Create Descript-like word scrolling subtitles for free using Python, Moviepy and OpenAI's Whisper

[amalyszek]

Great, work man. Thank you for your explanation. All the best to you!

[jackmartin]

Hi @Ramsri, great video. Would love to see any tutorial/resources on "Advanced Prompting" and "Prompt Design" covering techniques like Chain of Thought, Few Shot Examples using Vector DB etc.

[TalKatz-if]

Thanks @Ramsri, this is an awesome video!

I have a question for you,
I already have the script to my video and srt file, I just need to get the word-level timestamps, can I get it without doing transcribe? because Whisper speech recognition will change some of my words, and again, I already have the script of the video written.

[zero-code]

Great video Ram. Do you deploy these as Google Cloud Functions? Is it possible?

[Rapha_Carpio]

You build all those apps you mentioned?

[shortsfusion_]

Please is possible to use in local pc?

[skilledprogrammer]

please Sir make a video in which use Facebook's wav2vec2 model instead of open whisper. And we have option for user to upload video and user can change the font and color as per his wish.

[peppibua]

Does not work with any uploaded file or file on Google Drive!
Step 1 -> import os does work and showes me the right name:

import os

mp4videoURL = "/content/01-02. Supplies Needed for This Course (Video).mp4"
videofilename = mp4videoURL.split('/')[-1]
print (videofilename)
01-02. Supplies Needed for This Course (Video).mp4
time: 766 µs (started: 2023-05-26 20:09:54 +00:00)

But the next step generates the following error message:

import urllib.request
urllib.request.urlretrieve(mp4videoURL, videofilename)

ValueError Traceback (most recent call last)
in <cell line: 2>()
1 import urllib.request
----> 2 urllib.request.urlretrieve(mp4videoURL, videofilename)

5 frames
in urlretrieve(url, filename, reporthook, data)
239 url_type, path = _splittype(url)
240
--> 241 with contextlib.closing(urlopen(url, data)) as fp:
242 headers = fp.info()
243

in urlopen(url, data, timeout, cafile, capath, cadefault, context)
214 else:
215 opener = _opener
--> 216 return opener.open(url, data, timeout)
217
218 def install_opener(opener):

in open(self, fullurl, data, timeout)
501 # accept a URL or a Request object
502 if isinstance(fullurl, str):
--> 503 req = Request(fullurl, data)
504 else:
505 req = fullurl

in __init__(self, url, data, headers, origin_req_host, unverifiable, method)
320 origin_req_host=None, unverifiable=False,
321 method=None):
--> 322 self.full_url = url
323 self.headers = {}
324 self.unredirected_hdrs = {}

in full_url(self, url)
346 self._full_url = unwrap(url)
347 self._full_url, self.fragment = _splittag(self._full_url)
--> 348 self._parse()
349
350 @full_url.deleter

in _parse(self)
375 self.type, rest = _splittype(self._full_url)
376 if self.type is None:
--> 377 raise ValueError("unknown url type: %r" % self.full_url)
378 self.host, self.selector = _splithost(rest)
379 if self.host:

ValueError: unknown url type: '/content/01-02. Supplies Needed for This Course (Video).mp4'
time: 134 ms (started: 2023-05-26 20:10:02 +00:00)

And the most important of all! Where is the subtitle file? I can not do anything with the 1080x1080 square video!