Add subtitle dictionaries, Polish number normalizer, ComfyUI lifecycle control
Text processing before TTS:
- dictionary.txt rewrites words and phrases (km/h, np., units). Rules compile
into one alternation applied in a single pass, so replacements never chain
and longest-literal wins. Case-sensitive, whole-word.
- A per-queue second dictionary, written by the GUI into the first video's job
folder and merged with the shared one (--dictionary-extra).
- ALL-CAPS blocks become Title Case rather than sentence case: lowercasing
destroyed proper nouns inside captions ("NASA - MARGO MADISON") and invited
the normalizer to "correct" the name. Paired clips in output/clips/all-caps
showed the TTS reads both casings the same.
normalizer/ — digits to spoken Polish, in its own venv because it needs
transformers>=5, which must not touch ComfyUI's pinned versions:
- Two backends. Plain integers go to num2words for the value and a general
Polish LLM for the case; times, dates, decimals and identifiers go to a 0.6B
fine-tune. Measured on 40 real subtitle lines, each is clearly better than
the other on its half. 74% of numbered blocks take the first leg.
- Guards on every answer: runaway length, repetition loops, invalid bytes,
leftover digits, a lost minus sign, a changed number, a lost word. A failure
falls back to num2words, which is nominative but never wrong about the value.
- ComfyUI is asked to unload before normalizing; a normalizer that does not fit
in VRAM is spilled to RAM silently at 5x the cost.
Backend lifecycle:
- ComfyUI is found by port through /proc, so a backend started by hand is
managed too. SIGTERM escalates to SIGKILL, and a suspended process is killed
outright since it never runs its handlers.
- A clip that kills the backend is retried three times, each with a different
seed (the aborts are seed-dependent, not text-dependent), then skipped and
reported so an overnight queue keeps moving.
- start_comfyui.sh execs python, so the script's PID is ComfyUI's.
- GUI gains ComfyUI and Queue dictionary buttons; quitting stops the backend.
--resume-from now reuses the job folder's tts_input.srt instead of reparsing
the subtitles, so block N still pairs with clip_NNNN.flac.
Co-Authored-By: Claude Opus 5 <noreply@anthropic.com>
This commit is contained in:
@@ -0,0 +1,49 @@
|
||||
# lektor_ai_comfy — replacement dictionary
|
||||
#
|
||||
# Words and phrases rewritten in the subtitles before they are sent to the TTS
|
||||
# engine, for anything the model reads out wrong: abbreviations, units, symbols,
|
||||
# names spelled in a way it mispronounces.
|
||||
#
|
||||
# One rule per line:
|
||||
#
|
||||
# pattern = replacement
|
||||
#
|
||||
# * Matching is CASE-SENSITIVE and matches whole words only: "np." leaves
|
||||
# "znp." alone, and "Joshua" leaves "joshua" alone.
|
||||
# * The replacement is inserted exactly as written — nothing is re-capitalised
|
||||
# for you. Note that the first letter of every subtitle block is already
|
||||
# capitalised before these rules run, so if a word can start a block, add a
|
||||
# second rule for the capitalised form.
|
||||
# * Longer patterns win: "km/h" below is used even though a plain "km" rule
|
||||
# would also fit, whatever order the two are written in.
|
||||
# * The replacement is never re-scanned, so rules cannot feed into each other.
|
||||
# * A pattern wrapped in slashes is a regular expression, and \1 \2 … in the
|
||||
# replacement refer to its groups.
|
||||
# * A line starting with # or ; is a comment. An empty replacement deletes the
|
||||
# phrase.
|
||||
# * "->" and "=>" work as separators too, if you prefer them to "=".
|
||||
#
|
||||
# The changes are visible in clips/tts_input.srt after a run — check there if a
|
||||
# rule is not doing what you expect.
|
||||
|
||||
|
||||
# ── Units ──────────────────────────────────────────────────────────────────
|
||||
km/h = kilometrów na godzinę
|
||||
m/s = metrów na sekundę
|
||||
|
||||
# ── Abbreviations ──────────────────────────────────────────────────────────
|
||||
np. = na przykład
|
||||
tzn. = to znaczy
|
||||
# These two almost always close a sentence, so the dot is kept; drop it if you
|
||||
# see an unwanted pause mid-sentence.
|
||||
itp. = i tym podobne.
|
||||
itd. = i tak dalej.
|
||||
m.in. = między innymi
|
||||
24/7 = dwadzieścia cztery na siedem
|
||||
|
||||
# ── Symbols ────────────────────────────────────────────────────────────────
|
||||
/(\d+)\s*%/ = \1 procent
|
||||
/(\d+)\s*°C/ = \1 stopni Celsjusza
|
||||
& = i
|
||||
|
||||
# ── Your own rules below ───────────────────────────────────────────────────
|
||||
Reference in New Issue
Block a user