Convertion of custom pipeline to onnx #5657
Unanswered
Nikhil0307
asked this question in
Q&A
Replies: 2 comments 1 reply
-
I recommend Spox to write converts (I'm a maintainer of that project). It exposes the ONNX standard through a Pythonic interface. In essence, you have to re-implement the inference logic using ONNX operators. This section of the docs describes how you may organize your code to this end. |
Beta Was this translation helpful? Give feedback.
1 reply
-
Found a solution, For Pre and post processing I used onnxruntime-extensions with which custom operators can be created |
Beta Was this translation helpful? Give feedback.
0 replies
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
-
I have implemented an custom pipeline which has every operation done on the DataFrame from data cleaning, data extraction , data transformations and followed by ML operations ... where each and every step will be stored in a Pipeline object -> which will be used to infer where the input data goes through the pipeline and returns the results...
As mentioned i need to convert my custom pipeline to ONNX , have tried out but collapsed
Please help me out ..!!!
Thanks in advance
Beta Was this translation helpful? Give feedback.
All reactions