How to generate predictions from this model?
#3
by vivasvan100 - opened
sentence = tokenizer("god bless you",return_tensors="pt")
outputs = model(**inputs)
print(outputs.logits)
tensor([[ 1.9721, -1.8041]], grad_fn=<AddmmBackward0>)
is this how to generate the outputs?
inputs = tokenizer.encode("god bless you", return_tensors="pt")
outputs = model(inputs)[0]
probabilities = torch.softmax(outputs, dim=1)
predicted_class = torch.argmax(probabilities).item()
toxic_probability = round(probabilities[0][1].item() * 100, 2)
pc = ""
if predicted_class == 1:
pc = "Toxic"
else:
pc = "Not toxic"