Сan't add an existing model.

[PhantomKojima]

I'm trying to add a model that already exists on my computer. To do this, I enter the path to my model in the “Create a reference from local file path:” field. I entered D:\Prgrms\LM Studio\models\gpt4all\instruct and D:\Prgrms\LM Studio\models\gpt4all\instruct\mistral-7b-instruct-v0.1.Q4_0.gguf. None of the options worked and in cmd I got this:

[ParisNeo]

Which binding are you using? GGUF models are only supported by CTransformers and GPT4All bindings.

[PhantomKojima]

Im using GPT4All binding. update: I get the same thing with the CTransformers binding.

[FrigidICY3]

I encountered a similar issue, and spent the entire day attempting to resolve it. However, the system remains opaque to me. Nonetheless, to rectify this, simply copy the ".gguf" model file into the designated directory, "personal_models_path/gguf/". Upon starting the application, the model will automatically load and appear as an installed model in the "Settings/Models zoo". Select it, and it should function as intended. Use the "gpt4all" binding instead of "CTransformers" since it keeps crashing.

[PhantomKojima]

I encountered a similar issue, and spent the entire day attempting to resolve it. However, the system remains opaque to me. Nonetheless, to rectify this, simply copy the ".gguf" model file into the designated directory, "personal_models_path/gguf/". Upon starting the application, the model will automatically load and appear as an installed model in the "Settings/Models zoo". Select it, and it should function as intended. Use the "gpt4all" binding instead of "CTransformers" since it keeps crashing.

I already did this, but instead of a copy I made a hard link to the model. I can’t use the gpt4all binding because it often crashes the program, CTransformers works fine.

[FrigidICY3]

IRUTSD

I already did this, but instead of a copy I made a hard link to the model. I can’t use the gpt4all binding because it often crashes the program, CTransformers works fine.

I'm using Linux, and that's what worked for me, nonetheless still highly unstable and crashes very often, to be honest i'm using LM Studio now, cuz it's much more stable, even the beta version for linux, appreciate the hard work of ParisNeo though!

[ParisNeo]

Hi everyone. I am really sorry for the crushes. It is just a problem of time. I am working on this alone and only work at night. These days I have lots of family matters to attend to so I don't find time to focus. I'll try to see what I can do in the weekend.

[navarisun1982]

me too i have the (best for the gpu) thing and it doesn't appear at all now and i can't load the models (Couldn't load binding exllama2.)

[TurboLion]

I'm trying to add a model that already exists on my computer. To do this, I enter the path to my model in the “Create a reference from local file path:” field. I entered D:\Prgrms\LM Studio\models\gpt4all\instruct and D:\Prgrms\LM Studio\models\gpt4all\instruct\mistral-7b-instruct-v0.1.Q4_0.gguf. None of the options worked and in cmd I got this:

Had the same issue, it worked after I added double quotes around the path. So, instead of entering D:\Prgrms\..., I entered "D:\Prgrms\..."

[ParisNeo]

that happens if you have space in the path

[PhantomKojima]

Now I'm trying to add models from the D:\Prgrms\GGUF_MODELS folder. I use double quotes as written by TurboLion and no longer get an error in cmd, but after updating the list of models, models from the added folder do not appear in the list.

[navarisun1982]

Now I'm trying to add models from the D:\Prgrms\GGUF_MODELS folder. I use double quotes as written by TurboLion and no longer get an error in cmd, but after updating the list of models, models from the added folder do not appear in the list.

same here the models are not showing