update

distributed/rpc/parameter_server/README.md

@@ -5,4 +5,6 @@ This is a basic example of RPC-based training that uses several trainers remotel
 To run the example locally, run the following command worker for the server and each worker you wish to spawn, in separate terminal windows:
 `python rpc_parameter_server.py [world_size] [rank] [num_gpus]`. For example, for a master node with world size of 2, the command would be `python rpc_parameter_server.py 2 0 0`. The trainer can then be launched with the command `python rpc_parameter_server.py 2 1 0` in a separate window, and this will begin training with one server and a single trainer.
 
+Note that for demonstration purposes, this example supports only between 0-2 GPUs, although the general pattern can be extended to make use of all GPUs available on a node.
+
 You can pass in the command line arguments `--master_addr=<address>` and `master_port=PORT` to indicate the address:port that the master worker is listening on. All workers will contact the master for rendezvous during worker discovery. By default, `master_addr` will be `localhost` and `master_port` will be 29500.

distributed/rpc/parameter_server/rpc_parameter_server.py

@@ -262,6 +262,7 @@ def run_worker(rank, world_size, num_gpus, train_loader, test_loader):
 
     args = parser.parse_args()
     assert args.rank is not None, "must provide rank argument."
+    assert args.num_gpus <= 3, f"Only 0-2 GPUs currently supported (got {args.num_gpus})."
     os.environ['MASTER_ADDR'] = args.master_addr
     os.environ["MASTER_PORT"] = args.master_port
     processes = []