).get("lyrics") or {}
|
||||
return (
|
||||
lyrics.get("corrections_updated")
|
||||
or lyrics.get("corrections")
|
||||
or f"jobs/{job_id}/lyrics/corrections.json"
|
||||
)
|
||||
|
||||
|
||||
def fetch_recent_jobs(db: firestore.Client, limit: int) -> list[dict]:
|
||||
query = (
|
||||
db.collection("jobs")
|
||||
.where("status", "==", "complete")
|
||||
.order_by("created_at", direction=firestore.Query.DESCENDING)
|
||||
.limit(limit * 4)
|
||||
)
|
||||
jobs = []
|
||||
for doc in query.stream():
|
||||
data = doc.to_dict()
|
||||
data["_job_id"] = doc.id
|
||||
lyrics = (data.get("file_urls") or {}).get("lyrics") or {}
|
||||
if lyrics.get("corrections") or lyrics.get("corrections_updated"):
|
||||
jobs.append(data)
|
||||
if len(jobs) >= limit:
|
||||
break
|
||||
return jobs
|
||||
|
||||
|
||||
def fetch_job(db: firestore.Client, job_id: str) -> Optional[dict]:
|
||||
doc = db.collection("jobs").document(job_id).get()
|
||||
if not doc.exists:
|
||||
logger.warning("job %s not found in Firestore", job_id)
|
||||
return None
|
||||
data = doc.to_dict()
|
||||
data["_job_id"] = job_id
|
||||
return data
|
||||
|
||||
|
||||
def download_corrections(
|
||||
gcs: storage.Client, path: str
|
||||
) -> Optional[dict]:
|
||||
if path.startswith("gs://"):
|
||||
bucket_name, _, blob_path = path[5:].partition("/")
|
||||
else:
|
||||
bucket_name, blob_path = GCS_BUCKET, path
|
||||
try:
|
||||
blob = gcs.bucket(bucket_name).blob(blob_path)
|
||||
if not blob.exists():
|
||||
return None
|
||||
return json.loads(blob.download_as_bytes())
|
||||
except Exception as exc: # noqa: BLE001
|
||||
logger.warning("failed to read %s: %s", path, exc)
|
||||
return None
|
||||
|
||||
|
||||
def build_inputs(corrections: dict) -> Optional[tuple[list[dict], dict]]:
|
||||
"""Return (segments, reference_lyrics) in the shape suggest() expects."""
|
||||
segments = corrections.get("corrected_segments") or corrections.get("segments") or []
|
||||
reference_lyrics = corrections.get("reference_lyrics") or {}
|
||||
if not segments:
|
||||
logger.warning("no segments in corrections")
|
||||
return None
|
||||
if not reference_lyrics:
|
||||
logger.warning("no reference_lyrics in corrections — auto-correct needs >=1 source")
|
||||
return None
|
||||
# The service requires segment + word ids; older jobs may lack them.
|
||||
for seg in segments:
|
||||
if not seg.get("id") or any(not w.get("id") for w in (seg.get("words") or [])):
|
||||
logger.warning("segments missing word/segment ids — skipping job")
|
||||
return None
|
||||
return segments, reference_lyrics
|
||||
|
||||
|
||||
# ---------------------------------------------------------------------------
|
||||
# Measurement
|
||||
# ---------------------------------------------------------------------------
|
||||
|
||||
def measure_job(service, job: dict, gcs, models: list[str]) -> Optional[dict]:
|
||||
from backend.services.auto_correct.pricing import estimate_cost_usd
|
||||
from backend.services.auto_correct.prompts import (
|
||||
build_system_prompt,
|
||||
build_user_prompt,
|
||||
)
|
||||
from backend.services.auto_correct.settings import AutoCorrectSettings
|
||||
|
||||
job_id = job["_job_id"]
|
||||
artist = job.get("artist")
|
||||
title = job.get("title")
|
||||
path = _corrections_path(job, job_id)
|
||||
corrections = download_corrections(gcs, path)
|
||||
if not corrections:
|
||||
logger.warning("[%s] no corrections at %s — skipping", job_id, path)
|
||||
return None
|
||||
built = build_inputs(corrections)
|
||||
if not built:
|
||||
return None
|
||||
segments, reference_lyrics = built
|
||||
|
||||
settings = AutoCorrectSettings()
|
||||
system_prompt = build_system_prompt(settings)
|
||||
user_prompt = build_user_prompt(
|
||||
segments=segments,
|
||||
reference_lyrics=reference_lyrics,
|
||||
artist=artist,
|
||||
title=title,
|
||||
)
|
||||
|
||||
word_count = sum(len(seg.get("words") or []) for seg in segments)
|
||||
logger.info(
|
||||
"[%s] %s - %s | %d segments, %d words, %d ref sources",
|
||||
job_id, artist or "?", title or "?", len(segments), word_count,
|
||||
len(reference_lyrics),
|
||||
)
|
||||
|
||||
per_model = []
|
||||
for model in models:
|
||||
t0 = time.time()
|
||||
try:
|
||||
raw, usage = service._call_model(
|
||||
model, system_prompt, user_prompt, job_id=job_id
|
||||
)
|
||||
except Exception as exc: # noqa: BLE001
|
||||
logger.warning("[%s] model %s failed: %s", job_id, model, exc)
|
||||
per_model.append({"model": model, "error": str(exc)})
|
||||
continue
|
||||
elapsed = time.time() - t0
|
||||
cost = estimate_cost_usd(model, usage)
|
||||
n_sugg = len(raw.get("suggestions") or []) if isinstance(raw, dict) else 0
|
||||
per_model.append({
|
||||
"model": model,
|
||||
"usage": usage,
|
||||
"cost": cost,
|
||||
"suggestions": n_sugg,
|
||||
"elapsed": elapsed,
|
||||
})
|
||||
if usage:
|
||||
logger.info(
|
||||
" %-26s in=%-6d out=%-6d think=%-6s cost=%s (%.1fs, %d suggestions)",
|
||||
model,
|
||||
usage.get("input_tokens", 0),
|
||||
usage.get("output_tokens", 0),
|
||||
"na" if usage.get("thinking_tokens") is None else usage["thinking_tokens"],
|
||||
"?" if cost is None else f"${cost:.4f}",
|
||||
elapsed, n_sugg,
|
||||
)
|
||||
else:
|
||||
logger.info(" %-26s usage unavailable", model)
|
||||
|
||||
return {
|
||||
"job_id": job_id,
|
||||
"artist": artist,
|
||||
"title": title,
|
||||
"word_count": word_count,
|
||||
"per_model": per_model,
|
||||
}
|
||||
|
||||
|
||||
def print_report(results: list[dict], models: list[str]) -> None:
|
||||
print("\n" + "=" * 78)
|
||||
print("AUTO-CORRECT COST MEASUREMENT")
|
||||
print("=" * 78)
|
||||
track_totals = []
|
||||
model_cost_sums = {m: 0.0 for m in models}
|
||||
model_cost_known = {m: False for m in models}
|
||||
|
||||
for r in results:
|
||||
track_total = 0.0
|
||||
any_cost = False
|
||||
print(f"\n{r['artist'] or '?'} - {r['title'] or '?'} "
|
||||
f"({r['job_id']}, {r['word_count']} words)")
|
||||
for pm in r["per_model"]:
|
||||
if "error" in pm:
|
||||
print(f" {pm['model']:<26s} ERROR: {pm['error']}")
|
||||
continue
|
||||
u = pm["usage"] or {}
|
||||
cost = pm["cost"]
|
||||
if cost is not None:
|
||||
track_total += cost
|
||||
any_cost = True
|
||||
model_cost_sums[pm["model"]] += cost
|
||||
model_cost_known[pm["model"]] = True
|
||||
print(
|
||||
f" {pm['model']:<26s} "
|
||||
f"in={u.get('input_tokens', 0):<6d} "
|
||||
f"out={u.get('output_tokens', 0):<6d} "
|
||||
f"think={('na' if u.get('thinking_tokens') is None else u['thinking_tokens'])!s:<6s} "
|
||||
f"cost={'?' if cost is None else f'${cost:.4f}':<9s} "
|
||||
f"({pm['suggestions']} suggestions)"
|
||||
)
|
||||
print(f" {'TRACK TOTAL':<26s} {'':<26s}"
|
||||
f"{'$%.4f' % track_total if any_cost else 'unknown'}")
|
||||
if any_cost:
|
||||
track_totals.append(track_total)
|
||||
|
||||
print("\n" + "-" * 78)
|
||||
print("AGGREGATE")
|
||||
print("-" * 78)
|
||||
for m in models:
|
||||
if model_cost_known[m]:
|
||||
avg = model_cost_sums[m] / len(results) if results else 0.0
|
||||
print(f" {m:<26s} avg ${avg:.4f}/track")
|
||||
if track_totals:
|
||||
avg = sum(track_totals) / len(track_totals)
|
||||
hi = max(track_totals)
|
||||
print(f"\n Avg total cost/track: ${avg:.4f}")
|
||||
print(f" Max total cost/track: ${hi:.4f}")
|
||||
print(f" Jobs measured: {len(track_totals)}")
|
||||
print(f"\n Projected at 30 jobs/day: ${avg * 30:.2f}/day (~${avg * 30 * 30:.0f}/mo)")
|
||||
print(f" Projected at 100 jobs/day: ${avg * 100:.2f}/day (~${avg * 100 * 30:.0f}/mo)")
|
||||
print("\nNOTE: Gemini USD uses ESTIMATED rates (see pricing.py). Token counts are exact.")
|
||||
print("=" * 78)
|
||||
|
||||
|
||||
def main() -> int:
|
||||
parser = argparse.ArgumentParser(description=__doc__)
|
||||
parser.add_argument("--jobs", nargs="+", help="Specific job IDs to measure")
|
||||
parser.add_argument("--limit", type=int, default=3,
|
||||
help="Auto-pick N recent complete jobs (ignored if --jobs given)")
|
||||
parser.add_argument("--models", nargs="+", default=DEFAULT_MODELS,
|
||||
help=f"Models to measure (default: {' '.join(DEFAULT_MODELS)})")
|
||||
args = parser.parse_args()
|
||||
|
||||
if any(m.startswith("claude") for m in args.models) and not os.environ.get("ANTHROPIC_API_KEY"):
|
||||
logger.error("ANTHROPIC_API_KEY is not set — required for claude-* models")
|
||||
return 2
|
||||
|
||||
from backend.services.auto_correct.service import AutoCorrectService
|
||||
|
||||
db = firestore.Client(project=PROJECT)
|
||||
gcs = storage.Client(project=PROJECT)
|
||||
service = AutoCorrectService()
|
||||
|
||||
if args.jobs:
|
||||
jobs = [j for j in (fetch_job(db, jid) for jid in args.jobs) if j]
|
||||
else:
|
||||
jobs = fetch_recent_jobs(db, args.limit)
|
||||
logger.info("auto-picked %d recent complete jobs with corrections", len(jobs))
|
||||
|
||||
if not jobs:
|
||||
logger.error("no jobs to measure")
|
||||
return 1
|
||||
|
||||
results = []
|
||||
for job in jobs:
|
||||
r = measure_job(service, job, gcs, args.models)
|
||||
if r:
|
||||
results.append(r)
|
||||
|
||||
if not results:
|
||||
logger.error("no jobs produced measurable results")
|
||||
return 1
|
||||
|
||||
print_report(results, args.models)
|
||||
return 0
|
||||
|
||||
|
||||
if __name__ == "__main__":
|
||||
sys.exit(main())
|
||||
Reference in new issue
Block a user