I am trying to setup a Neural network through the Deepspeed framework and Pytorch.
I cannot get over this error and the internet is not helping very much... Someone saw this before? (Complete log: https://hastebin.com/ugeciyujid.sql)
'"hostname -I"' is not recognized as an internal or external command, operable program or batch file.
Here is a code sample to reproduce:
import torch
import torch.nn as nn
class SimpleModel(nn.Module):
def __init__(self, in_channels=3, out_channels=1):
super(SimpleModel, self).__init__()
self.fc1 = nn.Linear(in_channels, 12)
self.fc1 = nn.Linear(12, 24)
self.fc3 = nn.Linear(24, out_channels)
self.relu = nn.ReLU()
def forward(self, x):
x = self.fc1(x)
x = self.relu(x)
x = self.fc2(x)
x = self.relu(x)
x = self.fc3(x)
return self.relu(x)
import deepspeed
import torch
import torch.nn as nn
from model import SimpleModel
def setup():
model = SimpleModel()
arguments = {
"train_batch_size": 1,
"optimizer": {
"type": "Adam",
"params": {
"lr": 5e-04
}
},
"zero_optimization": "true"
}
engine, optimizer, _, _ = deepspeed.initialize(
args=arguments,
model=model,
model_parameters=model.parameters()
)
if __name__ == '__main__':
setup()
My setup is python 3.8, a GTX 1060 6GB and torch 1.9 with cuda 11.1.