Labels in the output
What do LABELS in the output correspond to ?
@anirudhpm Can you give more clarity. Like in the Example for "I love you. I like you." I am getting scores for a list of Labels. What is there significance in layman terms
Hi @seigishi .
The huggingface prediction function basically returns the last layer of the models which are logits. These are then converted to probabilities by going through a SoftMax Layer. (all 🤗 Transformers models output the logits, as the loss function for training will generally fuse the last activation function, such as SoftMax, with the actual loss function, such as cross entropy).
Note that this all happens in the background. What you see are the probability scores that your sentence/paragraph belongs to a certain label. These scores are always between 0 and 1 and add up to 1.
The higher the score, the more confident the model is that your text belongs to a certain label. For instance, your example: "I love you. I like you." has little to do with any of the Sustainable Development Goals, hence the model returns low probabilities for the labels.