Logits not between 0 and 1

#1
by JanMack - opened

Hi,
I tested the model and found that the output of the model for each property is not between 0 and 1.

I used the test code with the following results:

from transformers import AutoTokenizer, AutoModelForSequenceClassification

model = AutoModelForSequenceClassification.from_pretrained("vladinc/bigfive-regression-model")
tokenizer = AutoTokenizer.from_pretrained("vladinc/bigfive-regression-model")

text = "I enjoy reflecting on abstract concepts and trying new things."
inputs = tokenizer(text, return_tensors="pt")
outputs = model(**inputs)

print(outputs.logits) # Output: tensor([[4.3761, 3.5514, 3.5852, 3.7274, 2.3770]], grad_fn=)

Sign up or log in to comment