loadModel
Loads model into memory ahead of its first request, with POST /v1/models/{id}/load.
Returns when the model is in memory. Loading a large model from a cold disk takes a while; the gateway timeout covers it. A chat request loads its model automatically; this is for warming up. oMLX has no pull operation: it downloads models only through its admin dashboard.
Throws
if model is blank, before any request.
on a non-2xx response, for example an unknown model.