3 ms·
You should be able to first train it on generic text once, then duplicate the input layer and fine-tune on conversation.
by vanviegen 6mo ago
You should be able to first train it on generic text once, then duplicate the input layer and fine-tune on conversation.