Edward Grefenstette egrefen

egrefen / brew doctor

Created January 31, 2013 10:25

Your system is raring to brew.

egrefen / HOMEBREW_MAKE_JOBS=1 VERBOSE=1 brew install

Created January 31, 2013 13:42

Homebrew tbb install problems

	==> Downloading http://threadingbuildingblocks.org/sites/default/files/software_releases/source/tbb41_20121003oss_src.tgz
	Already downloaded: /Library/Caches/Homebrew/tbb-4.1u1.tgz
	/usr/bin/tar xf /Library/Caches/Homebrew/tbb-4.1u1.tgz
	==> make tbb_build_prefix=BUILDPREFIX CONLY=cc CPLUS=c++ arch=intel64
	make tbb_build_prefix=BUILDPREFIX CONLY=cc CPLUS=c++ arch=intel64
	Created ./build/BUILDPREFIX_release and ..._debug directories
	/usr/bin/make -C "./build/BUILDPREFIX_debug" -r -f ../../build/Makefile.tbb cfg=debug tbb_root=../..
	../../build/Makefile.tbb:39: CONFIG: cfg=debug arch=intel64 compiler=gcc os=macos runtime=cc4.2.1_os10.8.2
	c++ -c -MMD -g -O0 -DTBB_USE_DEBUG -DUSE_PTHREAD -m64 -fPIC -D__TBB_BUILD=1 -Wall -Wno-non-virtual-dtor -I../../src -I../../src/rml/include -I../../include ../../src/tbb/concurrent_hash_map.cpp
	brew: superenv removed: -g -O0 -m64 -Wall -Wno-non-virtual-dtor

egrefen / maml_train.py

Last active May 12, 2021 08:15

Train maml model with torchmeta and higher v0.2.

	# Based on the code in https://github.com/tristandeleu/pytorch-meta/tree/master/examples/maml
	# Basically, we only use the dataset loaders/helpers from TorchMeta and replace usage of MetaModules
	# with normal pytorch nn.Modules, letting higher deal with making the inner loop unrollable and the
	# optimizers differentiable. This makes it easier to use another optimizer than SGD, or any arbitrary
	# third-party model, when doing MAML using this codebase.

	import os
	import torch
	import torch.nn as nn
	import torch.nn.functional as F