classifier training error_solved

Hi all,

I am now experiencing some errors while trying to use the following:
qiime feature-classifier classify-sklearn

the error I get is the following:

ValueError: The scikit-learn version (0.21.2) used to generate this artifact does not match the current version of scikit-learn installed (0.24.1). Please retrain your classifier for your current deployment to prevent data-corruption errors

I have found the following guidelines from your documentation regarding feature classifier:
however I would need the fasta sequences and OTU assignments
where could I find those files?

using the following file:

Greengenes 13_8 99% OTUs full-length sequences (

the problem persists, so I was wondering if I could regenerate from fasta sequences and taxonomy as indicated in the tutoriale page, however I would need 99% instead of 85% identity

I report version details

qiime infoSystem versions

Python version: 3.8.10

QIIME 2 release: 2021.8

QIIME 2 version: 2021.8.0

q2cli version: 2021.8.0

Installed plugins

alignment: 2021.8.0

composition: 2021.8.0

cutadapt: 2021.8.0
dada2: 2021.8.0
deblur: 2021.8.0
demux: 2021.8.0
diversity: 2021.8.0
diversity-lib: 2021.8.0
emperor: 2021.8.0
feature-classifier: 2021.8.0
feature-table: 2021.8.0
fragment-insertion: 2021.8.0
gneiss: 2021.8.0
longitudinal: 2021.8.0
metadata: 2021.8.0
phylogeny: 2021.8.0
quality-control: 2021.8.0
quality-filter: 2021.8.0
sample-classifier: 2021.8.0
taxa: 2021.8.0
types: 2021.8.0
vsearch: 2021.8.0

thanks a lot,

I have read that it is possible to use classifiers already available, referred to the specific qiime version, I used the appropriate file in the appropriate version, available here:
and solved the issue


This topic was automatically closed 31 days after the last reply. New replies are no longer allowed.