Shang Yang

Did you find which part leads to the high CPU utilization on your machine? I also noticed a similar situation on a new environment. And it seems that this indeed...

**I would suggest you use "hashmap" mode if that is the case.** Since I did not meet the same illegal memory access problem when running the spvnas training manuscript, I...

Could you please provide more details or a short code snippet to reproduce this error? Thank you!

Hi All, the [pypi server](http://24.199.104.228/simple) has been restarted and can now be accessed. Sorry for the inconvenience. Many thanks to @kentang-mit .

I also met the same issue. I trained OLMo-1B with the provided config files. The batch size is: ```bash global_train_batch_size: 2048 device_train_microbatch_size: 8 ``` I used 8 NVIDIA A100 GPUs...