itsource

word2vec bin 파일을 텍스트로 변환

mycopycode 2022. 8. 21. 19:56
반응형

word2vec bin 파일을 텍스트로 변환

2vec 사이트에서 Google News-vectors-negative 300.bin.gz를 다운로드할 수 있습니다..bin 파일(약 3.4)GB)는 바이너리 형식이 아닙니다.Tomas Mikolov는 "바이너리 포맷을 텍스트 포맷으로 변환하는 것은 매우 간단할 것입니다(단, 디스크 공간이 더 많이 필요합니다).거리 도구에서 코드를 확인하십시오. 바이너리 파일을 읽는 것은 상당히 간단합니다."유감스럽게도 저는 http://word2vec.googlecode.com/svn/trunk/distance.c을 이해할 수 있을 만큼 C를 모릅니다.

아마 gensim도 할 수 있을 것 같은데, 내가 찾은 모든 튜토리얼은 텍스트에서 변환하는 것이지, 다른 방법은 아닌 것 같다.

누가 C 코드 수정이나 gensim이 텍스트를 출력하도록 지시할 수 있습니까?

이 코드를 사용하여 바이너리 모델을 로드하고 모델을 텍스트 파일에 저장합니다.

from gensim.models.keyedvectors import KeyedVectors

model = KeyedVectors.load_word2vec_format('path/to/GoogleNews-vectors-negative300.bin', binary=True)
model.save_word2vec_format('path/to/GoogleNews-vectors-negative300.txt', binary=False)

참고 자료: API눌레지.

주의:

위 코드는 새로운 버전의 gensim용입니다. 이전 버전에서는 다음 코드를 사용했습니다.

from gensim.models import word2vec

model = word2vec.Word2Vec.load_word2vec_format('path/to/GoogleNews-vectors-negative300.bin', binary=True)
model.save_word2vec_format('path/to/GoogleNews-vectors-negative300.txt', binary=False)

Word2vec-toolkit 메일링 리스트에서 Thomas Mensink는 .bin 파일을 텍스트로 변환하는 작은 C 프로그램의 형태로 답변을 제공했습니다.이것은 distance.c 파일의 수정입니다.원래 distance.c를 아래 Thomas의 코드로 대체하고 word2vec(make clean; make)을 재구축하여 컴파일된 거리를 readbin으로 이름을 변경하였습니다.그리고나서./readbin vector.binvector.bin 텍스트버전이 생성됩니다.

//  Copyright 2013 Google Inc. All Rights Reserved.
//
//  Licensed under the Apache License, Version 2.0 (the "License");
//  you may not use this file except in compliance with the License.
//  You may obtain a copy of the License at
//
//      http://www.apache.org/licenses/LICENSE-2.0
//
//  Unless required by applicable law or agreed to in writing, software
//  distributed under the License is distributed on an "AS IS" BASIS,
//  WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
//  See the License for the specific language governing permissions and
//  limitations under the License.

#include <stdio.h>
#include <string.h>
#include <math.h>
#include <malloc.h>

const long long max_size = 2000;         // max length of strings
const long long N = 40;                  // number of closest words that will be shown
const long long max_w = 50;              // max length of vocabulary entries

int main(int argc, char **argv) {
  FILE *f;
  char file_name[max_size];
  float len;
  long long words, size, a, b;
  char ch;
  float *M;
  char *vocab;
  if (argc < 2) {
    printf("Usage: ./distance <FILE>\nwhere FILE contains word projections in the BINARY FORMAT\n");
    return 0;
  }
  strcpy(file_name, argv[1]);
  f = fopen(file_name, "rb");
  if (f == NULL) {
    printf("Input file not found\n");
    return -1;
  }
  fscanf(f, "%lld", &words);
  fscanf(f, "%lld", &size);
  vocab = (char *)malloc((long long)words * max_w * sizeof(char));
  M = (float *)malloc((long long)words * (long long)size * sizeof(float));
  if (M == NULL) {
    printf("Cannot allocate memory: %lld MB    %lld  %lld\n", (long long)words * size * sizeof(float) / 1048576, words, size);
    return -1;
  }
  for (b = 0; b < words; b++) {
    fscanf(f, "%s%c", &vocab[b * max_w], &ch);
    for (a = 0; a < size; a++) fread(&M[a + b * size], sizeof(float), 1, f);
    len = 0;
    for (a = 0; a < size; a++) len += M[a + b * size] * M[a + b * size];
    len = sqrt(len);
    for (a = 0; a < size; a++) M[a + b * size] /= len;
  }
  fclose(f);
  //Code added by Thomas Mensink
  //output the vectors of the binary format in text
  printf("%lld %lld #File: %s\n",words,size,file_name);
  for (a = 0; a < words; a++){
    printf("%s ",&vocab[a * max_w]);
    for (b = 0; b< size; b++){ printf("%f ",M[a*size + b]); }
    printf("\b\b\n");
  }  

  return 0;
}

에서 "\b\b" 를 제거했습니다.printf.

덧붙여서, 결과 텍스트 파일에는 아직 텍스트 워드와 불필요한 공백이 포함되어 있어 수치 계산에는 필요 없었습니다.bash 명령어를 사용하여 각 행에서 첫 번째 텍스트 열과 후행 공백을 제거했습니다.

cut --complement -d ' ' -f 1 GoogleNews-vectors-negative300.txt > GoogleNews-vectors-negative300_tuples-only.txt
sed 's/ $//' GoogleNews-vectors-negative300_tuples-only.txt

형식은 IEEE 754 싱글소문자 바이너리 부동소수점 형식입니다.binary 32 http://en.wikipedia.org/wiki/Single-precision_floating-point_format 에서는 little-endian을 사용합니다.

예를 들어 보겠습니다.

  • 첫 번째 행은 문자열 형식: "3000000 300\n" (vocabSize & vecSize, getByte to byte=='\n')
  • 다음 행에는 먼저 vocab word가 포함되며, 그 다음 (300*4바이트의 부동값, 각 치수에 대해 4바이트)가 포함됩니다.

    getByte till byte==32 (space). (60 47 115 62 32 => <\s>[space])
    
  • 각 다음 4바이트는 하나의 플로트 번호를 나타냅니다.

    다음 4바이트: 0 - 108 58 => 0.001129150390625.

Wikipedia 링크를 통해 방법을 확인할 수 있습니다. 예를 들어 다음과 같습니다.

(리틀 엔디언 -> 역순서) 0011101010000000000000000000000000000

  • 첫 번째는 부호 비트 = > 부호 = 1(소수 = -1)입니다.
  • 다음 8 비트 => 117 => exp = 2 ^ (표준)
  • 다음 23비트 => pre = 0*2^(-1) + 0*2^(-2) + 1*2^(-3) + 1*2^(-5)

값 = 기호 * exp * pre

이진 파일을 word2vec로 로드한 다음 다음과 같이 텍스트 버전을 저장할 수 있습니다.

from gensim.models import word2vec
 model = word2vec.Word2Vec.load_word2vec_format('Path/to/GoogleNews-vectors-negative300.bin', binary=True)
 model.save("file.txt")

`

300을 사용하고 News-vectors-negative 300.bin을 포함하고 .binary = True모델을 로드하는 동안 플래그를 설정합니다.

from gensim import word2vec

model = word2vec.Word2Vec.load_word2vec_format('Path/to/GoogleNews-vectors-negative300.bin', binary=True) 

정상적으로 작동하는 것 같습니다.

에러가 표시되는 경우:

ImportError: No module named models.word2vec

API 업데이트가 있었기 때문입니다.이 조작은 유효합니다.

from gensim.models.keyedvectors import KeyedVectors

model = KeyedVectors.load_word2vec_format('./GoogleNews-vectors-negative300.bin', binary=True)
model.save_word2vec_format('./GoogleNews-vectors-negative300.txt', binary=False)

비슷한 문제가 있었습니다만, bin/non-bin(gensim) 모델의 출력을 CSV로 받고 싶었습니다.

다음은 python에서 이를 수행하는 코드이며, gensim이 설치되어 있다고 가정합니다.

https://gist.github.com/dav009/10a742de43246210f3ba

사용하는 코드는 다음과 같습니다.

import codecs
from gensim.models import Word2Vec

def main():
    path_to_model = 'GoogleNews-vectors-negative300.bin'
    output_file = 'GoogleNews-vectors-negative300_test.txt'
    export_to_file(path_to_model, output_file)


def export_to_file(path_to_model, output_file):
    output = codecs.open(output_file, 'w' , 'utf-8')
    model = Word2Vec.load_word2vec_format(path_to_model, binary=True)
    print('done loading Word2Vec')
    vocab = model.vocab
    for mid in vocab:
        #print(model[mid])
        #print(mid)
        vector = list()
        for dimension in model[mid]:
            vector.append(str(dimension))
        #line = { "mid": mid, "vector": vector  }
        vector_str = ",".join(vector)
        line = mid + "\t"  + vector_str
        #line = json.dumps(line)
        output.write(line + "\n")
    output.close()

if __name__ == "__main__":
    main()
    #cProfile.run('main()') # if you want to do some profiling

convertvec은 Word2vec 라이브러리의 다른 형식 간에 벡터를 변환하는 작은 도구입니다.

벡터를 이진 텍스트에서 일반 텍스트로 변환:

./vec bin2txt input.bin 출력.txt

일반 텍스트에서 이진 텍스트로 벡터 변환:

./vecvec txt2bin 입력.txt 출력.휴지통

지금 바로 업데이트하면 더 쉬운 방법이 있습니다.

「 」를 사용하고 word2vechttps://github.com/dav/word2vec 에서는, 라고 하는 추가 옵션이 있습니다.-binary 있다1 또는 바이너리 파일을 0텍스트 파일을 생성합니다.는 " "에서 합니다.demo-word.sh★★★★

time ./word2vec -train text8 -output vectors.bin -cbow 1 -size 200 -window 8 -negative 25 -hs 0 -sample 1e-4 -threads 20 -binary 0 -iter 15

언급URL : https://stackoverflow.com/questions/27324292/convert-word2vec-bin-file-to-text

반응형