Dataset csv creation from markdown file

I’m trying to create my first dataset csv file to use for question and answering. I’m following a blog of how to do this through OpenAI, markdown-to-train-data, but I want to do it through Huggingface instead. My issue is finding a working inference model I can post to on huggingface. How do I find an inference model that is running and is comparable to gpt-3.5-turbo for ? I have searched models for this, but there doesn’t seem to be one that is running?

3 Likes

The Serverless Inference API is currently in the middle of a complete overhaul, so I think there are many cases where it doesn’t work properly…

The following should at least give you a clue.

2 Likes

Thanks, this is helpful. Since this is just a small project(just one markdown file), do you recommend using a google transformer or some other transformer model based on google?

2 Likes

I thought you wanted to convert thousands of them…:sweat_smile:

If it’s just one or two, you might be able to get away with using Spaces for free.

1 Like