Hi,

Currently streaming with the python bindings for GptManager doesn't work as there is a bug in the python binding for InferenceRequest: toTrtLlm() is missing copying the streaming property from the binding object to the native InferenceRequest object.