exo
exo copied to clipboard
Nodes can see each other but task doesn't split
Hello, Team.
Have 2 nodes on the same host, but only one taking load. Any ideas? Running by:
CUDA_VISIBLE_DEVICES=0 exo --node-id=node1 --node-port=65001 --discovery-module manual --discovery-config-path n2.cfg --inference-engine=tinygrad
DEBUG=6 CUDA_VISIBLE_DEVICES=1 exo --node-id=node2 --node-port=65002 --discovery-module manual --discovery-config-path n2.cfg --inference-engine=tinygrad
Configuration:
{
"peers": {
"node1": {
"address": "127.0.0.1",
"port": 65001,
"device_capabilities": {
"model": "Linux Box",
"chip": "NVIDIA GEFORCE GTX 1080 TI",
"memory": 11264,
"flops": {
"fp32": 11.34,
"fp16": 0.18,
"int8": 45.36
}
}
},
"node2": {
"address": "127.0.0.1",
"port": 65002,
"device_capabilities": {
"model": "Linux Box",
"chip": "NVIDIA GEFORCE GTX 1080 TI",
"memory": 11000,
"flops": {
"fp32": 11.34,
"fp16": 0.177,
"int8": 45.36
}
}
}
}
}