How to run Text Generation model (GPT2) on Transformers-cli serve?

[nandakishorej8]

Hi,
I’m trying to run a text-generation model GPT2 on transformer-cli serve. I used below code
Creating API :
transformers-cli serve --task=text-generation --model=gpt2

Calling API:
curl -X POST http://localhost:8888/forward -H “accept: application/json” -H “Content-Type: application/json” -d ‘{“inputs”:“What is Deep Learning”,“parameters”:{“max_new_tokens”:20}}’

But I’m getting below error in the response. Looks like the API is expecting some specific input format.
{“detail”:[{“type”:“json_invalid”,“loc”:[“body”,0],“msg”:“JSON decode error”,“input”:{},“ctx”:{“error”:“Expecting value”}}]}

Let me know what is the write format to call the text-generation API