Will we release a R1-like model for code generation, specifically for competitive programming？

[phonism]

FYI：https://arxiv.org/pdf/2502.06807v1

[zacliu2023]

Yes, we can discuss more on this topic

[zacliu2023]

There is some distil data on HF:

[zacliu2023]

Maybe, we just need a highly efficient framework to train the 671B-R1

[phonism]

I am doing some experiments on Qwen-7B. I hope to have the opportunity to experiment with larger models.

[zacliu2023]

I am doing some experiments on Qwen-7B. I hope to have the opportunity to experiment with larger models.

How are the experimental results?