Import trained model to ElastichSearch

Elastic Stack Elasticsearch
1

hello everyone. I have a question about importing my model to Elasticsearch.
When i imported the model for task text embedding and started deployment mode on Kibana web, i got an error message
" Couldn't start trained model deployment, the following nodes failed with errors [ Failed to connect to PyTorch process to job]".

I struggled to get the answer on Google but didn't any information helpful. Can anyone help me to fix that?

2

Hi @Khanh_Dao_Minh Welcome to the community.

A couple things.

First basics:

What version of the Stack?

Is the node enabled with machine learning role?

Do you have a platinum or Trial license enabled?

Also can you show the exact commands and outputs?

Then perhaps we can help..

1 Like
3

Thank @stephenb for your response. I fix that by setting xpack.ml.enabled = true in yml config file and down version Pytorch.

1 Like
4

This topic was automatically closed 28 days after the last reply. New replies are no longer allowed.