Error loading model
When trying to run this model in LMstudio, the program returns the error: (Exit code: -36861). Unknown error. Try a different model and/or config.
The same error occurred when I tried to run Phi 3.5 vision, so I think the problem is in the visual part of the model. Although there were no problems with the work of MiniCPM-V 2.6 and Llava-v1.5-7B.
Does anyone know how to solve this problem?
Thanks! Since our model's architecture is different from traditional Llava model, LMStudio/llama.cpp may not be able to support our model. You can have a try using sdk from our company: https://github.com/NexaAI/nexa-sdk and try nexa run omnivision. Also, we host the huggingface space to test the model: https://huggingface.co./spaces/NexaAIDev/omnivlm-dpo-demo
Thanks for the help!