Monarch

https://meta-pytorch.org/monarch/

Admin

Monarch is a distributed programming framework for PyTorch based on scalable actor messaging,...

Tokens:318

Snippets:5

Trust Score:6.4

Update:6 months ago

Show doc for...

Tokens:

Raw

### Spawn Processes and Actors in Monarch

Source: https://meta-pytorch.org/monarch/index.html

This example demonstrates how to spawn multiple trainer processes, define a Trainer actor, and then spawn instances of this actor on the created processes. It also shows how to send a message to all trainers and wait for their completion.

```python
from monarch.actor import Actor, endpoint, this_host

# spawn 8 trainer processes one for each gpu
training_procs = this_host().spawn_procs({"gpus": 8})


# define the actor to run on each process
class Trainer(Actor):
    @endpoint
    def train(self, step: int): ...


# create the trainers
trainers = training_procs.spawn("trainers", Trainer)

# tell all the trainers to take a step
fut = trainers.train.call(step=0)

# wait for all trainers to complete
fut.get()

```

Tokens:

Raw