Ai model ekak

[baba-chuti]

mchanla mata oni audio ekak wachana ain karnna ex: fuck vage explicit words AI use karla. prompt eka type karama wachna ai eken ain karala output enna. ubala danna ai models or sites tiyei da. mama danata dakka ekama site eka https://songcleaner.com/ but meke editing quality madi. dawskata 2i free tiyenne. @Nidarshana_k @රෝසි ආච්චි @infernocus @olu bakka

 

[Heshan Daminda]

ethakota ethana wachane wenuwata mkkda danne?

 

[Yeezus]

mchanla mata oni audio ekak wachana ain karnna ex: fuck vage explicit words AI use karla. prompt eka type karama wachna ai eken ain karala output enna. ubala danna ai models or sites tiyei da. mama danata dakka ekama site eka https://songcleaner.com/ but meke editing quality madi. dawskata 2i free tiyenne. @Nidarshana_k @රෝසි ආච්චි @infernocus @olu bakka

Gemini 3.7 flash use කරලා tool එකක් හදාගන්න.
මේ අවුරුද්ද ඉවර වෙනකල් 50% discount එකක් යනවා token usage.

Open router ගත්තොත් තව 50% ආසන්න ගානක් අඩු වෙනවා ටිකක්.බලන්න. model එක release කරෙත් ඊයේ.

 

[baba-chuti]

ethakota ethana wachane wenuwata mkkda danne?

e kalla auto mute wenawa background music play wenne. dj karna kattiya num podi shift karnawa first letter eke prounce karnawa ekkama. uba english song wala clean version ahala tiyei da meeta kalin? anna e vage pattern ekak.

Gemini 3.7 flash use කරලා tool එකක් හදාගන්න.
මේ අවුරුද්ද ඉවර වෙනකල් 50% discount එකක් යනවා token usage.

Open router ගත්තොත් තව 50% ආසන්න ගානක් අඩු වෙනවා ටිකක්.බලන්න. model එක release කරෙත් ඊයේ.

Gemini 3.7 flash use කරලා tool එකක් හදාගන්න. -- ma gawa antigravity eka tiyei. meka try kala hari yanne na machan. songcleaner ugema model ekak tiyenne man hithanne. mata oni thridpary ekak

 

[හෙළයෙක්]

ethakota ethana wachane wenuwata mkkda danne?

"ඊය" කියලා දැම්ම නන් හරි.

 

[helplesser]

from faster_whisper import WhisperModel
from pydub import AudioSegment
from pydub.generators import Sine

TARGETS = {"damn", "hell"}
PAD_MS = 60  # timestamps are slightly loose; pad both sides

model = WhisperModel("small", compute_type="int8")
segments, _ = model.transcribe("input.mp3", word_timestamps=True)
audio = AudioSegment.from_file("input.mp3")

spans = [
    (int(w.start * 1000) - PAD_MS, int(w.end * 1000) + PAD_MS)
    for seg in segments for w in seg.words
    if w.word.strip().lower().strip(".,!?'\"") in TARGETS
]

for start, end in spans:
    start, end = max(0, start), min(len(audio), end)
    patch = Sine(1000).to_audio_segment(duration=end - start).apply_gain(-12)
    # or: patch = AudioSegment.silent(duration=end - start)
    audio = audio[:start] + patch + audio[end:]

audio.export("output.mp3", format="mp3")

 

[baba-chuti]

from faster_whisper import WhisperModel
from pydub import AudioSegment
from pydub.generators import Sine

TARGETS = {"damn", "hell"}
PAD_MS = 60  # timestamps are slightly loose; pad both sides

model = WhisperModel("small", compute_type="int8")
segments, _ = model.transcribe("input.mp3", word_timestamps=True)
audio = AudioSegment.from_file("input.mp3")

spans = [
    (int(w.start * 1000) - PAD_MS, int(w.end * 1000) + PAD_MS)
    for seg in segments for w in seg.words
    if w.word.strip().lower().strip(".,!?'\"") in TARGETS
]

for start, end in spans:
    start, end = max(0, start), min(len(audio), end)
    patch = Sine(1000).to_audio_segment(duration=end - start).apply_gain(-12)
    # or: patch = AudioSegment.silent(duration=end - start)
    audio = audio[:start] + patch + audio[end:]

audio.export("output.mp3", format="mp3")

try ekak deela balnnum.

 

[helplesser]

try ekak deela balnnum.

ela ela.
claude eken gatte.
oke audio eke length eka wenas wenne nalu. wachane play wena tika silent wena ekalu wenne.

 

[Honey Bunch]

Bump machan

 

[Heshan Daminda]

ඕක මම කරන්නේ virtual DJ එකෙන්,. එතනට break effect එකක් දානවා.

 

[Nidarshana_k]

Transcribe karala gnna puluwan unoth nmN8n /Make.com use karala LLM ekakta dala Explicit words ain karala output eka ganna puluwan wei. JSON widiyata aran site ekata daganna puluwan

 

[olu bakka]

ඕකට ලේසිම වැඩේ මුලින්ම single word වලට timestamps එක්ක transcribe කරපන් audio එක. ඊට පස්සෙ ඉතින් මහ දෙයක් නෑ අදාල වචනෙ තියෙන timestamp එක දන්නවනෙ. එතනට mute effect එකක් දාගන්න program කරගනින්.