Skip to content

Instantly share code, notes, and snippets.

Setepenre Delaunay

  • Montréal
Block or report user

Report or block Delaunay

Hide content and notifications from this user.

Learn more about blocking users

Contact Support about this user’s behavior.

Learn more about reporting abuse

Report abuse
View GitHub Profile
View log.err
HIP_DB=3 HCC_SERIALIZE_KERNEL=0x6 python test/test_cuda.py TestCuda.test_multinomial_invalid_probs_cuda > log.out
HIP_DB=0x3 [api+sync]
hip-api pid:22173 tid:1:HIP initialized short_tid#1 (maps to full_tid: 0x7fd84c378740)
hip-sync pid:22173 tid:1:locking criticalData=0x7fd7ee7da798 for ctx:0x7fd7ee7da780.dev:0..
hip-sync pid:22173 tid:1: addStream: stream:0.0
hip-sync pid:22173 tid:1:created ctx with defaultStream=0x7fd7eee38670 (stream:0.0)
hip-sync pid:22173 tid:1:auto-unlocking criticalData=0x7fd7ee7da798 for ctx:0x7fd7ee7da780.dev:0...
hip-sync pid:22173 tid:1:pid=22173 <ihipInit> g_numLogicalThreads=4
<<hip-api pid:22173 tid:1.1 hipGetDeviceCount (0x7ffcaee0a1c4) @370289587147796
hip-api pid:22173 tid:1.1 hipGetDeviceCount ret= 0 (hipSuccess)>> +3657 ns
View ImageNet_HDF5.py
import torchvision
import torchvision.transforms as transforms
import time
import torch
import os
import h5py
import numpy as np
from MixedPrecision.tools.stats import StatStream
@Delaunay
Delaunay / pool.cpp
Last active Apr 11, 2018
Simple Pool Implementation (gcc 8, C++2a)
View pool.cpp
#include <thread>
#include <mutex>
#include <tuple>
#include <list>
#include <memory>
namespace pool{
namespace internal {
template<typename T>
You can’t perform that action at this time.