This repository has been archived by the owner on Sep 18, 2024. It is now read-only.
Add OpEvo example #2549
Merged
Add OpEvo example #2549
Conversation
This file contains 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
|
@gxiaotian could you write description of this pr, which makes it easy for others to review |
|
@QuanluZhang Description added. Would like to add more if needed. Thanks. |
QuanluZhang
reviewed
Jun 22, 2020
README_zh_CN.md
Outdated
| 代码库遵循 [MIT 许可协议](LICENSE) |
There was a problem hiding this comment.
please revert the changes to chinese doc
There was a problem hiding this comment.
I didn't change this file in fact. I believe there might be some invisible characters modified by the editor. I will try to figure it out.
QuanluZhang
reviewed
Jun 22, 2020
|
|
||
| Abundant applications raise the demands of training and inference deep neural networks (DNNs) efficiently on diverse hardware platforms ranging from cloud servers to embedded devices. Moreover, computational graph-level optimization of deep neural network, like tensor operator fusion, may introduce new tensor operators. Thus, manually optimized tensor operators provided by hardware-specific libraries have limitations in terms of supporting new hardware platforms or supporting new operators, so automatically optimizing tensor operators on diverse hardware platforms is essential for large-scale deployment and application of deep learning technologies in the real-world problems. | ||
|
|
||
| Tensor operator optimization is substantially a combinatorial optimization problem. The objective function is the performance of a tensor operator on specific hardware platform, which should be maximized with respect to the hyper-parameters of corresponding device code, such as how to tile a matrix or whether to unroll a loop. This example illustrates how to automatically tune tensor operators with NNI. Three tuning algorithms, OpEvo, G-BFS and N-A2C are provided. Please refer to [OpEvo: An Evolutionary Method for Tensor Operator Optimization](https://128.84.21.199/abs/2006.05664) for detailed explanation about these algorithms. |
There was a problem hiding this comment.
could you use arxiv link for the paper, not the IP one?
QuanluZhang
reviewed
Jun 22, 2020
|
|
||
| ## Run Experiments: | ||
|
|
||
| Three representative kinds of tensor operators, matrix multiplication, batched matrix multiplication and 2D convolution, are chosen from BERT and AlexNet, and tuned with NNI. The `Trial` code for all tensor operators is `/root/compiler_auto_tune_stable.py`, and `Search Space` files and `config` files for each tuning algorithm locate in `/root/experiments/`, which are categorized by tensor operators. Here `/root` refers to the root of the container. |
There was a problem hiding this comment.
"matrix multiplication, batched matrix multiplication and 2D convolution" -> "matrix multiplication, batched matrix multiplication and 2D convolution"
QuanluZhang
reviewed
Jun 22, 2020
|
|
||
| Three representative kinds of tensor operators, matrix multiplication, batched matrix multiplication and 2D convolution, are chosen from BERT and AlexNet, and tuned with NNI. The `Trial` code for all tensor operators is `/root/compiler_auto_tune_stable.py`, and `Search Space` files and `config` files for each tuning algorithm locate in `/root/experiments/`, which are categorized by tensor operators. Here `/root` refers to the root of the container. | ||
|
|
||
| For tuning the operators of matrix multiplication, please run below commands from `/root`: |
There was a problem hiding this comment.
how to start the container?
There was a problem hiding this comment.
got it. make cuda-env will directly start the container, right?
QuanluZhang
approved these changes
Jun 22, 2020
|
looks great to me. |
liuzhe-lz
approved these changes
Jun 30, 2020
chicm-ms
added a commit
to chicm-ms/nni
that referenced
this pull request
Jun 30, 2020
Add OpEvo example (microsoft#2549)
Sign up for free
to subscribe to this conversation on GitHub.
Already have an account?
Sign in.
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.
This pr aims to provide examples of tuning tensor operators with NNI. Three tuning algorithms, OpEvo, G-BFS and N-A2C are implemented with NNI. All examples locate at
examples/trials/systems/opevo/, and a doc filedocs/en_US/TrialExample/OpEvoExamples.mdis also provided for demonstrating how to set up experiment environment and run these experiments. Please refer to the doc for detailed explanation of this pr.