cannot get 405B-model to run

#30
by hAI-hades - opened

Hi,
I downloaded meta-llama/Llama-3.1-405B-Instruct and cannot get it to run, i tried really many different setups and parameters, but I got always the error:
ValueError: Supplied state dict for model.layers.0.mlp.gate_proj.weight does not contain bitsandbytes__* and possibly other quantized_stats components.
I tried other Llama3.x models with the same code: no problem
Help regarding this would be highly appreciated
Best, Patrick

Sign up or log in to comment