Spaces:
Running
on
CPU Upgrade
Model upload failed
H! Yesterday we submitted some models to the leaderboard.
I have verified that both models can be loaded using AutoModel and AutoTokenizer.
But in open-llm-leaderboard/requests it shows as FAILED.
(porkorbeef/Llama-2-13b-sf, porkorbeef/Llama-2-13b etc)
How can I diagnose what's going on here?
Hi!
Two of your models succeeded and had problems during results upload (it's been fixed), and there was a problem when loading porkorbeef/Llama-2-13b from the hub (an error 500, which tells us nothing sadly).
Can you convert your weights to safetensors and ping me again?
(Btw, for your next issue, could you point directly towards the requests files as indicated in the FAQ to help me help you faster?)
Thank you very much for your kind reply.
Next time I have an issue, I'll ask in the FAQ. thank you :)
Sorry, I've been unclear 😅: creating a new issue is exactly what you need to do!
However, it would help me a lot if you could point to the correct request file for your model : for example, your failed model's request file is here. It allows me to access the logs faster :)
I got it! I hope the next my upload will be successful :)