Autosculptor 1.7.0

New Features

Hugging Face Access User Token

Added option to add user token to get higher priority in requests (on ZeroGPU-powered spaces).

Enhancements

Updated Gradio Client File Handling

Switched from file()to new handle_file() function from Gradio Python Client last major update.

Bug Fixes

API Pipelines Endpoints

Fixed DreamGaussian client endpoint.

Full Changelog: 1.6.0...1.7.0

Autosculptor 1.6.0

New Features

Threading:

Added threading option (experimental) to allow generation to be executed in another thread.

Enhancements

Estimated Time Display:

Updating dynamically estimated generation time based during batch generation usage.

Code Structure:

Refactored code structure for a better maintainability.

Bug Fixes

Files Handling

Fixed bugs in pipelines during sending files through Gradio client.

Full Changelog: 1.5.0...1.6.0

Autosculptor 1.5.0

New Features

Custom Image Dimensions:

Added support for specifying custom dimensions for the generated images:

• New properties: image_width and image_height to set custom width and height.

Estimated Time Display:

Added an estimated time display for model generation in the UI.

Enhancements

Improved Model Generation Functions:

Refactored the model generation functions to accept custom image dimensions.

Enhanced UI:

Added an estimated_time property to display the estimated generation time based on the selected model type.

Bug Fixes

Fixed the default category value in bl_info from "Object" to "Add Mesh".

Full Changelog: 1.4.0...1.5.0

Autosculptor 1.4.0

New Features

Added Batch Generation

Added support for batch generation of 3D models.

Enhancements

Updated Dependencies Handling

Removed the ensure_gradio_installed and install_gradio functions.
Dependencies are now handled implicitly within the script.

Updated Material Handling

Added option to dislabe auto-texturing after generation.

Full Changelog: 1.3.0...1.4.0

Autosculptor 1.3.0

New Features

Added models support:

Added support for generating models using multiple generative models:

• SDXL + TripoSR: ByteDance/Hyper-SDXL-1Step-T2I combined with stabilityai/TripoSR

Full Changelog: 1.2.0...1.3.0

Autosculptor 1.2.0

New Features

Added Prompt Enhancer:

Added an option to enhance the user-provided prompt for better 3D model generation results.

New Menu Structure:

Introduced a collapsible section for advanced settings in the UI, allowing users to toggle advanced options.

Enhancements

Enhanced Context Properties:

Now imports properties from the Blender context to streamline user inputs and operations.

API Integration

Integrating MagicPrompt API:

Integrated with the MagicPrompt API to enhance prompts before generating 3D models.

Full Changelog: 1.1.0...1.2.0

Autosculptor 1.1.0

New Features

Multiple Model Support:

Added support for generating models using multiple generative models:

• SDXL + Shape-E: ByteDance/Hyper-SDXL-1Step-T2I combined with hysts/Shap-E
• SDXL + DreamGaussian: ByteDance/Hyper-SDXL-1Step-T2I combined with jiawei011/dreamgaussian
• SDXL + InstantMesh: ByteDance/Hyper-SDXL-1Step-T2I combined with TencentARC/InstantMesh

Enhancements

Refactored Code Structure:

Separated the model generation logic into individual functions:

• generate_shape_e_model
• generate_sdxl_shape_e_model
• generate_sdxl_dreamgaussian_model
• generate_sdxl_instantmesh_model
  Added generate_model method to handle model type selection and dispatch to appropriate generation function.

Error Handling Improvements:

Added detailed error messages for invalid model type and issues during model import.
Ensured proper cleanup and user feedback in case of errors during the generation process.

Full Changelog: 1.0.0...1.1.0

Autosculptor 1.0.0

Initial release with basic functionality to generate 3D models using hysts/Shap-E model.
Basic UI for setting generation parameters like prompt, seed, guidance scale, and number of inference steps.
Import and material assignment for the generated model.