Failure of chargoddard/mixtralnt-4x7b-test

#452
by chargoddard - opened

Hi there!

I was wondering why this particular model failed. It is in safetensors format and able to be loaded with AutoModel.

It's an unconventional Mixtral config, but it works locally.

Here is the requests file in question:
https://huggingface.co/datasets/open-llm-leaderboard/requests/blob/main/chargoddard/mixtralnt-4x7b-test_eval_request_False_bfloat16_Original.json

Thanks!

Open LLM Leaderboard org

Hi!
Thank you for your issue!
This job was cancelled this morning, as your model was downloading (during cluster maintenance) - I passed it to pending again.

clefourrier changed discussion status to closed

Thanks for the prompt response!

Unfortunately it looks like it failed again. What was the cause this time?

Thanks again for your time.

chargoddard changed discussion status to open
Open LLM Leaderboard org

Hi!

Your model failed while loading the checkpoint's shards. I relaunched it in case it was a hardware failure, but if it fails again, it's probably an issue with your model. Did you follow all the pre-submission steps?

clefourrier changed discussion status to closed

Sign up or log in to comment