ray.runtime_context.RuntimeContext.get_resource_ids

RuntimeContext.get_resource_ids() → Dict[str, List[str]]

Get the current worker’s GPU, accelerator and TPU ids.

Returns

A dictionary keyed by the resource name. The values are lists of ids. .. rubric:: Example

{‘GPU’: [‘0’, ‘1’], ‘neuron_cores’: [‘0’, ‘1’], ‘TPU’: [‘0’, ‘1’]}.