gpt4all

gpt4all copied to clipboard

Added the option to modify the default server port in the application settings

#1635

Checklist before requesting a review

• [x] I have performed a self-review of my code.
• [ ] If it is a core feature, I have added thorough tests.
• [ ] I have added thorough documentation for my code.
• [ ] I have tagged PR with relevant project labels. I acknowledge that a PR without labels may be dismissed.
• [ ] If this PR addresses a bug, I have provided both a screenshot/video of the original bug and the working solution.

Notes

For the configuration to take effect, you need to restart the application

Could you provide the original input image so that we could perform further analysis? Thanks.

give me your email address. I can send you my test images

Thanks for your test images.

• I notice that the areas around the given prompt points are quite similar to that of background, i.e., both are in white, which make the segmentation of such mask quite difficult. Segmenting the main area of the image by TinySAM obtains reasonable results:
• You can also provide additional negative prompt points to obtain the following result:

input_point = np.array([[1300, 100], [200, 300], [2400, 300], [1300, 1100], [100, 100], [2500, 1200]])
input_label = np.array([1, 1, 1, 1, 0, 0])

• Your above results of HQ-SAM are obtained with a relative large model, i.e., ViT-L backbone, which is much larger than TinySAM. The Light HQ-SAM obtains the following result:

yes,i've got better results with negative points,thank you for your reply