samsunglabs / eagle Goto Github PK
View Code? Open in Web Editor NEWMeasuring and predicting on-device metrics (latency, power, etc.) of machine learning models
License: Apache License 2.0
Measuring and predicting on-device metrics (latency, power, etc.) of machine learning models
License: Apache License 2.0
I've got logs below while installing eagle:
$ pip install eagle
Collecting eagle
Using cached eagle-0.2.tar.gz (417 kB)
ERROR: Command errored out with exit status 1:
command: /home/andy/anaconda3/envs/pytorch/bin/python -c 'import io, os, sys, setuptools, tokenize; sys.argv[0] = '"'"'/tmp/pip-install-vw3svb83/eagle_ca7b4cba24b942c19a8c35b503fdfc19/setup.py'"'"'; __file__='"'"'/tmp/pip-install-vw3svb83/eagle_ca7b4cba24b942c19a8c35b503fdfc19/setup.py'"'"';f = getattr(tokenize, '"'"'open'"'"', open)(__file__) if os.path.exists(__file__) else io.StringIO('"'"'from setuptools import setup; setup()'"'"');code = f.read().replace('"'"'\r\n'"'"', '"'"'\n'"'"');f.close();exec(compile(code, __file__, '"'"'exec'"'"'))' egg_info --egg-base /tmp/pip-pip-egg-info-_uca8aqj
cwd: /tmp/pip-install-vw3svb83/eagle_ca7b4cba24b942c19a8c35b503fdfc19/
Complete output (8 lines):
Traceback (most recent call last):
File "<string>", line 1, in <module>
File "/tmp/pip-install-vw3svb83/eagle_ca7b4cba24b942c19a8c35b503fdfc19/setup.py", line 3, in <module>
import ez_setup
File "/tmp/pip-install-vw3svb83/eagle_ca7b4cba24b942c19a8c35b503fdfc19/ez_setup/__init__.py", line 176
print "Setuptools version",version,"or greater has been installed."
^
SyntaxError: Missing parentheses in call to 'print'. Did you mean print("Setuptools version",version,"or greater has been installed.")?
----------------------------------------
WARNING: Discarding https://files.pythonhosted.org/packages/dd/5c/4318a37b78339ab35e8340d51ddcb76f8005261dffcd5d1e9ab7cec0b5cd/eagle-0.2.tar.gz#sha256=7af5dffd0f1f2b5c46dcc19f5b5a47e96dc3ce387f1de05199df249e8d2bed59 (from https://pypi.org/simple/eagle/). Command errored out with exit status 1: python setup.py egg_info Check the logs for full command output.
ERROR: Could not find a version that satisfies the requirement eagle (from versions: 0.2)
ERROR: No matching distribution found for eagle
It shows that there's a print syntax error in init.py line 176
The syntax of print "Setuptools version",version,"or greater has been installed." is compatible for python2 but not python3.
Please help to download the package.
I've been working with the predictor for a while, and something I've realized is that I'm not quite sure what the ±0.1,0.01,etc. accuracy results mean in the final log. I know about top-1 accuracy in CIFAR10/100, but I'm not sure what the ± part is supposed to symbolize here. I looked at the corresponding paper, but it doesn't seem to describe the results in these terms.
Firstly, I'd like to thank the authors for making this NAS benchmark public - such works really contribute to HW-Aware NAS research.
I had a query regarding the data for DARTS and NasBench101 spaces - I noticed that the project page mentioned these are coming soon - and just wanted to check regarding the same. Any information would really help :)
Pickle is a security issue.
Hi, I didn't find accuracy dataset, seems the Google Drive link only contains latency dataset.
To compare the performance of the binary relation predictor with our proposed predictor, can you share the pre-trained desktop GPU latency predictor and FLOPS predictor?
When using the sample nasbench201 predictor commands, I keep getting this error:
ValueError: Not enough points in the dataset: expected at least 15284 but got 14664
I can't figure out what's assigning this point_limit value, since it doesn't seem to show up in any of the code.
Hello,
I downloaded the pickle files and tested them as shown below:
>>> [len(pickle.load(open(x,'rb'))) for x in os.listdir('./') if x.__contains__(".pickle")]
>>>[15284, 15261, 15252, 15284, 15261, 15284, 15261, 15252, 15284, 15284, 15284, 15284, 15284]
Why are there < 15625 architectures? Am I missing something?
Thanks!
Hi guys,
thanks for the nice dataset. In the paper you report 6 devices, but in the readme you list 12 files.
embedded-gpu-jetson-nono-fp16.pickle
and embedded-gpu-jetson-nono-fp32.pickle
, i.e. what does fp
stand for?eagle/eagle/predictors/__main__.py
Line 449 in 355f3c7
This results is a "search trace" which we later feed into our generic NAS toolkit, the toolkit selects models according to the provided search trace, querying a standard NAS-Bench-201 dataset to obtain accuracy values
As mentioned above in readme, I didn't find the generic NAS toolkit, so I can't reproduce the accuracy results in figure2, figure 3, figure 5, figure 6, table 3 in paper, could you please tell me how to reproduce the nas accuracy results in those figures?
Is there a way to convert the structure of the generated models to a readable textfile? Currently when I try to do so the output is a bunch of symbols.
I'm getting this error when I try to run results_calc.py on the combined pickle dataset. Not sure why it's happening now since I was able to run it before.
nasbench-201 contains networks same as the above figure, same cells connect to form a long network, but I notice that in brp-nas code, adjacency matrix and feature matrix only consider the condition in one cell:
I wonder why only considering one cell's condition can predict latency and accuracy in such a long network, does the latency dataset only measure latency in one cell instead of the whole network?
https://github.com/thomasccp/eagle/blob/b9833fedaa203da012fea136d2b4e1446bc08249/setup.py#L43
I guess here misses a quote at the end of the line?
A declarative, efficient, and flexible JavaScript library for building user interfaces.
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
An Open Source Machine Learning Framework for Everyone
The Web framework for perfectionists with deadlines.
A PHP framework for web artisans
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
Some thing interesting about web. New door for the world.
A server is a program made to process requests and deliver data to clients.
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
Some thing interesting about visualization, use data art
Some thing interesting about game, make everyone happy.
We are working to build community through open source technology. NB: members must have two-factor auth.
Open source projects and samples from Microsoft.
Google ❤️ Open Source for everyone.
Alibaba Open Source for everyone
Data-Driven Documents codes.
China tencent open source team.