A PyTorch LMS user recently opened an issue asking for community support to integrate LMS into an official version of PyTorch:
This is a good opportunity to gather any and all user testimonials and success stories to document the value of LMS in a public place. Please feel free to share your support and any thoughts there in the conversation.
PyTorch Large Model Support (LMS) is a feature in the PyTorch provided by IBM Watson Machine Learning Community Edition (WML CE) that allows the successful training of deep learning models that would otherwise exhaust GPU memory and abort with "out-of-memory" errors. LMS manages this oversubscription of GPU memory by temporarily swapping tensors to host memory when they are not needed.
One or more elements of a deep learning model can lead to GPU memory exhaustion.
These include:
- Model depth and complexity
- Base data size (for example, high-resolution images)
- Batch size
Traditionally, the solution to this problem has been to modify the model until it fits in GPU memory. This approach, however, can negatively impact accuracy – especially if concessions are made by reducing data fidelity or model complexity.
With LMS, deep learning models can scale significantly beyond what was previously possible and, ultimately, generate more accurate results.
LMS is built into the pytorch conda package so it is installed by
default when you install the GPU enabled PyTorch from WML CE.
The support is currently available in the WML CE conda channel.
For more information on this channel, how to add channels, and install
frameworks see this WML CE install documentation.
The LMS functionality is disabled by default in PyTorch and needs to be enabled before your model creates tensors. Enabling LMS is as simple as calling the enablement API at the start of your program:
import torch
torch.cuda.set_enabled_lms(True)The ManyModel.py example, found in the PyTorch LMS examples, uses synthetic random images with multiple models provided by PyTorch's torchvision to allow users a fast hands-on experience with LMS. The example allows users to change the image size, explore auto-tuning, and manually set the LMS tunable parameters on various model architectures.
If you are utilizing a single GPU it is recommended to use NUMA pinning to pin
the process to the CPU and memory that is on the same system socket as the
GPU being used. Pinning the process allows the fastest connection paths between
system memory and GPU memory, which reduces the training or inferencing time.
WML CE includes the numactl utility that can be used to do this pinning. It
can be installed with the conda install numactl command. The following
example shows how to specify a single GPU to be used and how to pin the
process to use the CPU cores and memory that are on the same socket
as the specified GPU:
export CUDA_VISIBLE_DEVICES=0
numactl --cpunodebind=0 --membind=0 python train.pyIt is recommended to use Horovod distribution when using more than one GPU because Horovod creates a separate process per GPU and automatically sets the process have socket affinity with the GPU which allows the fastest connection paths between system memory and GPU memory, which reduces the training or inferencing time.
LMS adds a few statistics to the GPU memory statistics API such as the distribution of allocation sources (free-list, cudaMalloc, LMS reclaim, etc), the amount of memory swapped, and more. For more information on the new statistics and examples of their usage see the PyTorch LMS examples.
The patches directory contains git patches for the LMS code. The file names correspond to tag levels in the PyTorch source. To build PyTorch from source with Large Model Support, check out the specific PyTorch git tag and then apply the corresponding PyTorch Large Model Support patch file.
For example:
git clone https://github.com/pytorch/pytorch
cd pytorch
git checkout v1.4.0
git am /pytorch-large-model-support/patches/pytorch_v1.4.0_large_model_support.patchIf you want to contribute to PyTorch Large Model Support please read the contribution guidelines.