resp = client.inference.put_model(
task_type="text_embedding",
inference_id="my-e5-model",
body={
"service": "elasticsearch",
"service_settings": {
"num_allocations": 1,
"num_threads": 1,
"model_id": ".multilingual-e5-small",
},
},
)
print(resp)