sft time
#5
by
hanxinyan
- opened
I’m wondering how long it takes you to sft Qwen2.5-Math-7B-Instruct on OpenR1-Math-220k using your config. I’ve been trying to reproduce it but find that it takes a lot of time. Any insights would be appreciated. Thank you!
Hi, were you able to resolve the training time issue for SFT on Qwen2.5-Math-7B-Instruct with OpenR1-Math-220k? I'd be curious to hear what worked for you, or if you found any optimizations. Thanks!