mlx-examples: "missing 5 required positional arguments" when I try to train using lora on llama2

When I run :

python lora.py \
   --model ./mlx_model \
   --train \
   --iters 600 \
   --batch-size 1 \
   --lora-layers 4

I get the following error:

Loading pretrained model
Traceback (most recent call last):
  File "/Users/macstudiosct/projects/finetune_mlx/mlx-examples/lora/lora.py", line 297, in <module>
    model, tokenizer, _ = models.load(args.model)
  File "/Users/macstudiosct/projects/finetune_mlx/mlx-examples/lora/models.py", line 319, in load
    model_args = ModelArgs.from_dict(config)
  File "/Users/macstudiosct/projects/finetune_mlx/mlx-examples/lora/models.py", line 46, in from_dict
    return cls(
TypeError: ModelArgs.__init__() missing 5 required positional arguments: 'hidden_size', 'num_hidden_layers', 'intermediate_size', 'num_attention_heads', and 'rms_norm_eps'

I have also tried the arguments in the order mentioned in python lora.py -h, no dice

System:

M2 ultra with 128gb memory; 76 cores GPU.

Edit:

My mlx version is 0.07