franzko03

Hi, you can test with: for i in range(1,10): myTweet= tweetsDF.values[i,5] kategorie= classify_tweet(myTweet) print(kategorie + ':' + myTweet) tweetsDF is a pandas.DataFrame

hi, with torchtext 0.8.1 chapter 5 skript works as follows: (sample size and vocab size is arbitrary) `tweetsDF.sample(20000).to_csv("train-processed-sample.csv", header=None, index=None) LABEL = data.LabelField() TWEET = **data.Field(tokenizer_language='en', lower=True)** fields = [('score',None),...