feat: Fixed compatibility with Blackwell (RTX 50+ Series) #109
Conversation
This file contains hidden or bidirectional Unicode text that may be interpreted or compiled differently than what appears below. To review, open the file in an editor that reveals hidden Unicode characters.
Learn more about bidirectional Unicode characters
added 2 commits
August 20, 2025 19:24
- Add PowerShell equivalents for all batch files (setenv.ps1, Start.ps1, Start_Portable.ps1, etc.) - Switch from bundled Python to external Python 3.11 with virtual environment - Fix PyTorch 2.8.0+ compatibility issues by using Python 3.11 - Create install_requirements.ps1 for virtual environment package management - Update PowerShell_Usage.md with comprehensive documentation - Maintain full compatibility with existing batch file functionality - Support modern typing features (typing.Self) required by PyTorch 2.8.0+ Files added/modified: - scripts/setenv.ps1 (new) - Start.ps1 (new) - Start_Portable.ps1 (new) - Update_Portable.ps1 (new) - scripts/update_cu118.ps1 (new) - scripts/update_cu124.ps1 (new) - scripts/update_cu128.ps1 (new) - scripts/install_requirements.ps1 (new) - app/ui/core/convert_ui_to_py.ps1 (new) - PowerShell_Usage.md (new) This resolves the 'typing.Self is not valid as type argument' error by using Python 3.11+ which fully supports modern typing annotations required by PyTorch 2.8.0+ and torchvision 0.23.0+.
- Document infrastructure upgrades for modern GPU architecture support - Cover compatibility from RTX 20 series to future generations - Detail 32-bit CUDA deprecation and 64-bit environment requirements - Explain Python 3.11+ and PyTorch 2.8.0+ benefits for all GPU architectures - Provide performance improvement matrix for RTX 20/30/40/50 series - Include installation paths and testing results by GPU generation - Position changes as universal GPU compatibility improvements, not just stability fixes This document explains how the PowerShell and virtual environment upgrades benefit all modern NVIDIA GPUs by addressing fundamental architectural limitations and preparing for future GPU generations.
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
1 participant
Add this suggestion to a batch that can be applied as a single commit.
This suggestion is invalid because no changes were made to the code.
Suggestions cannot be applied while the pull request is closed.
Suggestions cannot be applied while viewing a subset of changes.
Only one suggestion per line can be applied in a batch.
Add this suggestion to a batch that can be applied as a single commit.
Applying suggestions on deleted lines is not supported.
You must change the existing code in this line in order to create a valid suggestion.
Outdated suggestions cannot be applied.
This suggestion has been applied or marked resolved.
Suggestions cannot be applied from pending reviews.
Suggestions cannot be applied on multi-line comments.
Suggestions cannot be applied while the pull request is queued to merge.
Suggestion cannot be applied right now. Please check back later.
You can see much more detail on what was upgraded and why @ https://github.com/jcwicked/VisoMaster/blob/feature-rtx-5090-update/GPU_Architecture_Upgrade_Compatibility.md
Files added/modified:
This resolves the 'typing.Self is not valid as type argument' error by using Python 3.11+ which fully supports modern typing annotations required by PyTorch 2.8.0+ and torchvision 0.23.0+.