I'm noticing a rather tight confidence score on most of these. Methinkst the model needs more tuning
Login to reply
Replies (2)
high confidence with google's model, low confidence with Qwen run locally