Error when loading model: ValueError: Received 240 parameters not in model
#3
by
felipemoran - opened
Hey, I'm trying to load this model with LM Studio but I'm getting this error. I tried downloading the model via the tool and separately with hf but both result in the same error. Any ideas what I could be doing wrong?
I'm on a M3 Max with 128GB and able to run other MLX models (like this).
2026-03-15 20:53:08 [DEBUG]
lmstudio-llama-cpp: failed to load model. Error: Error when loading model: ValueError: Received 240 parameters not in model:
backbone.layers.1.mixer.fc1_latent_proj.biases,
backbone.layers.1.mixer.fc1_latent_proj.scales,
backbone.layers.1.mixer.fc1_latent_proj.weight,
backbone.layers.1.mixer.fc2_latent_proj.biases,
backbone.layers.1.mixer.fc2_latent_proj.scales,
backbone.layers.1.mixer.fc2_latent_proj.weight,
backbone.layers.10.mixer.fc1_latent_proj.biases,
backbone.layers.10.mixer.fc1_latent_proj.scales,
backbone.layers.10.mixer.fc1_latent_proj.weight,
backbone.layers.10.mixer.fc2_latent_proj.biases,
backbone.layers.10.mixer.fc2_latent_proj.scales,
backbone.layers.10.mixer.fc2_latent_proj.weight,
backbone.layers.12.mixer.fc1_latent_proj.biases,
backbone.layers.12.mixer.fc1_latent_proj.scales,
backbone.layers.12.mixer.fc1_latent_proj.weight,
backbone.layers.12.mixer.fc2_latent_proj.biases,
backbone.layers.12.mixer.fc2_latent_proj.scales,
backbone.layers.12.mixer.fc2_latent_proj.weight,
backbone.layers.14.mixer.fc1_latent_proj.biases,
backbone.layers.14.mixer.fc1_latent_proj.scales,
backbone.layers.14.mixer.fc1_latent_proj.weight,
backbone.layers.14.mixer.fc2_latent_proj.biases,
backbone.layers.14.mixer.fc2_latent_proj.scales,
backbone.layers.14.mixer.fc2_latent_proj.weight,
backbone.layers.17.mixer.fc1_latent_proj.biases,
backbone.layers.17.mixer.fc1_latent_proj.scales,
backbone.layers.17.mixer.fc1_latent_proj.weight,
backbone.layers.17.mixer.fc2_latent_proj.biases,
backbone.layers.17.mixer.fc2_latent_proj.scales,
backbone.layers.17.mixer.fc2_latent_proj.weight,
backbone.layers.19.mixer.fc1_latent_proj.biases,
backbone.layers.19.mixer.fc1_latent_proj.scales,
backbone.layers.19.mixer.fc1_latent_proj.weight,
backbone.layers.19.mixer.fc2_latent_proj.biases,
backbone.layers.19.mixer.fc2_latent_proj.scales,
backbone.layers.19.mixer.fc2_latent_proj.weight,
backbone.layers.21.mixer.fc1_latent_proj.biases,
backbone.layers.21.mixer.fc1_latent_proj.scales,
backbone.layers.21.mixer.fc1_latent_proj.weight,
backbone.layers.21.mixer.fc2_latent_proj.biases,
backbone.layers.21.mixer.fc2_latent_proj.scales,
backbone.layers.21.mixer.fc2_latent_proj.weight,
backbone.layers.23.mixer.fc1_latent_proj.biases,
backbone.layers.23.mixer.fc1_latent_proj.scales,
backbone.layers.23.mixer.fc1_latent_proj.weight,
backbone.layers.23.mixer.fc2_latent_proj.biases,
backbone.layers.23.mixer.fc2_latent_proj.scales,
backbone.layers.23.mixer.fc2_latent_proj.weight,
backbone.layers.26.mixer.fc1_latent_proj.biases,
backbone.layers.26.mixer.fc1_latent_proj.scales,
backbone.layers.26.mixer.fc1_latent_proj.weight,
backbone.layers.26.mixer.fc2_latent_proj.biases,
backbone.layers.26.mixer.fc2_latent_proj.scales,
backbone.layers.26.mixer.fc2_latent_proj.weight,
backbone.layers.28.mixer.fc1_latent_proj.biases,
backbone.layers.28.mixer.fc1_latent_proj.scales,
backbone.layers.28.mixer.fc1_latent_proj.weight,
backbone.layers.28.mixer.fc2_latent_proj.biases,
backbone.layers.28.mixer.fc2_latent_proj.scales,
backbone.layers.28.mixer.fc2_latent_proj.weight,
backbone.layers.3.mixer.fc1_latent_proj.biases,
backbone.layers.3.mixer.fc1_latent_proj.scales,
backbone.layers.3.mixer.fc1_latent_proj.weight,
backbone.layers.3.mixer.fc2_latent_proj.biases,
backbone.layers.3.mixer.fc2_latent_proj.scales,
backbone.layers.3.mixer.fc2_latent_proj.weight,
backbone.layers.30.mixer.fc1_latent_proj.biases,
backbone.layers.30.mixer.fc1_latent_proj.scales,
backbone.layers.30.mixer.fc1_latent_proj.weight,
backbone.layers.30.mixer.fc2_latent_proj.biases,
backbone.layers.30.mixer.fc2_latent_proj.scales,
backbone.layers.30.mixer.fc2_latent_proj.weight,
backbone.layers.32.mixer.fc1_latent_proj.biases,
backbone.layers.32.mixer.fc1_latent_proj.scales,
backbone.layers.32.mixer.fc1_latent_proj.weight,
backbone.layers.32.mixer.fc2_latent_proj.biases,
backbone.layers.32.mixer.fc2_latent_proj.scales,
backbone.layers.32.mixer.fc2_latent_proj.weight,
backbone.layers.34.mixer.fc1_latent_proj.biases,
backbone.layers.34.mixer.fc1_latent_proj.scales,
backbone.layers.34.mixer.fc1_latent_proj.weight,
backbone.layers.34.mixer.fc2_latent_proj.biases,
backbone.layers.34.mixer.fc2_latent_proj.scales,
backbone.layers.34.mixer.fc2_latent_proj.weight,
backbone.layers.37.mixer.fc1_latent_proj.biases,
backbone.layers.37.mixer.fc1_latent_proj.scales,
backbone.layers.37.mixer.fc1_latent_proj.weight,
backbone.layers.37.mixer.fc2_latent_proj.biases,
backbone.layers.37.mixer.fc2_latent_proj.scales,
backbone.layers.37.mixer.fc2_latent_proj.weight,
backbone.layers.39.mixer.fc1_latent_proj.biases,
backbone.layers.39.mixer.fc1_latent_proj.scales,
backbone.layers.39.mixer.fc1_latent_proj.weight,
backbone.layers.39.mixer.fc2_latent_proj.biases,
backbone.layers.39.mixer.fc2_latent_proj.scales,
backbone.layers.39.mixer.fc2_latent_proj.weight,
backbone.layers.41.mixer.fc1_latent_proj.biases,
backbone.layers.41.mixer.fc1_latent_proj.scales,
backbone.layers.41.mixer.fc1_latent_proj.weight,
backbone.layers.41.mixer.fc2_latent_proj.biases,
backbone.layers.41.mixer.fc2_latent_proj.scales,
backbone.layers.41.mixer.fc2_latent_proj.weight,
backbone.layers.43.mixer.fc1_latent_proj.biases,
backbone.layers.43.mixer.fc1_latent_proj.scales,
backbone.layers.43.mixer.fc1_latent_proj.weight,
backbone.layers.43.mixer.fc2_latent_proj.biases,
backbone.layers.43.mixer.fc2_latent_proj.scales,
backbone.layers.43.mixer.fc2_latent_proj.weight,
backbone.layers.45.mixer.fc1_latent_proj.biases,
backbone.layers.45.mixer.fc1_latent_proj.scales,
backbone.layers.45.mixer.fc1_latent_proj.weight,
backbone.layers.45.mixer.fc2_latent_proj.biases,
backbone.layers.45.mixer.fc2_latent_proj.scales,
backbone.layers.45.mixer.fc2_latent_proj.weight,
backbone.layers.48.mixer.fc1_latent_proj.biases,
backbone.layers.48.mixer.fc1_latent_proj.scales,
backbone.layers.48.mixer.fc1_latent_proj.weight,
backbone.layers.48.mixer.fc2_latent_proj.biases,
backbone.layers.48.mixer.fc2_latent_proj.scales,
backbone.layers.48.mixer.fc2_latent_proj.weight,
backbone.layers.5.mixer.fc1_latent_proj.biases,
backbone.layers.5.mixer.fc1_latent_proj.scales,
backbone.layers.5.mixer.fc1_latent_proj.weight,
backbone.layers.5.mixer.fc2_latent_proj.biases,
backbone.layers.5.mixer.fc2_latent_proj.scales,
backbone.layers.5.mixer.fc2_latent_proj.weight,
backbone.layers.50.mixer.fc1_latent_proj.biases,
backbone.layers.50.mixer.fc1_latent_proj.scales,
backbone.layers.50.mixer.fc1_latent_proj.weight,
backbone.layers.50.mixer.fc2_latent_proj.biases,
backbone.layers.50.mixer.fc2_latent_proj.scales,
backbone.layers.50.mixer.fc2_latent_proj.weight,
backbone.layers.52.mixer.fc1_latent_proj.biases,
backbone.layers.52.mixer.fc1_latent_proj.scales,
backbone.layers.52.mixer.fc1_latent_proj.weight,
backbone.layers.52.mixer.fc2_latent_proj.biases,
backbone.layers.52.mixer.fc2_latent_proj.scales,
backbone.layers.52.mixer.fc2_latent_proj.weight,
backbone.layers.54.mixer.fc1_latent_proj.biases,
backbone.layers.54.mixer.fc1_latent_proj.scales,
backbone.layers.54.mixer.fc1_latent_proj.weight,
backbone.layers.54.mixer.fc2_latent_proj.biases,
backbone.layers.54.mixer.fc2_latent_proj.scales,
backbone.layers.54.mixer.fc2_latent_proj.weight,
backbone.layers.56.mixer.fc1_latent_proj.biases,
backbone.layers.56.mixer.fc1_latent_proj.scales,
backbone.layers.56.mixer.fc1_latent_proj.weight,
backbone.layers.56.mixer.fc2_latent_proj.biases,
backbone.layers.56.mixer.fc2_latent_proj.scales,
backbone.layers.56.mixer.fc2_latent_proj.weight,
backbone.layers.59.mixer.fc1_latent_proj.biases,
backbone.layers.59.mixer.fc1_latent_proj.scales,
backbone.layers.59.mixer.fc1_latent_proj.weight,
backbone.layers.59.mixer.fc2_latent_proj.biases,
backbone.layers.59.mixer.fc2_latent_proj.scales,
backbone.layers.59.mixer.fc2_latent_proj.weight,
backbone.layers.61.mixer.fc1_latent_proj.biases,
backbone.layers.61.mixer.fc1_latent_proj.scales,
backbone.layers.61.mixer.fc1_latent_proj.weight,
backbone.layers.61.mixer.fc2_latent_proj.biases,
backbone.layers.61.mixer.fc2_latent_proj.scales,
backbone.layers.61.mixer.fc2_latent_proj.weight,
backbone.layers.63.mixer.fc1_latent_proj.biases,
backbone.layers.63.mixer.fc1_latent_proj.scales,
backbone.layers.63.mixer.fc1_latent_proj.weight,
backbone.layers.63.mixer.fc2_latent_proj.biases,
backbone.layers.63.mixer.fc2_latent_proj.scales,
backbone.layers.63.mixer.fc2_latent_proj.weight,
backbone.layers.65.mixer.fc1_latent_proj.biases,
backbone.layers.65.mixer.fc1_latent_proj.scales,
backbone.layers.65.mixer.fc1_latent_proj.weight,
backbone.layers.65.mixer.fc2_latent_proj.biases,
backbone.layers.65.mixer.fc2_latent_proj.scales,
backbone.layers.65.mixer.fc2_latent_proj.weight,
backbone.layers.67.mixer.fc1_latent_proj.biases,
backbone.layers.67.mixer.fc1_latent_proj.scales,
backbone.layers.67.mixer.fc1_latent_proj.weight,
backbone.layers.67.mixer.fc2_latent_proj.biases,
backbone.layers.67.mixer.fc2_latent_proj.scales,
backbone.layers.67.mixer.fc2_latent_proj.weight,
backbone.layers.70.mixer.fc1_latent_proj.biases,
backbone.layers.70.mixer.fc1_latent_proj.scales,
backbone.layers.70.mixer.fc1_latent_proj.weight,
backbone.layers.70.mixer.fc2_latent_proj.biases,
backbone.layers.70.mixer.fc2_latent_proj.scales,
backbone.layers.70.mixer.fc2_latent_proj.weight,
backbone.layers.72.mixer.fc1_latent_proj.biases,
backbone.layers.72.mixer.fc1_latent_proj.scales,
backbone.layers.72.mixer.fc1_latent_proj.weight,
backbone.layers.72.mixer.fc2_latent_proj.biases,
backbone.layers.72.mixer.fc2_latent_proj.scales,
backbone.layers.72.mixer.fc2_latent_proj.weight,
backbone.layers.74.mixer.fc1_latent_proj.biases,
backbone.layers.74.mixer.fc1_latent_proj.scales,
backbone.layers.74.mixer.fc1_latent_proj.weight,
backbone.layers.74.mixer.fc2_latent_proj.biases,
backbone.layers.74.mixer.fc2_latent_proj.scales,
backbone.layers.74.mixer.fc2_latent_proj.weight,
backbone.layers.76.mixer.fc1_latent_proj.biases,
backbone.layers.76.mixer.fc1_latent_proj.scales,
backbone.layers.76.mixer.fc1_latent_proj.weight,
backbone.layers.76.mixer.fc2_latent_proj.biases,
backbone.layers.76.mixer.fc2_latent_proj.scales,
backbone.layers.76.mixer.fc2_latent_proj.weight,
backbone.layers.79.mixer.fc1_latent_proj.biases,
backbone.layers.79.mixer.fc1_latent_proj.scales,
backbone.layers.79.mixer.fc1_latent_proj.weight,
backbone.layers.79.mixer.fc2_latent_proj.biases,
backbone.layers.79.mixer.fc2_latent_proj.scales,
backbone.layers.79.mixer.fc2_latent_proj.weight,
backbone.layers.8.mixer.fc1_latent_proj.biases,
backbone.layers.8.mixer.fc1_latent_proj.scales,
backbone.layers.8.mixer.fc1_latent_proj.weight,
backbone.layers.8.mixer.fc2_latent_proj.biases,
backbone.layers.8.mixer.fc2_latent_proj.scales,
backbone.layers.8.mixer.fc2_latent_proj.weight,
backbone.layers.81.mixer.fc1_latent_proj.biases,
backbone.layers.81.mixer.fc1_latent_proj.scales,
backbone.layers.81.mixer.fc1_latent_proj.weight,
backbone.layers.81.mixer.fc2_latent_proj.biases,
backbone.layers.81.mixer.fc2_latent_proj.scales,
backbone.layers.81.mixer.fc2_latent_proj.weight,
backbone.layers.83.mixer.fc1_latent_proj.biases,
backbone.layers.83.mixer.fc1_latent_proj.scales,
backbone.layers.83.mixer.fc1_latent_proj.weight,
backbone.layers.83.mixer.fc2_latent_proj.biases,
backbone.layers.83.mixer.fc2_latent_proj.scales,
backbone.layers.83.mixer.fc2_latent_proj.weight,
backbone.layers.85.mixer.fc1_latent_proj.biases,
backbone.layers.85.mixer.fc1_latent_proj.scales,
backbone.layers.85.mixer.fc1_latent_proj.weight,
backbone.layers.85.mixer.fc2_latent_proj.biases,
backbone.layers.85.mixer.fc2_latent_proj.scales,
backbone.layers.85.mixer.fc2_latent_proj.weight,
backbone.layers.87.mixer.fc1_latent_proj.biases,
backbone.layers.87.mixer.fc1_latent_proj.scales,
backbone.layers.87.mixer.fc1_latent_proj.weight,
backbone.layers.87.mixer.fc2_latent_proj.biases,
backbone.layers.87.mixer.fc2_latent_proj.scales,
backbone.layers.87.mixer.fc2_latent_proj.weight.
<same as above but ending with the following>
At:
/Users/USER/.lmstudio/extensions/backends/vendor/_amphibian/app-mlx-generate-mac14-arm64@19/lib/python3.11/site-packages/mlx/nn/layers/base.py(185): load_weights
/Users/USER/.lmstudio/extensions/backends/vendor/_amphibian/app-mlx-generate-mac14-arm64@19/lib/python3.11/site-packages/mlx_lm/utils.py(403): load_model
/Users/USER/.lmstudio/extensions/backends/vendor/_amphibian/app-mlx-generate-mac14-arm64@19/lib/python3.11/site-packages/mlx_lm/utils.py(479): load
/Users/USER/.lmstudio/extensions/backends/vendor/_amphibian/app-mlx-generate-mac14-arm64@19/lib/python3.11/site-packages/mlx_engine/generate.py(205): is_batchable
/Users/USER/.lmstudio/extensions/backends/vendor/_amphibian/app-mlx-generate-mac14-arm64@19/lib/python3.11/site-packages/mlx_engine/generate.py(230): load_model
Hi, it's not currently supported by LM, best to contact their support to see when they'll implement it.
In the meantime, it's available on Inferencer.
Got it, thanks for the suggestion! Got it working beautifully right on the first try.