Predicting a missing word in a sentence
How do I use bloom model to predict a missing word in the middle of a sentence? I'd like as answer a list of possible words and a metric representing the goodness of the replacement, such as a probability.
Consider this instance:
- I took my ____ for a walk.
I would like an answer like this:
- Dog - 0.92
- Cat - 0.06
- Horse - 0.02
Thanks
Try using the text in between the underscore as a prompt.
Sample sentence: "I took my ____ for a walk."
List of possible words and a percentage representing the goodness of the replacement:
1.
I tried it and it gave back this:
Sample sentence: "I took my ____ for a walk."
List of possible words and a percentage representing the goodness of the replacement:
- Child 75%
- Kid 65%
- Boy 50%
- Son 25%
Hope this is what you were looking for.
You could use BLOOMZ
and e.g. given
Sample sentence: "I took my ____ for a walk."
What does ____ most likely to stand for?
it generates dog.
If you want probabilities, you can just look at the raw outputs that you get when using the model in code with one of the provided code snippets on the BLOOMZ model card.
Closing as this seems resolved.