Any pointers to how to use this model for inference ( user provided text rather than the validation review text in csv)

[mancunian1792]

Hello,

I want to use this model in the inference regime, but I want to provide my text instead of looking at the validation text. However, the code has many moving parts, and I need some help.
This is the flow in which I would expect a typical inference to work. (text (set of reviews) -> pre-processing -> model -> output_summary)
Any tips for translating the above flow in your code?

Thanks.

[abrazinskas]

Hello, you could format your text into CSV files that the system expects. I will create a simpler interface for inference but it will take some time.

[mancunian1792]

Can I even run this code without gold summaries? The data pipeline expects the summaries1, summaries2, summaries3 columns that I don't have during inference.