This is probably a dumb question, but would this be possible with tinyML and something like an esp32? I wouldn't need the super low latency, just the transcription on device.
@@finnless2540 no question is dumb, and I’m glad you asked. I’m afraid the answer is no. The model weights alone would not fit in the ESP32 memory, not even with an external memory chip.
So interesting
technology is awesomeeeeeeeeeeeeeeeeeeee
Super interesting! 🎉 just subscribed ❤
I'm wondering what computing device u r using.
@@emranio Thank you! It’s a Jetson Nano with 4GB.
This is probably a dumb question, but would this be possible with tinyML and something like an esp32? I wouldn't need the super low latency, just the transcription on device.
Too weak, you need something a bit powerful to run Whisper
@@finnless2540 no question is dumb, and I’m glad you asked. I’m afraid the answer is no. The model weights alone would not fit in the ESP32 memory, not even with an external memory chip.