Finn

请问有对openllama和llama比较过基础能力吗？我看预训练数据还是比较少的，有考虑加入redpajama数据集么？

save model nccl timeout

3

When saving the model after training, I always get the message "nccl timeout". The model is large and the saving time is long. How can I solve this problem? I...

[BUG]save model NCCL timeout

2

**Describe the bug** When the model is large, the watchdog times out when saving the model， model is large, gpt-oss-120B, It timeout about 10 min, so the init setting seems...