Base model: gpt-2 large
Epoch: 20
Loss: 0.15
Dataset:Sherlock (https://www.kaggle.com/idevji1/sherlock-holmes-stories)
License:CC0: Public Domain

this model is based on gpt2-large and fine tuned with sherlock script dataset.
it generates sherlock scrip related to your input keyword.
first, input your keyword
second, wait for result

three ways to use Gpt2 sherlock
	1.  CLI
	2.  Swagger
	3.  Demo

keyword: input your keyword than you will check out the sherlock script related to keyword

generated text

curl -X GET "https://master-gpt2-sherlock-ha-mulan.endpoint.ainize.ai/api/?keyword=murder" -H "accept: string"

    murderous outbursts and it seemed to me that
    to speak at all was very painful to him and that his will all
    through was overriding his inclinations

    it is a very delicate thing said he one does not like to speak of
    ones domestic affairs to strangers it seems dreadful to discuss the
    conduct of ones wife with two men whom i have never seen before
    it is horrible to have to do it but i have got to the end of my tether
    and i must have advice

    my dear mr grant munro began holmes

    our visitor sprang from his chair what he cried you know my
    name

    if you wish to preserve your incognito said holmes smiling i suggest that you cease to
    write your name upon the lining of your hat or else that you turn the crown
    towards the person whom you are addressing i was about to say that my friend and i
    have been expecting you come round for some months
    therefore i had no difficulty in finding you

    the doctor flushed with anger

    i have been expecting you for some time said he but you must have forgotten my advice or i was not to