Comments (4)
No we haven't investigated the additional overhead of the single methods systematically since they are all negligible (seconds) compared to the time of a single function evaluation (hours). Also note, that an optimizer's overhead strongly depends on how efficient it is implemented and which hyperparameters you use.
Having said that, since we do an exhaustive amount of function evaluations (5000), model-based approaches such as SMAC might need a few hours to generated the full trajectory in Figure 7 (even though the optimizer overhead are only a few seconds).
from nasbench.
As you said, SMAC takes a few hours to generate the full trajectory in Figure 7. This result is based TPU, right?
Which implementation of SMAC you used in your experiments?
From my experience, the open source implementation of SMAC takes a very long time(1 hour for only 1 repeat ) to generate the trajectory(around 2500 evaluations) on my laptop(Intel® Core™ i5-5200U CPU @ 2.20GHz × 4 ). So in order to compute the mean performance of 500 independent runs as a function of the estimated training time just like the Figure 7 in the origin paper, it will take 500 hours which is a very big number. Unlike SMAC, Random Search and Evolutionary Search are much faster. That's why I am concerned about the additional overhead of these algorithms.
from nasbench.
I used the implementation from here
For the comparison we parallelized all 500 runs of SMAC on 500 different cores which means in actual wall-clock time it took only a few hours (everything was on CPU).
Even though one could probably improve SMAC's optimization overhead it will never be as fast as random search, which is ok since it's made for expensive optimization problems. Keep in mind running SMAC for one hour corresponds to running the original benchmark for 2778 hours.
For the paper we used 500 runs to get very solid estimate of the optimizers' performance. However, in case you just want to play around or get some quick estimates of method's performance, you probably get away with less runs (e.g 50 or so).
from nasbench.
Anyway, nice work!
And thanks for your reply.
from nasbench.
Related Issues (20)
- How much mean test accuracy can the best architecure achieve? HOT 1
- Query neighbors of an architecture HOT 1
- y-axis in Fig 7(left) HOT 4
- Reproduce Figure 8 in the NASBench paper
- Are there any api of turning the 01 spec_matrix into model_architecture? HOT 1
- Generating a dataset (tfrecord files) HOT 1
- Not compatible with tensorflow 2.0 HOT 3
- how to create model instance using matrix?
- RMSProp epsilon=1.0, why?
- How to get the FLOPs of a model HOT 2
- Lower than expected final test accuracies for some models? HOT 1
- Data Query is non-deterministic HOT 1
- hdf5 file format
- tensorflow 2.3.1 incompatibility HOT 1
- How to extract all the architectures and metrics from the records? HOT 1
- source code for calculating the graph edit distance HOT 1
- what is "total_time" in the tfrecord file?
- is_valid is true, query function report errors HOT 1
- how to get the model spec by accuracy?
Recommend Projects
-
React
A declarative, efficient, and flexible JavaScript library for building user interfaces.
-
Vue.js
🖖 Vue.js is a progressive, incrementally-adoptable JavaScript framework for building UI on the web.
-
Typescript
TypeScript is a superset of JavaScript that compiles to clean JavaScript output.
-
TensorFlow
An Open Source Machine Learning Framework for Everyone
-
Django
The Web framework for perfectionists with deadlines.
-
Laravel
A PHP framework for web artisans
-
D3
Bring data to life with SVG, Canvas and HTML. 📊📈🎉
-
Recommend Topics
-
javascript
JavaScript (JS) is a lightweight interpreted programming language with first-class functions.
-
web
Some thing interesting about web. New door for the world.
-
server
A server is a program made to process requests and deliver data to clients.
-
Machine learning
Machine learning is a way of modeling and interpreting data that allows a piece of software to respond intelligently.
-
Visualization
Some thing interesting about visualization, use data art
-
Game
Some thing interesting about game, make everyone happy.
Recommend Org
-
Facebook
We are working to build community through open source technology. NB: members must have two-factor auth.
-
Microsoft
Open source projects and samples from Microsoft.
-
Google
Google ❤️ Open Source for everyone.
-
Alibaba
Alibaba Open Source for everyone
-
D3
Data-Driven Documents codes.
-
Tencent
China tencent open source team.
from nasbench.