Merge pull request #91 from facebookresearch/prepare_wmt

Prepare scripts for WMT14 (#88)

README.md

@@ -49,7 +49,7 @@ The following command-line tools are available:
 ## Evaluating Pre-trained Models
 First, download a pre-trained model along with its vocabularies:
 ```
-$ curl https://s3.amazonaws.com/fairseq-py/models/wmt14.en-fr.fconv-py.tar.bz2 | tar xvjf -
+$ curl https://s3.amazonaws.com/fairseq-py/models/wmt14.v2.en-fr.fconv-py.tar.bz2 | tar xvjf -
 ```
 
 This model uses a [Byte Pair Encoding (BPE) vocabulary](https://arxiv.org/abs/1508.07909), so we'll have to apply the encoding to the source text before it can be translated.
@@ -135,30 +135,30 @@ BPE continuation markers can be removed with the `--remove-bpe` flag.
 
 We provide the following pre-trained fully convolutional sequence-to-sequence models:
 
-* [wmt14.en-fr.fconv-py.tar.bz2](https://s3.amazonaws.com/fairseq-py/models/wmt14.en-fr.fconv-py.tar.bz2): Pre-trained model for [WMT14 English-French](http://statmt.org/wmt14/translation-task.html#Download) including vocabularies
-* [wmt14.en-de.fconv-py.tar.bz2](https://s3.amazonaws.com/fairseq-py/models/wmt14.en-de.fconv-py.tar.bz2): Pre-trained model for [WMT14 English-German](https://nlp.stanford.edu/projects/nmt) including vocabularies
+* [wmt14.en-fr.fconv-py.tar.bz2](https://s3.amazonaws.com/fairseq-py/models/wmt14.v2.en-fr.fconv-py.tar.bz2): Pre-trained model for [WMT14 English-French](http://statmt.org/wmt14/translation-task.html#Download) including vocabularies
+* [wmt14.en-de.fconv-py.tar.bz2](https://s3.amazonaws.com/fairseq-py/models/wmt14.v2.en-de.fconv-py.tar.bz2): Pre-trained model for [WMT14 English-German](https://nlp.stanford.edu/projects/nmt) including vocabularies
 
 In addition, we provide pre-processed and binarized test sets for the models above:
-* [wmt14.en-fr.newstest2014.tar.bz2](https://s3.amazonaws.com/fairseq-py/data/wmt14.en-fr.newstest2014.tar.bz2): newstest2014 test set for WMT14 English-French
-* [wmt14.en-fr.ntst1213.tar.bz2](https://s3.amazonaws.com/fairseq-py/data/wmt14.en-fr.ntst1213.tar.bz2): newstest2012 and newstest2013 test sets for WMT14 English-French
-* [wmt14.en-de.newstest2014.tar.bz2](https://s3.amazonaws.com/fairseq-py/data/wmt14.en-de.newstest2014.tar.bz2): newstest2014 test set for WMT14 English-German
+* [wmt14.en-fr.newstest2014.tar.bz2](https://s3.amazonaws.com/fairseq-py/data/wmt14.v2.en-fr.newstest2014.tar.bz2): newstest2014 test set for WMT14 English-French
+* [wmt14.en-fr.ntst1213.tar.bz2](https://s3.amazonaws.com/fairseq-py/data/wmt14.v2.en-fr.ntst1213.tar.bz2): newstest2012 and newstest2013 test sets for WMT14 English-French
+* [wmt14.en-de.newstest2014.tar.bz2](https://s3.amazonaws.com/fairseq-py/data/wmt14.v2.en-de.newstest2014.tar.bz2): newstest2014 test set for WMT14 English-German
 
 Generation with the binarized test sets can be run in batch mode as follows, e.g. for English-French on a GTX-1080ti:
 ```
-$ curl https://s3.amazonaws.com/fairseq-py/models/wmt14.en-fr.fconv-py.tar.bz2 | tar xvjf - -C data-bin
-$ curl https://s3.amazonaws.com/fairseq-py/data/wmt14.en-fr.newstest2014.tar.bz2 | tar xvjf - -C data-bin
+$ curl https://s3.amazonaws.com/fairseq-py/models/wmt14.v2.en-fr.fconv-py.tar.bz2 | tar xvjf - -C data-bin
+$ curl https://s3.amazonaws.com/fairseq-py/data/wmt14.v2.en-fr.newstest2014.tar.bz2 | tar xvjf - -C data-bin
 $ python generate.py data-bin/wmt14.en-fr.newstest2014  \
   --path data-bin/wmt14.en-fr.fconv-py/model.pt \
   --beam 5 --batch-size 128 --remove-bpe | tee /tmp/gen.out
 ...
-| Translated 3003 sentences (95451 tokens) in 93.8s (1018.09 tokens/s)
-| Generate test with beam=5: BLEU4 = 40.67, 67.7/46.7/34.2/25.3 (BP=1.000, ratio=0.998, syslen=81377, reflen=81194)
+| Translated 3003 sentences (96311 tokens) in 166.0s (580.04 tokens/s)
+| Generate test with beam=5: BLEU4 = 40.83, 67.5/46.9/34.4/25.5 (BP=1.000, ratio=1.006, syslen=83262, reflen=82787)
 
 # Scoring with score.py:
 $ grep ^H /tmp/gen.out | cut -f3- > /tmp/gen.out.sys
 $ grep ^T /tmp/gen.out | cut -f2- > /tmp/gen.out.ref
 $ python score.py --sys /tmp/gen.out.sys --ref /tmp/gen.out.ref
-BLEU4 = 40.67, 67.7/46.7/34.2/25.3 (BP=1.000, ratio=0.998, syslen=81377, reflen=81194)
+BLEU4 = 40.83, 67.5/46.9/34.4/25.5 (BP=1.000, ratio=1.006, syslen=83262, reflen=82787)
 ```
 
 # Join the fairseq community

data/README.md

+Sample data processing scripts for the FAIR Sequence-to-Sequence Toolkit
+
+These scripts provide an example of pre-processing data for the NMT task.
+
+# prepare-iwslt14.sh
+
+Provides an example of pre-processing for IWSLT'14 German to English translation task: ["Report on the 11th IWSLT evaluation campaign" by Cettolo et al.](http://workshop2014.iwslt.org/downloads/proceeding.pdf)
+
+Example usage:
+```
+$ cd data/
+$ bash prepare-iwslt14.sh
+$ cd ..
+
+# Binarize the dataset:
+$ TEXT=data/iwslt14.tokenized.de-en
+$ python preprocess.py --source-lang de --target-lang en \
+  --trainpref $TEXT/train --validpref $TEXT/valid --testpref $TEXT/test \
+  --destdir data-bin/iwslt14.tokenized.de-en
+
+# Train the model:
+$ mkdir -p checkpoints/fconv
+$ CUDA_VISIBLE_DEVICES=0 python train.py data-bin/iwslt14.tokenized.de-en \
+  --lr 0.25 --clip-norm 0.1 --dropout 0.2 --max-tokens 4000 \
+  --label-smoothing 0.1 --force-anneal 200 \
+  --arch fconv_iwslt_de_en --save-dir checkpoints/fconv
+
+# Generate:
+$ python generate.py data-bin/iwslt14.tokenized.de-en \
+  --path checkpoints/fconv/checkpoint_best.pt \
+  --batch-size 128 --beam 5 --remove-bpe
+
+```
+
+
+# prepare-wmt14en2de.sh
+
+Provides an example of pre-processing for the WMT'14 English to German translation task. By default it will produce a dataset that was modeled after ["Attention Is All You Need" by Vaswani et al.](https://arxiv.org/abs/1706.03762) that includes news-commentary-v12 data.
+
+To use only data awailable in WMT'14 or to replicate results obtained in the original paper ["Convolutional Sequence to Sequence Learning" by Gehring et al.](https://arxiv.org/abs/1705.03122) run it with --icml17 instead:
+
+```
+$ bash prepare-wmt14en2de.sh --icml17
+```
+
+Example usage:
+
+```
+$ cd data/
+$ bash prepare-wmt14en2de.sh
+$ cd ..
+
+# Binarize the dataset:
+$ TEXT=data/wmt14_en_de
+$ python preprocess.py --source-lang en --target-lang de \
+  --trainpref $TEXT/train --validpref $TEXT/valid --testpref $TEXT/test \
+  --destdir data-bin/wmt14_en_de --thresholdtgt 0 --thresholdsrc 0
+
+# Train the model:
+# If it runs out of memory, try to set --max-tokens 1500 instead
+$ mkdir -p checkpoints/fconv_wmt_en_de
+$ python train.py data-bin/wmt14_en_de \
+  --lr 0.5 --clip-norm 0.1 --dropout 0.2 --max-tokens 4000 \
+  --label-smoothing 0.1 --force-anneal 50 \
+  --arch fconv_wmt_en_de --save-dir checkpoints/fconv_wmt_en_de
+
+# Generate:
+$ python generate.py data-bin/wmt14_en_de \
+  --path checkpoints/fconv_wmt_en_de/checkpoint_best.pt --beam 5 --remove-bpe
+
+```
+
+# prepare-wmt14en2fr.sh
+
+Provides an example of pre-processing for the WMT'14 English to French translation task.
+
+Example usage:
+
+```
+$ cd data/
+$ bash prepare-wmt14en2fr.sh
+$ cd ..
+
+# Binarize the dataset:
+$ TEXT=data/wmt14_en_fr
+$ python preprocess.py --source-lang en --target-lang fr \
+  --trainpref $TEXT/train --validpref $TEXT/valid --testpref $TEXT/test \
+  --destdir data-bin/wmt14_en_fr --thresholdtgt 0 --thresholdsrc 0
+
+# Train the model:
+# If it runs out of memory, try to set --max-tokens 1000 instead
+$ mkdir -p checkpoints/fconv_wmt_en_fr
+$ python train.py data-bin/wmt14_en_fr \
+  --lr 0.5 --clip-norm 0.1 --dropout 0.1 --max-tokens 3000 \
+  --label-smoothing 0.1 --force-anneal 50 \
+  --arch fconv_wmt_en_fr --save-dir checkpoints/fconv_wmt_en_fr
+
+# Generate:
+$ python generate.py data-bin/fconv_wmt_en_fr \
+  --path checkpoints/fconv_wmt_en_fr/checkpoint_best.pt --beam 5 --remove-bpe
+
+```
+

data/prepare-wmt14en2de.sh

+#!/bin/bash
+# Adapted from https://github.com/facebookresearch/MIXER/blob/master/prepareData.sh
+
+echo 'Cloning Moses github repository (for tokenization scripts)...'
+git clone https://github.com/moses-smt/mosesdecoder.git
+
+echo 'Cloning Subword NMT repository (for BPE pre-processing)...'
+git clone https://github.com/rsennrich/subword-nmt.git
+
+SCRIPTS=mosesdecoder/scripts
+TOKENIZER=$SCRIPTS/tokenizer/tokenizer.perl
+CLEAN=$SCRIPTS/training/clean-corpus-n.perl
+NORM_PUNC=$SCRIPTS/tokenizer/normalize-punctuation.perl
+REM_NON_PRINT_CHAR=$SCRIPTS/tokenizer/remove-non-printing-char.perl
+BPEROOT=subword-nmt
+BPE_TOKENS=40000
+
+URLS=(
+    "http://statmt.org/wmt13/training-parallel-europarl-v7.tgz"
+    "http://statmt.org/wmt13/training-parallel-commoncrawl.tgz"
+    "http://data.statmt.org/wmt17/translation-task/training-parallel-nc-v12.tgz"
+    "http://statmt.org/wmt14/test-full.tgz"
+)
+FILES=(
+    "training-parallel-europarl-v7.tgz"
+    "training-parallel-commoncrawl.tgz"
+    "training-parallel-nc-v12.tgz"
+    "test-full.tgz"
+)
+CORPORA=(
+    "training/europarl-v7.de-en"
+    "commoncrawl.de-en"
+    "training/news-commentary-v12.de-en"
+)
+
+# This will make the dataset compatible to the one used in "Convolutional Sequence to Sequence Learning"
+# https://arxiv.org/abs/1705.03122
+if [ "$1" == "--icml17" ]; then
+    URLS[2]="http://statmt.org/wmt14/training-parallel-nc-v9.tgz"
+    FILES[2]="training-parallel-nc-v9.tgz"
+    CORPORA[2]="training/news-commentary-v9.de-en"
+fi
+
+if [ ! -d "$SCRIPTS" ]; then
+    echo "Please set SCRIPTS variable correctly to point to Moses scripts."
+    exit
+fi
+
+src=en
+tgt=de
+lang=en-de
+prep=wmt14_en_de
+tmp=$prep/tmp
+orig=orig
+
+mkdir -p $orig $tmp $prep
+
+cd $orig
+
+for ((i=0;i<${#URLS[@]};++i)); do
+    file=${FILES[i]}
+    if [ -f $file ]; then
+        echo "$file already exists, skipping download"
+    else
+        url=${URLS[i]}
+        wget "$url"
+        if [ -f $file ]; then
+            echo "$url successfully downloaded."
+        else
+            echo "$url not successfully downloaded."
+            exit -1
+        fi
+        if [ ${file: -4} == ".tgz" ]; then
+            tar zxvf $file
+        elif [ ${file: -4} == ".tar" ]; then
+            tar xvf $file
+        fi
+    fi
+done
+cd ..
+
+echo "pre-processing train data..."
+for l in $src $tgt; do
+    rm $tmp/train.tags.$lang.tok.$l
+    for f in "${CORPORA[@]}"; do
+        cat $orig/$f.$l | \
+            perl $NORM_PUNC $l | \
+            perl $REM_NON_PRINT_CHAR | \
+            perl $TOKENIZER -threads 8 -a -l $l >> $tmp/train.tags.$lang.tok.$l
+    done
+done
+
+echo "pre-processing test data..."
+for l in $src $tgt; do
+    if [ "$l" == "$src" ]; then
+        t="src"
+    else
+        t="ref"
+    fi
+    grep '<seg id' $orig/test-full/newstest2014-deen-$t.$l.sgm | \
+        sed -e 's/<seg id="[0-9]*">\s*//g' | \
+        sed -e 's/\s*<\/seg>\s*//g' | \
+        sed -e "s/\’/\'/g" | \
+    perl $TOKENIZER -threads 8 -a -l $l > $tmp/test.$l
+    echo ""
+done
+
+echo "splitting train and valid..."
+for l in $src $tgt; do
+    awk '{if (NR%100 == 0)  print $0; }' $tmp/train.tags.$lang.tok.$l > $tmp/valid.$l
+    awk '{if (NR%100 != 0)  print $0; }' $tmp/train.tags.$lang.tok.$l > $tmp/train.$l
+done
+
+TRAIN=$tmp/train.de-en
+BPE_CODE=$prep/code
+rm -f $TRAIN
+for l in $src $tgt; do
+    cat $tmp/train.$l >> $TRAIN
+done
+
+echo "learn_bpe.py on ${TRAIN}..."
+python $BPEROOT/learn_bpe.py -s $BPE_TOKENS < $TRAIN > $BPE_CODE
+
+for L in $src $tgt; do
+    for f in train.$L valid.$L test.$L; do
+        echo "apply_bpe.py to ${f}..."
+        python $BPEROOT/apply_bpe.py -c $BPE_CODE < $tmp/$f > $tmp/bpe.$f
+    done
+done
+
+perl $CLEAN -ratio 1.5 $tmp/bpe.train $src $tgt $prep/train 1 250
+perl $CLEAN -ratio 1.5 $tmp/bpe.valid $src $tgt $prep/valid 1 250
+
+for L in $src $tgt; do
+    cp $tmp/bpe.test.$L $prep/test.$L
+done

data/prepare-wmt14en2fr.sh

+#!/bin/bash
+# Adapted from https://github.com/facebookresearch/MIXER/blob/master/prepareData.sh
+
+echo 'Cloning Moses github repository (for tokenization scripts)...'
+git clone https://github.com/moses-smt/mosesdecoder.git
+
+echo 'Cloning Subword NMT repository (for BPE pre-processing)...'
+git clone https://github.com/rsennrich/subword-nmt.git
+
+SCRIPTS=mosesdecoder/scripts
+TOKENIZER=$SCRIPTS/tokenizer/tokenizer.perl
+CLEAN=$SCRIPTS/training/clean-corpus-n.perl
+NORM_PUNC=$SCRIPTS/tokenizer/normalize-punctuation.perl
+REM_NON_PRINT_CHAR=$SCRIPTS/tokenizer/remove-non-printing-char.perl
+BPEROOT=subword-nmt
+BPE_TOKENS=40000
+
+URLS=(
+    "http://statmt.org/wmt13/training-parallel-europarl-v7.tgz"
+    "http://statmt.org/wmt13/training-parallel-commoncrawl.tgz"
+    "http://statmt.org/wmt13/training-parallel-un.tgz"
+    "http://statmt.org/wmt14/training-parallel-nc-v9.tgz"
+    "http://statmt.org/wmt10/training-giga-fren.tar"
+    "http://statmt.org/wmt14/test-full.tgz"
+)
+FILES=(
+    "training-parallel-europarl-v7.tgz"
+    "training-parallel-commoncrawl.tgz"
+    "training-parallel-un.tgz"
+    "training-parallel-nc-v9.tgz"
+    "training-giga-fren.tar"
+    "test-full.tgz"
+)
+CORPORA=(
+    "training/europarl-v7.fr-en"
+    "commoncrawl.fr-en"
+    "un/undoc.2000.fr-en"
+    "training/news-commentary-v9.fr-en"
+    "giga-fren.release2.fixed"
+)
+
+if [ ! -d "$SCRIPTS" ]; then
+    echo "Please set SCRIPTS variable correctly to point to Moses scripts."
+    exit
+fi
+
+src=en
+tgt=fr
+lang=en-fr
+prep=wmt14_en_fr
+tmp=$prep/tmp
+orig=orig
+
+mkdir -p $orig $tmp $prep
+
+cd $orig
+
+for ((i=0;i<${#URLS[@]};++i)); do
+    file=${FILES[i]}
+    if [ -f $file ]; then
+        echo "$file already exists, skipping download"
+    else
+        url=${URLS[i]}
+        wget "$url"
+        if [ -f $file ]; then
+            echo "$url successfully downloaded."
+        else
+            echo "$url not successfully downloaded."
+            exit -1
+        fi
+        if [ ${file: -4} == ".tgz" ]; then
+            tar zxvf $file
+        elif [ ${file: -4} == ".tar" ]; then
+            tar xvf $file
+        fi
+    fi
+done
+
+gunzip giga-fren.release2.fixed.*.gz
+cd ..
+
+echo "pre-processing train data..."
+for l in $src $tgt; do
+    rm $tmp/train.tags.$lang.tok.$l
+    for f in "${CORPORA[@]}"; do
+        cat $orig/$f.$l | \
+            perl $NORM_PUNC $l | \
+            perl $REM_NON_PRINT_CHAR | \
+            perl $TOKENIZER -threads 8 -a -l $l >> $tmp/train.tags.$lang.tok.$l
+    done
+done
+
+echo "pre-processing test data..."
+for l in $src $tgt; do
+    if [ "$l" == "$src" ]; then
+        t="src"
+    else
+        t="ref"
+    fi
+    grep '<seg id' $orig/test-full/newstest2014-fren-$t.$l.sgm | \
+        sed -e 's/<seg id="[0-9]*">\s*//g' | \
+        sed -e 's/\s*<\/seg>\s*//g' | \
+        sed -e "s/\’/\'/g" | \
+    perl $TOKENIZER -threads 8 -a -l $l > $tmp/test.$l
+    echo ""
+done
+
+echo "splitting train and valid..."
+for l in $src $tgt; do
+    awk '{if (NR%1333 == 0)  print $0; }' $tmp/train.tags.$lang.tok.$l > $tmp/valid.$l
+    awk '{if (NR%1333 != 0)  print $0; }' $tmp/train.tags.$lang.tok.$l > $tmp/train.$l
+done
+
+TRAIN=$tmp/train.fr-en
+BPE_CODE=$prep/code
+rm -f $TRAIN
+for l in $src $tgt; do
+    cat $tmp/train.$l >> $TRAIN
+done
+
+echo "learn_bpe.py on ${TRAIN}..."
+python $BPEROOT/learn_bpe.py -s $BPE_TOKENS < $TRAIN > $BPE_CODE
+
+for L in $src $tgt; do
+    for f in train.$L valid.$L test.$L; do
+        echo "apply_bpe.py to ${f}..."
+        python $BPEROOT/apply_bpe.py -c $BPE_CODE < $tmp/$f > $tmp/bpe.$f
+    done
+done
+
+perl $CLEAN -ratio 1.5 $tmp/bpe.train $src $tgt $prep/train 1 250
+perl $CLEAN -ratio 1.5 $tmp/bpe.valid $src $tgt $prep/valid 1 250
+
+for L in $src $tgt; do
+    cp $tmp/bpe.test.$L $prep/test.$L
+done

fairseq/tokenizer.py

@@ -13,12 +13,11 @@ import torch
 
 from fairseq import dictionary
 
+SPACE_NORMALIZER = re.compile("\s+")
 
 def tokenize_line(line):
-    line = re.sub(r"\t", "", line)
-    line = re.sub(r"^\s+", "", line)
-    line = re.sub(r"\s+$", "", line)
-    line = re.sub(r"\s+", " ", line)
+    line = SPACE_NORMALIZER.sub(" ", line)
+    line = line.strip()
     return line.split()