Phi 3.5 Vision Instruct Fails to Load "Trust remote code" error

[YorkieDev]

mlx-community/Phi-3.5-vision-instruct-4bit fails to load in the current MLX build of LM Studio.

🥲 Failed to load the model

Failed to load model

Error when loading model: ValueError: Loading /Users/name/.cache/lm-studio/models/mlx-community/Phi-3.5-vision-instruct-4bit requires you to execute the configuration file in that repo on your local machine. Make sure you have read the code there to avoid malicious use, then set the option `trust_remote_code=True` to remove this error.

LM Studio version: 0.3.5(Build 4)
Sonoma 14.3/128gb M3 Max.

[neilmehta24]

Right now, by default we disable trusting remote code. We will need to update the LM Studio app UI so that users can configure their trust settings. Will update this issue when we are ready to ship this feature.

[stelterlab]

Teuken 7B has the same problem. It also needs to trust remote code.

Would love to see this feature in the future, too.

[dwdcth]

mlx-community/Molmo-7B-D-0924-4bit has the same problem,

[Zer0xPoint]

Florence-2-large-ft-bf16 also

[emilianomonzaia]

@neilmehta24 what is the schedule to roll it out?

We're facing the same with Molmo-7B-D-0924-4bit, which is core for our next implementation.

[delphi1989]

any updates ? thanks QQ

[9Somboon]

up

[rizrmd]

how to do this without UI ?

[mscinar]

up

[jason-engage]

up

[dinaftc]

up

[ah3243]

having the same problem with phi 3.5, any updates?

[wmerritt72]

Has there been any updates on this? If not, what is the workaround for LM studio to allow loading og the models?

[CaliLuke]

same here, I'm having issues loading Florence and I can't figure out where to put the parameter to load the model using the UI.