How is this model able to be better than Gemma?
#13
by
ashiep
- opened
I gave it some unstructured garbage data and asked it to transform the data for a specific use case. It nailed the output. I copied and pasted the exact same prompt to Gemma- complete failure.
What makes this abliterated model so much more capable? I would really appreciate any information.
btw, I am speaking specifically about the Ollama model pidrilkin/gemma3_27b_abliterated:Q4_K_M which links here from the ollama site.