marsggbo / hyperbox Goto Github PK

View Code? Open in Web Editor NEW

24.0 24.0 3.0 59.47 MB

https://hyperbox-doc.readthedocs.io/en/latest/

License: MIT License

Dockerfile 0.08% Python 99.56% Shell 0.36%

hyperbox's Introduction

Hi 👋, I'm marsggbo

A student who keeps slim and smart

Personal Website

Education

🇸🇬 Research Fellow. School of Computing (SOC), National University of Singapore, 2023-now
🇭🇰 Ph.D. Department of Computer Science, Hong Kong Baptist University, 2018-2023
🇨🇳 B.E. School of Electronic Information and Communications, Huazhong University of Science and Technology, 2014-2018

My current research focuses automated machine learning (AutoML) and distributed training and inference. Should you seek collaboration opportunities, please do not hesitate to reach out to me.

Project

AutoML Framework: https://github.com/marsggbo/hyperbox
CVPR2021 Neural Architecture Search 1st lightweight NAS challenge and moving beyond: 7th pride

Publications

He X, Chu X. MedPipe: End-to-End Joint Search of Data Augmentation Policy and Neural Architecture for 3D Medical Image Classification[C]. IEEE MedAI, 2023.
He, X., Yao, J., Wang, Y., Tang, Z., Cheung, K. C., See, S., ... & Chu, X. NAS-LID: Efficient Neural Architecture Search with Local Intrinsic Dimension. AAAI 2023.
Ying G, He X, Gao B, et al. EAGAN: Efficient Two-stage Evolutionary Architecture Search for GANs[C]. ECCV 2022. (co-first author)
He, X., Ying, G., Zhang, J., & Chu, X.. Evolutionary Multi-objective Architecture Search Framework: Application to COVID-19 3D CT Classification. MICCAI 2022.
Tang, Z., Zhang, Y., Shi, S., He, X., Han, B., & Chu, X. (2022). Virtual Homogeneity Learning: Defending against Data Heterogeneity in Federated Learning. ICML 2022.
He X, Zhao K, Chu X. AutoML: A Survey of the State-of-the-Art[J]. Knowledge-Based Systems, 2021, 212: 106622. (1000+citations)
He, X., Wang, S., Chu, X., Shi, S., Tang, J., Liu, X., Yan, C., Zhang, J., & Ding, G. Automated Model Design and Benchmarking of Deep Learning Models for COVID-19 Detection with Chest CT Scans. AAAI, 2021.
Wang Y, Wang Q, Shi S, He X, et al. Benchmarking the performance and energy efficiency of ai accelerators for ai training[C]//2020 20th IEEE/ACM International Symposium on Cluster, Cloud and Internet Computing (CCGRID). IEEE, 2020: 744-751.
He X, Wang S, Shi S, et al. Computer-Aided Clinical Skin Disease Diagnosis Using CNN and Object Detection Models[C]//2019 IEEE International Conference on Big Data (Big Data). IEEE, 2019: 4839-4844.

Preprints

He X, Wang S, Shi S, et al. Benchmarking deep learning models and automated model design for covid-19 detection with chest ct scans[J]. medRxiv, 2020.

Invited Reviewer for Journals/Conferences

IEEE Transactions on Pattern Analysis and Machine Intelligence (TPAMI)
IEEE Transactions on Medical Imaging (TMI)
IEEE Journal of Biomedical and Health Informatics (JBHI)
Expert Systems with Applications
AAAI Conference on Artificial Intelligence (AAAI) 2020/2022
European Conference on Computer Vision (ECCV) 2022
Computer Vision and Pattern Recognition Conference (CVPR) 2023
International Conference on Computer Vision (ICCV) 2023

Awards

2020/21 Computer Science Department RPg Performance Award, Hong Kong Baptist University. Link
2020/21 Best Presentation Award of 2021 PG day
2020/21 semester 1, Excellent Teaching Assistant Performance Awards (COMP 7800 Analytic Models in IT Management), Hong Kong Baptist University.
2019/20 semester 2, Excellent Teaching Assistant Performance Awards (COMP 7540 IT Management: Principles & Practice), Hong Kong Baptist University.
2019/20 semester 1, Excellent Teaching Assistant Performance Awards (COMP 7180 Quantitative Methods for Data Analytics & Artificial Intelligence), Hong Kong Baptist University.

Work/Intern Experience

09/2020-11/2020, Huawei Noah'S Ark Lab, Shenzhen.
06/2021-now, NVIDIA AI Tech Center Joint Collaboration Program.

Contact Me

⚡ Email: [email protected]
😄 My Cnblog: marsggbo
🔭 My Zhihu Column（知乎专栏）: AutoML
👯 My WeChat Official Account（微信公众号）: AutoML机器学习

hyperbox's People

Contributors

Stargazers

Watchers

Forkers

ai-jie01 overfittingstudyroom axxqwerr

hyperbox's Issues

support `NAS-Bench-101`, `NAS-Bench-201`,`NAS-Bench-301`

Error executing job with overrides: []
Traceback (most recent call last):
File "D:/kk/autoML/hyperbox/hyperbox/run.py", line 31, in main
utils.print_config(config, resolve=True)
File "C:\Users\kk\anaconda3\envs\hydra\lib\site-packages\pytorch_lightning\utilities\distributed.py", line 48, in wrapped_fn
return fn(*args, **kwargs)
File "D:\kk\autoML\hyperbox\hyperbox\hyperbox\utils\utils.py", line 150, in print_config
rich.print(tree, file=fp)
File "C:\Users\kk\anaconda3\envs\hydra\lib\site-packages\rich_init_.py", line 68, in print
return write_console.print(*objects, sep=sep, end=end)
File "C:\Users\kk\anaconda3\envs\hydra\lib\site-packages\rich\console.py", line 1615, in print
self._buffer.extend(new_segments)
File "C:\Users\kk\anaconda3\envs\hydra\lib\site-packages\rich\console.py", line 825, in exit
self._exit_buffer()
File "C:\Users\kk\anaconda3\envs\hydra\lib\site-packages\rich\console.py", line 784, in _exit_buffer
self._check_buffer()
File "C:\Users\kk\anaconda3\envs\hydra\lib\site-packages\rich\console.py", line 1836, in _check_buffer
write(line)
UnicodeEncodeError: 'gbk' codec can't encode character '\u2699' in position 0: illegal multibyte sequence
*** You may need to add PYTHONIOENCODING=utf-8 to your environment ***

Set the environment variable HYDRA_FULL_ERROR=1 for a complete stack trace.

Process finished with exit code 1

操作系统：Windows10
8月13的版本直接运行python run.py没有此问题,8月18版本直接运行python run.py运行出现该问题。在linux下没有此问题

ValueError "some key" not found in decision cache.

In Pytorch Lightning, Model will validate the network on the first time to make sure there is no bug in the validation process.

when validating the network on the first time, the _cache object in Mutator is None and Mutator is not properly initiated.

we can naively add self.mutator.reset() to initiate the Mutator object. But in some NAS algorithem, this would not work well becasue maybe some algorithm would like to run particular subnet.

So we change the logic in the forward function of OperationSpace.

old:

    def forward(self, *inputs):
        if self.is_search:
            if hasattr(self, "mutator"):
                out, mask = self.mutator.on_forward_operation_space(self, *inputs)
        else:
            out = self.choices[0](*inputs)
            mask = torch.tensor([True])
        if self.return_mask:
            return out, mask
        else:
            return out

new:

    def forward(self, *inputs):
        if self.is_search and hasattr(self, "mutator") and self.mutator._cache:
                out, mask = self.mutator.on_forward_operation_space(self, *inputs)
        else:
            out = self.choices[0](*inputs)
            mask = torch.tensor([True])
        if self.return_mask:
            return out, mask
        else:
            return out

数据集网址问题

直接运行python run.py

Downloading http://yann.lecun.com/exdb/mnist/https://ossci-datasets.s3.amazonaws.com/mnist/train-images-idx3-ubyte.gz
Failed to download (trying next):
HTTP Error 503: Service Unavailable

Downloading https://ossci-datasets.s3.amazonaws.com/mnist/https://ossci-datasets.s3.amazonaws.com/mnist/train-images-idx3-ubyte.gz
Failed to download (trying next):
HTTP Error 404: Not Found
…………
…………
最后报
RuntimeError: Error downloading https://ossci-datasets.s3.amazonaws.com/mnist/train-images-idx3-ubyte.gz

注意到问题是下载网址问题，自动将两个网址拼接起来了，导致网址错误。

操作系统：win10，linux均有此问题

解决方法：自行访问数据集的网址https://ossci-datasets.s3.amazonaws.com/mnist/train-images-idx3-ubyte.gz
并将数据集放于如图路径

再将datamodules.init.py里的所有内容注释掉，即可。不确定将__init__.py注释掉的其他后果。

Bug about tensor type mismatch

Bug Description:

Runtime Error: Input type (torch.cuda.FloatTensor) and weight type (torch.FloatTensor) should be the same

self.network = self.network.cuda() did not work

Error while running the code

Hi!
While running the run.py file i got this error.

D:\Codes\hyperbox-master\venv\lib\site-packages\hydra_internal\defaults_list.py:251: UserWarning: In 'config.yaml': Defaults list is missing _self_. See https://hydra.cc/docs/upgrades/1.0_to_1.1/default_composition_order for more information
warnings.warn(msg, UserWarning)
[2022-09-28 13:10:58] [INFO] [D:\Codes\hyperbox-master\hyperbox\utils\utils.py:99 (hyperbox.utils.utils)] Disabling python warnings! <config.ignore_warnings=True>
Error executing job with overrides: []
Traceback (most recent call last):
File "D:\Codes\hyperbox-master\hyperbox\run.py", line 36, in main
utils.print_config(config, resolve=True)
File "D:\Codes\hyperbox-master\venv\lib\site-packages\pytorch_lightning\utilities\rank_zero.py", line 32, in wrapped_fn
return fn(*args, **kwargs)
File "D:\Codes\hyperbox-master\hyperbox\utils\utils.py", line 164, in print_config
rich.print(tree, file=fp)
File "D:\Codes\hyperbox-master\venv\lib\site-packages\rich_init_.py", line 73, in print
return write_console.print(*objects, sep=sep, end=end)
File "D:\Codes\hyperbox-master\venv\lib\site-packages\rich\console.py", line 1694, in print
self._buffer.extend(new_segments)
File "D:\Codes\hyperbox-master\venv\lib\site-packages\rich\console.py", line 848, in exit
self._exit_buffer()
File "D:\Codes\hyperbox-master\venv\lib\site-packages\rich\console.py", line 806, in _exit_buffer
self._check_buffer()
File "D:\Codes\hyperbox-master\venv\lib\site-packages\rich\console.py", line 2009, in _check_buffer
write(line)
File "C:\Users\junai\miniconda3\lib\encodings\cp1252.py", line 19, in encode
return codecs.charmap_encode(input,self.errors,encoding_table)[0]
UnicodeEncodeError: 'charmap' codec can't encode character '\u2699' in position 0: character maps to
*** You may need to add PYTHONIOENCODING=utf-8 to your environment ***