AayushManShrestha

Why is only the text model being fine-tuned in the Finetuning.ipynb notebook?

1

I noticed that only the text_model of Moondream was being trained. Is there a reason for this?  What will training the entire Moondream or training only the vision_encoder result...

I was trying to fine-tune Moondream2 by using LoRA. But, I got weird loss curves. Here is the link to the code: [LoRA-finetune](https://colab.research.google.com/drive/1cNSme3Tc3vynxUmov-JN3uEzbU3mbG9W?usp=sharing) Here is the loss curve when doing...

I have a Gaussian splat .ply file of an environment created using vanilla 3dgs repo. Can this custom environment splat be segmented using LangSplat?