Vincent QB, PhD vincentqb

Leaving a note here to adjust the solution above for cell magic (e.g. %%time), with nbconvert > 6.0.

In nbconvert_config.py:

def comment_magics(input, **kwargs):
    if input.startswith("%"):
        input = "# " + input
    return input

	perf record -F 320 --call-graph dwarf python some_script.py
	perf script > out.perf

	# Visualize as flame graphs: https://github.com/brendangregg/FlameGraph
	~/FlameGraph/stackcollapse-perf.pl out.perf > out.folded
	~/FlameGraph/flamegraph.pl out.folded > out.svg

	google-chrome out.svg

	import pandas as pd
	import json


	def read_json(filename):

	with open(filename, "r") as f:
	data = f.read()

	# TODO remove characters before first "{" from log directly instead

	#! /bin/bash

	#SBATCH --job-name=torchaudiomodel
	#SBATCH --output=/checkpoint/%u/jobs/audio-%j.out
	#SBATCH --error=/checkpoint/%u/jobs/audio-%j.err
	#SBATCH --signal=USR1@600
	#SBATCH --open-mode=append
	#SBATCH --partition=learnfair
	#SBATCH --time=4320
	#SBATCH --mem-per-cpu=5120

	import torch
	from .optimizer import Optimizer


	class RMSprop(Optimizer):
	r"""Implements RMSprop algorithm.
	Proposed by G. Hinton in his
	`course <http://www.cs.toronto.edu/~tijmen/csc321/slides/lecture_slides_lec6.pdf>`_.
	The centered version first appears in `Generating Sequences
	With Recurrent Neural Networks <https://arxiv.org/pdf/1308.0850v5.pdf>`_.

	from itertools import repeat


	class StridedBuffer:

	def __init__(self, generator, stride, length):
	self._generator = generator
	self._stride = stride
	self._length = length
	self._buffer = [[] for _ in repeat(None, stride)]