R
RunPod•4mo ago
caseus

Linux kernel version is 5.4.0

per accelerate: https://github.com/huggingface/accelerate/blob/85a75d4c3d0deffde2fc8b917d9b1ae1cb580eb2/src/accelerate/utils/other.py#L314C1-L331C1 
Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.
Detected kernel version 5.4.0, which is below the recommended minimum of 5.5.0; this can cause the process to hang. It is recommended to upgrade the kernel to the minimum version or higher.
basically H100's are currently unusable as it hangs for me using accelerate to train models.
GitHub
accelerate/src/accelerate/utils/other.py at 85a75d4c3d0deffde2fc8b9...
🚀 A simple way to train and use PyTorch models with multi-GPU, TPU, mixed-precision - huggingface/accelerate
1 Reply
caseus
caseus•4mo ago
@Alpay Ariyak
Want results from more Discord servers?
Add your server
More Posts
How to scale pod GPU count properly?Hello, we have some pods running with 2x4090, and what is the best way to increase this to e.g 4x409Is execution timeout per request or per worker execution?https://docs.runpod.io/serverless/endpoints/send-requests#--execution-policy "Execution Timeout: SS3 ENV does not work as described in the Runpod DocumentionHi all, I have a serverless function and also all env variable as its written in documention. But itGPU type prioritization seems to have stopped working on 13th of MarchI have an endpoint with 3 cheapest GPU types selected in the order of their price (i.e. 4090 is my 3distributed trainingIs it possible to set up a slurm cluster for distributed training on Runpod?How can i bulk download all my images generated in my Output FolderHow can i bulk download all my images generated in my Output Folder (Fooocus)? I'm in the Jupyter La