My experience as a new working student at Axelera#1

Nice work ​@Abdi ! Whisper is huge - that’s where the natural language interaction begins, and that’s where AI really integrates with humanity! Being able to just chat with each other. Mind blowing stuff that we’re already starting to take for granted.

Would love to see your progress or experimentation with it - feel free to share it!

Also, you get bonus points for your profile image! 🖖 (Did you watch Starfleet Academy yet? I’m keeping my powder dry for now, but so far I’m a little underwhelmed… especially after how awesome SNW turned out to be!)

Thanks for the kinds words ​@Spanner 

Fellow work student ​@Pablo Herrero made a neat bash script that get this task done

#!/bin/bash

WHISPER_DIR=~/whisper.cpp
VOYAGER_DIR=~/voyager-sdk
AUDIO_FILE=$1

# Store transcription
echo "Whisper Transcription:"
TRANSCRIPT=$($WHISPER_DIR/build/bin/whisper-cli -m $WHISPER_DIR/models/ggml-base.en.bin -f "$AUDIO_FILE" -nt -np)

echo "$TRANSCRIPT"
echo ""

# Feed to Llama inside its environment
echo "Llama Response:"
cd $VOYAGER_DIR
source venv/bin/activate
axllm llama-3-2-1b-1024-4core-static --prompt "$TRANSCRIPT" | sed -n '/Assistant:/,$p'

This script assumes: You have whisper installed and you are running voyager-SDK 1.5

#trek stuff
I haven’t watched star fleet academy yet ,I’ll keep my expectations to a minimum. 
I did hear good words from SNW ,I may give it a shot!
 

Nice! So this let’s Voyager run Whisper locally?

Yeah.  Two stages process: the script sends the audio files to whisper
whisper turns into text , the script passes the text to voyager-SDK which runs the LLM.