英文:
Problem to load spacy in Google Cloud Functions but not locally
问题
我有一个类似这样的云函数：
import spacy
def entities():
NER = spacy.load("en_core_web_sm")
text='Looking for a Recruiting Data Analyst who can follow written instructions to correctly identify qualifying attributes from a candidates profile, load and properly code candidates in our recruiting database/CRM'
text_entities= NER(text)
entity_vector=[[word.text,word.label_] for word in text_entities.ents]
return print('Function finished')
entities()
由于某种原因，如果我在我的计算机上执行它，没有问题，但如果我在我的函数中加载它，然后我会遇到以下问题：
OSError: 无法找到模型'en_core_web_sm'。它似乎不是一个Python包或有效的数据目录路径。
请访问https://cloud.google.com/functions/docs/troubleshooting以获取深入的故障排除文档。
我尝试直接导入'en_core_web_sm'，但它引发了其他问题。我想知道是否可以手动获取Google函数保存spacy的位置，因为我猜当我搜索它时，环境在错误的位置查找。
英文:
I have a Cloud Function like this one:
import spacy
def entities():
NER = spacy.load("en_core_web_sm")
text='Looking for a Recruiting Data Analyst who can follow written instructions to correctly identify qualifying attributes from a candidates profile, load and properly code candidates in our recruiting database/CRM'
text_entities= NER(text)
entity_vector=[[word.text,word.label_] for word in text_entities.ents]
return print('Function finished')
entities()
For some reason, If I execute this in my computer, there is no issues, but if I load it in my function, then I get:
> OSError: [E050] Can't find model 'en_core_web_sm'. It doesn't seem to be a Python package or a valid path to a data directory.
> Please visit https://cloud.google.com/functions/docs/troubleshooting for in-depth troubleshooting documentation.
I tried importing 'en_core_web_sm' directly, but I had other issues with it. I wonder If I can manually get where is Google Function saving spacy, becouse I guess that when I look for it, the environment is looking for it in the wrong place.
答案1
得分: 1
以下是翻译好的部分：
这可能会解决你的问题：
import spacy
import os
os.system("python -m spacy download en_core_web_sm")
def entities():
NER = spacy.load("en_core_web_sm")
text='Looking for a Recruiting Data Analyst who can follow written instructions to correctly identify qualifying attributes from a candidates profile, load and properly code candidates in our recruiting database/CRM'
text_entities= NER(text)
entity_vector=[[word.text,word.label_] for word in text_entities.ents]
return print('Function finished')
entities()
英文:
This may solve your problem:
import spacy
import os
os.system("python -m spacy download en_core_web_sm")
def entities():
NER = spacy.load("en_core_web_sm")
text='Looking for a Recruiting Data Analyst who can follow written instructions to correctly identify qualifying attributes from a candidates profile, load and properly code candidates in our recruiting database/CRM'
text_entities= NER(text)
entity_vector=[[word.text,word.label_] for word in text_entities.ents]
return print('Function finished')
entities()
答案2
得分: 1
以下是要翻译的内容：
For me it worked to add the model to the requirements.txt (as the download python -m spacy command is not executed during deployment):
spacy<3.0.0
https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.3.0/en_core_web_sm-2.3.0.tar.gz#egg=en_core_web_sm
and then load it by
import en_core_web_sm
NER = en_core_web_sm.load()
Perhaps also the current version 3.5 of spaCy and the model works, but I use firebase and that version was causing a dependency conflict with the firebase_functions package, so I kept v2.3 for now...
英文:
For me it worked to add the model to the requirements.txt (as the download python -m spacy command is not executed during deployment:
spacy<3.0.0
https://github.com/explosion/spacy-models/releases/download/en_core_web_sm-2.3.0/en_core_web_sm-2.3.0.tar.gz#egg=en_core_web_sm
and then load it by
import en_core_web_sm
NER = en_core_web_sm.load()
Perhaps also the current version 3.5 of spaCy and the model works, but I use firebase and that version was causing a dependency conflict with the firebase_functions package, so I kept v2.3 for now...
通过集体智慧和协作来改善编程学习和解决问题的方式。致力于成为全球开发者共同参与的知识库，让每个人都能够通过互相帮助和分享经验来进步。
评论