4 ms·
yes. Finetuning a whisper model on a RPi 5 is ~2x faster than on the RPi 4. Other stages involving data pre-processing with HF dataset is again 2x-3x faster.
by jafermarq 3y ago
yes. Finetuning a whisper model on a RPi 5 is ~2x faster than on the RPi 4. Other stages involving data pre-processing with HF dataset is again 2x-3x faster.