Credit: (V)(;,,,;)(V)

Step 1
In the folder with the ”Run” file used to start the server, there’s a folder labeled “Tools”.

Open that folder, and inside there’s several other files. Find “enable_local_whisper”.

Run it, and a command window will open that looks like this:

Type “yes”, and you’ll be prompted to pick the model you want to use. The synopsis for each model is in the command window.
Pick which one you want to use. The command window will close. Give it about 5 to 10 minutes to complete its installation.
After waiting, run the server file, and you should see something like this:

If you do, then you can skip to Step 4.
If you don’t, then continue on to Step 2.

Step 2
In file explorer, go to:
\\wsl.localhost\DwemerAI4Skyrim2\home\dwemer\speech-rest-api
In there, there’s three files that, depending on which model you chose, need to be edited.

The numbers above indicate which model you chose. For the purposes of this guide, we’ll be editing medium_model_en.
Open it in a text editor (preferably Notepad++) and find the following line:

(If editing the other models, the corresponding model lines will say “large-v2” for the large model, and “medium” for the regular medium model).
Change “float16” to “int8”.

Save, then exit the text editor.

(This is also where you can change it from using your gpu to instead use your cpu (performance will vary). Just change “cuda” to “cpu”, then save and exit.)
Redo Step 1, and it should say that the Local Whisper Server is started, and skip to Step 4.
If it still shows that it isn’t started, continue to Step 3.

Step 3
If the server command window still says that it hasn’t started, then it could just be the trigger not communicating properly.
Go to the same folder with the models and find the “log” file. Open it in a text editor.

If it shows this:

Then it’s working fine, and the server command window is just having a bit of a hiccup. You can continue to Step 4.

Step 4
Go to the server webpage that opens whenever running the server.
Along the top, “Configuration”, then “Configuration Wizard”.

Along the right side in the Quick Access menu click on “Speech-to-Text Service”.

This will auto-scroll down to the Speech-to-Text section.
Make sure that the “STTFUNCTION” is set to “localwhisper”.

Scroll down to the “Local Whisper Speech-to-Text” section.

Change the “127.0.0.1:3000” to what the server command window tells you, leaving the /transcribe”.

Click on “Check” and a small window should pop up at the top saying the URL is ok and available.

Scroll down to the bottom of the page (or in the “Quick Access” menu) and hit “Save”.
Step 5
In a file explorer, go to:
\\wsl.localhost\DwemerAI4Skyrim2\var\www\html\HerikaServer\stt

Open “stt-localwhisper” in a text editor and find this line:

Change “file\” to “audio\”.

Save, then exit the text editor.

Now it should work in game. Just make sure that you’re using the “record voice” hotkey, not the “chat with Herika” hotkey.