You can not select more than 25 topics Topics must start with a letter or number, can include dashes ('-') and can be up to 35 characters long.
Soothspider 67224853c1 AI Captions for Dec. 19 11 months ago
guest.speaking AI Captions for CHD 2023-09-22 talk 11 months ago
tools Fix README generator. Remove trailing / and extra -. 11 months ago
twitch AI Captions for Dec. 19 11 months ago
youtube Folder structure and readmes. 12 months ago
LICENSE Initial commit 1 year ago
README.md guest.speaking link 11 months ago

README.md

Stream.Transcripts

Gigaohm Biological High Resistance Low Noise Information Transcripts

The purpose of this repo is to place AI generate captions in one location for easy searching. How this repo will change will depend on how well it is working for our purposes.

Note: Due to when this repo got created, it would not be possible to generate all the captions for all the streams (at least initially). JC has his own archive of all the streams, so this is possible, but it is a matter of time and effort.

Organization

All transcripts will be

Other folders:

  • tools - tools and scripts

Conventions

In each transcript "top" folder, will be a new folder for each stream. These folders will typically start with the stream ID and include the title and date of ths stream. Inside each folder will be the transcript (probably in VTT format). This provides some flexibility in case we need to put additional files there. It will also have a ReadMe.md file which should include some information like a link back to the known streams.

Plans

Initially this will server as a repository for AI generated (and corrected) transcripts from any streams that Gigaohm Biological appears for the purpose of searching. However, over time, we might develop new capabilities.

Javascript + Static Site

One idea is to generate (using github pages) a static site using javascript and pre-built indexes to make the searching a lot better/easier (including hot-linking to the right timestamp from either the Twitch or alternative sourced stream). This would make finding the content that one wants to share much faster.

(This is a far-off-in-the-future plan...)