File size: 2,017 Bytes
95456ed | 1 2 3 4 5 6 7 8 9 10 11 12 13 14 15 16 17 18 19 20 21 22 23 24 25 26 27 28 29 30 31 32 33 34 35 36 37 38 39 40 41 42 43 44 45 46 47 48 49 50 51 52 53 54 55 56 57 58 59 60 61 62 63 64 65 66 67 68 69 70 71 72 73 74 75 76 77 78 79 80 81 82 83 84 85 86 | """
Helpers for distributed training.
"""
import io
import os
import socket
import blobfile as bf
import torch as th
import torch.distributed as dist
# Change this to reflect your cluster layout.
def setup_dist():
"""
Setup a distributed process group.
"""
if dist.is_initialized():
return
# UNCOMMENT IF ON LINUX/MAC
# backend = "gloo" if not th.cuda.is_available() else "nccl"
backend = "gloo"
if backend == "gloo":
hostname = "localhost"
else:
hostname = socket.gethostbyname(socket.getfqdn())
if os.environ.get("LOCAL_RANK") is None:
os.environ["MASTER_ADDR"] = hostname
os.environ["RANK"] = str(0)
os.environ["WORLD_SIZE"] = str(1)
port = _find_free_port()
os.environ["MASTER_PORT"] = str(port)
os.environ["LOCAL_RANK"] = str(0)
dist.init_process_group(backend=backend, init_method="env://")
if th.cuda.is_available(): # This clears remaining caches in GPU 0
th.cuda.set_device(dev())
th.cuda.empty_cache()
def dev():
"""
Get the device to use for torch.distributed.
"""
if th.cuda.is_available():
return th.device(f"cuda:{os.environ['LOCAL_RANK']}")
return th.device("cpu")
def load_state_dict(path, **kwargs):
"""
Load a PyTorch file.
"""
# if int(os.environ['LOCAL_RANK']) == 0:
with bf.BlobFile(path, "rb") as f:
data = f.read()
return th.load(io.BytesIO(data), **kwargs)
def sync_params(params):
"""
Synchronize a sequence of Tensors across ranks from rank 0.
"""
for p in params:
with th.no_grad():
dist.broadcast(p, 0)
def _find_free_port():
try:
s = socket.socket(socket.AF_INET, socket.SOCK_STREAM)
s.bind(("", 0))
s.setsockopt(socket.SOL_SOCKET, socket.SO_REUSEADDR, 1)
return s.getsockname()[1]
finally:
s.close()
|