Basic Text Categorization fails in 3.2 #9732
-
|
A very basic test for Text Categorization fails. import spacy
nlp = spacy.load("en_core_web_md")
doc = nlp("This is a sentence.")
textcat = nlp.add_pipe("textcat")
processed = textcat(doc)I get the following error: I tried using en_core_web_sm, and another piece of code dervied from the API docs as follows. I still get the same error. nlp = spacy.load("en_core_web_sm")
from spacy.pipeline.textcat import DEFAULT_SINGLE_TEXTCAT_MODEL
config = {
"threshold": 0.5,
"model": DEFAULT_SINGLE_TEXTCAT_MODEL,
}
nlp.add_pipe("textcat", config=config)
nlp("spacy.io usage spacy-101")Info about spaCy
|
Beta Was this translation helpful? Give feedback.
Replies: 1 comment 3 replies
-
|
The |
Beta Was this translation helpful? Give feedback.
-
|
Thank you Adriane. Here's some of the output: Now I'm trying to load up the model and test it with some example sentences, but I can't seem to get anywhere. Is this the right way to go about this: nlp = spacy.load("/Users/vrun/Desktop/textcat_demo/training/model-best")
nlp("what does this eval to?")But that just echoes back the input sentence. |
Beta Was this translation helpful? Give feedback.
-
The categories are saved in When you pass text to the |
Beta Was this translation helpful? Give feedback.
-
|
Got it. Thanks! |
Beta Was this translation helpful? Give feedback.
The
textcatcomponent that you've just added this way hasn't been initialized (it doesn't even know which labels it's supposed to predict) and it hasn't been trained. I'd suggest having a look at the training docs (https://spacy.io/usage/spacy-101#training, https://spacy.io/usage/training) and trying out a textcat demo project (spacy project clone pipelines/textcat_demo, https://spacy.io/usage/projects)