erica

it was very helpful. Now it works perfectly on my local machine. Thank you so much☺️ I have one final question!
when using the deepseek r1 model as MODEL_COMPLETION, For all the data points in the completion column, I've noticed that the < * think> tags are disappearing. I'm carefully examining the source code to find the cause of this issue, but I haven't been able to identify it yet. Do you have any idea why this might be happening?
I've seen the prompt_style definition with [### Response: < * think> {}] . So I understand it's not a big deal, it's just out of curiosity!

Hello, thank you so much for the excellent tutorials and solutions! If possible, could you please create a tutorial on running a synthetic data generator locally? As others have mentioned, Llama can be quite unstable, and I'd like to try running it locally through Ollama. However, I'm having some trouble doing it on my own. Even just some tips on the config settings would be a huge help!