How to generate a fix amount of tokens when calling the model

[JeremyGe07]

I have read the doc but sitll cannot find how to generate a fix amount of tokens in call function. Lets say, e.g., how to gurantee to generate 50 tokens in the following code?

model = llama_cpp.Llama(model_path="path/to/model")
print(model("The quick brown fox jumps ")["choices"][0]["text"])

In llama.cpp ，we can use the -n(--n-predict) parameter to set the number of tokens to predict when generating text, and use the --ignore-eos parameter to keep generating until the set nubmer length as this doc said.
Thanks for your help in advance.