Model performance

#1
by spanspek - opened

Thanks @noctrex for the work done to quantize this model

Ignoring this version of it, has anyone tried this LFM2? I gave it a quick relatively simple coding task and it did horribly! Gave the same task to gpt-oss:20b and gave both responses to Grok to grade and LFM2 got a 4.5/10 and gpt-oss got 8.5/10. It was a straightforward task, 20 lines of pyspark needed to be converted to duckdb in Python.

I'm not impressed, I don't see any use-cases for this model given the ability of it's competition but maybe I'm being too narrow - has anyone found any good use cases for this model?

Please ignore, let me rather post this on the Liquid AI page

spanspek changed discussion status to closed
Owner

Well this model is a general use model, it's not for coding.
I'll use it successfully on the app KaraKeep for summarization and tag creation.

Sign up or log in to comment