Confusion about the baseline models

[puck-24]

Thanks for your open source for this awesome work. The improvement of applying Dec2Enc is very significant!

However, I have a confusion that is the baseline model (e.g. Qwen2.5-0.5B, Qwen2.5-1.5B in Table 1 and Table 2) also fine-tuned on the tested datasets?

Your clarification would be greatly appreciated.

[HuangZixian]

Thank you for your attention!

Yes, all the baselines were fine-tuned on the training set of the same datasets for a fair comparison.