I met a problem after downloading model checkpoint

#1
by panzeyu2013 - opened

Python raised an error when trying to use pipeline, AutoModel, and LlamaForCausalLM, indicating that "Error(s) in loading state_dict for LlamaForCausalLM:\n\tsize mismatch for model.layers.0.self_attn.k_proj.weight: copying a param with shape torch.Size([256, 2048]) from checkpoint, the shape in current model is torch.Size([2048, 2048])." from layer0 to layer21. How can I solve this problem

TinyLlama org

Make sure you have transformers>=4.30

I have tested some versions of transformers. It seems that it only works well on transformer==4.32.0. Still thanks for your help

Sign up or log in to comment