Add Dockerfile for Blackwell GPU support (RTX 50xx)#1371
Open
TroyHernandez wants to merge 1 commit intodeepbeepmeep:mainfrom
Open
Add Dockerfile for Blackwell GPU support (RTX 50xx)#1371TroyHernandez wants to merge 1 commit intodeepbeepmeep:mainfrom
TroyHernandez wants to merge 1 commit intodeepbeepmeep:mainfrom
Conversation
Blackwell architecture (sm_120) requires: - CUDA 12.8+ (12.4 doesn't support sm_120) - PyTorch 2.7.0+ (first version with Blackwell support) Key changes from main Dockerfile: - Base image: nvidia/cuda:12.8.1-cudnn-devel-ubuntu22.04 - PyTorch: 2.7.1+cu128 - CUDA arch: 12.0 (sm_120) Tested on RTX 5060 Ti 16GB with: - LTX-2 19B Distilled: ~3.5 min per video - VACE 1.3B: ~10 min per video - Profile 4 (low VRAM mode) works correctly Co-Authored-By: Claude Opus 4.5 <[email protected]>
|
Need to make it cuda 13, to make sure it's possible to take advantage of nv fp4, also need sage 2.2+ rather than 1 |
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
Sign up for free
to join this conversation on GitHub.
Already have an account?
Sign in to comment
2 participants
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Outdated suggestions cannot be applied.This suggestion has been applied or marked resolved.Suggestions cannot be applied from pending reviews.Suggestions cannot be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.
Summary
Dockerfile.blackwellfor NVIDIA Blackwell architecture (RTX 5060 Ti, 5070, 5080, 5090)Why this is needed
Blackwell GPUs (sm_120) require:
The existing Dockerfile uses CUDA 12.4 and PyTorch 2.6, which won't recognize Blackwell GPUs.
Changes
Testing
Tested on RTX 5060 Ti (16GB VRAM) with Profile 4:
All models load and run correctly with
--profile 4.Build instructions
docker build -f Dockerfile.blackwell -t wan2gp:blackwell .🤖 Generated with Claude Code