KeyError: 'model.embed_tokens.weight' when converting .safetensors to ggml

[Jake36921]

(base) PS E:\Games\llama.cpp> python3 convert.py OPT-13B-Erebus-4bit-128g.safetensors --outtype q4_1 --outfile 4ggml.bin Loading model file OPT-13B-Erebus-4bit-128g.safetensors Loading vocab file tokenizer.model Traceback (most recent call last): File "E:\Games\llama.cpp\convert.py", line 1147, in main() File "E:\Games\llama.cpp\convert.py", line 1137, in main model = do_necessary_conversions(model) File "E:\Games\llama.cpp\convert.py", line 983, in do_necessary_conversions model = convert_transformers_to_orig(model) File "E:\Games\llama.cpp\convert.py", line 588, in convert_transformers_to_orig out["tok_embeddings.weight"] = model["model.embed_tokens.weight"] KeyError: 'model.embed_tokens.weight' (base) PS E:\Games\llama.cpp>

Model is from here: https://huggingface.co/notstoic/OPT-13B-Erebus-4bit-128g

[jon-chuang]

I don't think OPT 13B is currently supported.

[github-actions[bot]]

This issue was closed because it has been inactive for 14 days since being marked as stale.