thomwolf

class MetaLearner(nn.Module):
    """ Bare Meta-learner class
        Should be added: intialization, hidden states, more control over everything
    """
    def __init__(self, model):
        super(MetaLearner, self).__init__()
        self.weights = Parameter(torch.Tensor(1, 2))

    def forward(self, forward_model, backward_model):
        """ Forward optimizer with a simple linear neural net

thomwolf

def train(forward_model, backward_model, optimizer, meta_optimizer, train_data, meta_epochs):
  """ Train a meta-learner
  Inputs:
    forward_model, backward_model: Two identical PyTorch modules (can have shared Tensors)
    optimizer: a neural net to be used as optimizer (an instance of the MetaLearner class)
    meta_optimizer: an optimizer for the optimizer neural net, e.g. ADAM
    train_data: an iterator over an epoch of training data
    meta_epochs: meta-training steps
  To be added: intialization, early stopping, checkpointing, more control over everything
  """

thomwolf

import cProfile
import pstats
import my_slow_module
cProfile.run('my_slow_module.run()', 'restats')
p = pstats.Stats('restats')
p.sort_stats('cumulative').print_stats(30)

thomwolf

class Rectangle:
    def __init__(self, w, h):
        self.w = w
        self.h = h
    def area(self):
        return self.w * self.h

thomwolf

from random import random

class Rectangle:
    def __init__(self, w, h):
        self.w = w
        self.h = h
    def area(self):
        return self.w * self.h

def check_rectangles(rectangles, threshold):

thomwolf

from cymem.cymem cimport Pool
from random import random

cdef struct Rectangle:
    float w
    float h

cdef int check_rectangles(Rectangle* rectangles, int n_rectangles, float threshold):
    cdef int n_out = 0
    # C arrays contain no size information => we need to give it explicitly

thomwolf

cdef struct Rectangle:
    float w
    float h

thomwolf

def slow_loop(doc_list, word, tag):
    n_out = 0
    for doc in doc_list:
        for tok in doc:
            if tok.lower_ == word and tok.tag_ == tag:
                n_out += 1
    return n_out

def main_nlp_slow(doc_list):
    n_out = slow_loop(doc_list, 'run', 'NN')

thomwolf

%%cython -+
import numpy # Sometime we have a fail to import numpy compilation error if we don't import numpy
from cymem.cymem cimport Pool
from spacy.tokens.doc cimport Doc
from spacy.typedefs cimport hash_t
from spacy.structs cimport TokenC

cdef struct DocElement:
    TokenC* c
    int length

thomwolf

import urllib.request
import spacy

with urllib.request.urlopen('https://raw.githubusercontent.com/pytorch/examples/master/word_language_model/data/wikitext-2/valid.txt') as response:
   text = response.read()
nlp = spacy.load('en')
doc_list = list(nlp(text[:800000].decode('utf8')) for i in range(10))