Create speaker labeled transcripts with Google Colab!

Create speaker labeled transcripts with Google Colab! | Deepgram Studios

Рет қаралды 1,770

Күн бұрын

Sign up for Deepgram, and receive up to 45,000 minutes of free transcription. No credit card needed: dpgr.am/youtube
----------------------------------
Resources!
Link to notebook: colab.research.google.com/dri...
See the notebook on Github: github.com/deepgram-devs/diar...
Deepgram Nova video: dpgr.am/b775182
Demos: dpgr.am/853cef6
--------------------------------
Try Deepgram without having to write any code: dpgr.am/23552b6
Follow us on Twitter: / deepgramai
Follow us on LinkedIn: / deepgram
Follow us on TikTok: www.tiktok.com/@deepgram_ai
Check out our Github: github.com/deepgram
---------------------------
Chapters:
0:00 Intro
0:19 Speaker labeling explained
0:52 The Python Notebook explained
1:16 The 1st notebook cell
1:24 The 2nd notebook cell
1:31 The 3rd notebook cell (and vars to modify)
2:31 The final cell and your speaker-labeled transcript
2:45 Example results
3:37 Get creative!
4:21 Outro

Пікірлер: 9

@camelendezl 5 ай бұрын

Will there be a new notebook for the SDK 3? It might be my lack of experience with Deepgram, but it doesn't seem to be working with the current version of SDK :(

@arturgomes1654 7 ай бұрын

Excellent video! I'm using deepgram for some projects and it works much better than the OpenAI API. Regarding diarization, is there any way to associate each speaker with a name? For example: Speaker 0 : John, Speaker 1 : Mary.

@Seedley 7 ай бұрын

Would you make a video on how to train a model for speaker dirarization

@fabcotechnologies489 Жыл бұрын

nova video is not available (link in description)

@Deepgram Жыл бұрын

Fixed! Thanks for pointing that out :D Here's another link for convenience: kzfaq.info/get/bejne/gcmPn7d8vbmneas.html

@kilexmo Жыл бұрын

It doesn't work for another language. When generating the transcription, the txt appears blank

@Deepgram Жыл бұрын

The .txt usually appears blank if the transcription cell was run before the audio file is finished uploading! This might be a race condition issue. You can also specify the language you're using in the config. Doing so improves the accuracy of the output! Check out this documentation to see how: developers.deepgram.com/docs/language? And, of course, if you wish to try out the API without having to code anything, you can use this playground: playground.deepgram.com/? And, of course, if you have any further issues or if you'd like us to help debug, there's a feedback form at the bottom of the playground page where we can continue speaking! (Or you could also DM us on Twitter, link in description 😄)

@Raystopmotions 8 ай бұрын

@@Deepgramcan you make a tutorial on how to config the language specifically using it for google collab?