Transformers backend: Auto-downloaded model can not be used

[mcd01]

I am facing an issue when using the transformers backend with huggingface models and was hoping that someone can provide additional insights. I am quite sure it is just a small thing on my end, but I already tried quite a bunch of combinations and none of them worked.

LocalAI version:
v2.20.1 with localai/localai:latest-gpu-nvidia-cuda-12

Environment, CPU architecture, OS, and Version:
TBD

Describe the bug
When trying to interact with the model downloaded by the backend, I get the following error:

{
  "error": {
    "code": 500,
    "message": "could not load model (no success): Unexpected err=OSError(\"Can't load the model for 'facebook/opt-125m'. If you were trying to load it from 'https://huggingface.co/models', make sure you don't have a local directory with the same name. Otherwise, make sure 'facebook/opt-125m' is the correct path to a directory containing a file named pytorch_model.bin, tf_model.h5, model.ckpt or flax_model.msgpack.\"), type(err)=<class 'OSError'>",
    "type": ""
  }

The model is apparently successfully downloaded into the models directory. The model path is also correctly configured (e.g. manually downloaded .gguf models can be used with llama-cpp backend). It is just that the docs on GPT with transformers backend seem to be out-of-date, at least I can not reproduce their results.

To Reproduce
Work through the docs on GPT for the transformers backend and try to implement the example.

name: transformers
backend: transformers
parameters:
    model: "facebook/opt-125m"
type: AutoModelForCausalLM
quantization: bnb_4bit

Make sure the backend is configured and correctly loaded:

EXTRA_BACKENDS: backend/python/transformers
Preparing backend: backend/python/transformers
make: Entering directory '/build/backend/python/transformers'
bash install.sh
Initializing libbackend for transformers
virtualenv activated
activated virtualenv has been ensured
starting requirements install for /build/backend/python/transformers/requirements.txt
Audited 4 packages in 73ms
finished requirements install for /build/backend/python/transformers/requirements.txt
make: Leaving directory '/build/backend/python/transformers'

Expected behavior
As per the docs, I would have expected the model to be downloaded (which apparently happens) and be correctly detected, loaded, and used by the respective backend.

Logs
None

Additional context
I also carefully read this issue (and posted a detailed comment there) but it didn't solve my problem.