Convert OpenAssistant/oasst-sft-6-llama-30b-xor

[DawoodTouseef]

I'm trying to convert to GGUF the new OpenAssistant/oasst-sft-6-llama-30b-xor

python3 convert.py /Downloads/phi-1_5
Loading model file /Downloads/ooost/pytorch_model.bin
Traceback (most recent call last):
File "/llama.cpp/convert.py", line 1196, in
main()
File "/llama.cpp/convert.py", line 1145, in main
params = Params.load(model_plus)
^^^^^^^^^^^^^^^^^^^^^^^
File "/llama.cpp/convert.py", line 299, in load
params = Params.loadHFTransformerJson(model_plus.model, hf_config_path)
^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^^
File "/llama.cpp/convert.py", line 209, in loadHFTransformerJson
n_embd = config["hidden_size"]
~~~~~~^^^^^^^^^^^^^^^
KeyError: 'hidden_size'

[ianscrivener]

You ran pip install -r requirements?
What version of python are you using?
Personally I've had some issues with python 3.9.x & 3.11.x... so use python 3.10.12 as my default for AI & LLMs.

[KerfuffleV2]

Did you un-XOR it first? It looks like even the metadata like JSON files are XORed.

[DawoodTouseef]

Idk I Got it from hugging face

[KerfuffleV2]

Idk I Got it from hugging face

Okay, then that's a "no". Please read the model card for the model you're talking about having issues with: https://huggingface.co/OpenAssistant/oasst-sft-6-llama-30b-xor

It's basically encrypted. You'll need to follow the instructions there to decrypt it.