lllyasviel/ControlNet

Low VRAM tests (now 8GB ok, begin to solve 6gb)

Open

#3 opened on Feb 11, 2023

View on GitHub
 (21 comments) (3 reactions) (0 assignees)Python (27,526 stars) (2,500 forks)batch import
help wanted

Description

RuntimeError: CUDA out of memory. Tried to allocate 58.00 MiB (GPU 0; 8.00 GiB total capacity; 7.14 GiB already allocated; 0 bytes free; 7.26 GiB reserved in total by PyTorch) If reserved memory is >> allocated memory try setting max_split_size_mb to avoid fragmentation. See documentation for Memory Management and PYTORCH_CUDA_ALLOC_CONF

Contributor guide

Tech stack
pythonpytorch
Domain
machine learningperformance
Issue type
performance
DifficultyEstimated implementation difficulty for a new contributor, from 1 for very small changes to 5 for expert-level work.
4
Estimated timeA rough time range for an experienced contributor to investigate, implement, test, and prepare a pull request.
3-5 days
Activity statusHow available the issue appears right now: fresh, active, stale, blocked, or waiting on maintainer input.
active
ClarityHow clearly the issue explains the expected change, acceptance criteria, and next step.
mostly clear
Prerequisites
PythonPyTorchCUDA basics
Newbie friendlinessA 1-100 score estimating how approachable this issue is for first-time contributors.
40
Research direction
Investigate memory allocation in PyTorch and identify areas where memory can be saved, such as using gradient checkpointing, reducing batch size, or optimizing model layers. Look at current memory usage with nvidia smi and PyTorch memory profiler. Explore settings like max split size mb as mentioned in the error. The goal is to make the code work on 6GB GPUs.