text-vllm.mdx

export const TextVllm = ({ modelId, samplingParams, maxTokens }) => (

Install:

{`pip install vllm==0.14`.split('\n').map((line, i) => {line}{'\n'})}

Run:

{`from vllm import LLM, SamplingParams
llm = LLM(model="${modelId}")
sampling_params = SamplingParams(${samplingParams}max_tokens=${maxTokens || 512})
output = llm.chat("What is machine learning?", sampling_params)
print(output[0].outputs[0].text)`.split('\n').map((line, i) => {line}{'\n'})}

);