Comments (7)
Try LSUV init https://github.com/ducha-aiki/LSUVinit/blob/master/tools/extra/lsuv_init.py
from squeezenet.
It turns out that I needed to reduce the learning rate. After reducing the learning rate by 10x and increasing the effective batch size by 2x, I was able to train from scratch. Less extreme measures are most likely sufficient.
@ducha-aiki thanks. LSUV does seem to have a slightly faster start; in this case, my biggest problem was the learning rate.
from squeezenet.
@liuyipei with LSUV I was able to converge with big lr. But it is good, that other ways work as well :)
See https://github.com/ducha-aiki/caffenet-benchmark/blob/master/prototxt/architectures/SqueezeNet128_lsuv.prototxt
from squeezenet.
I like how you have trainval and solver in one file. Does Caffe accept that
as-is, or did you customize Caffe to allow it?
Anyway, it looks convenient!
On Mar 3, 2016 8:50 PM, "Dmytro Mishkin" [email protected] wrote:
@liuyipei https://github.com/liuyipei with LSUV I was able to converge
with big lr. But it is good, that other ways work as well :)
See
https://github.com/ducha-aiki/caffenet-benchmark/blob/master/prototxt/architectures/SqueezeNet128_lsuv.prototxt—
Reply to this email directly or view it on GitHub
#4 (comment).
from squeezenet.
@forresti it accepts, see example in caffe master branch: https://github.com/BVLC/caffe/blob/master/examples/mnist/lenet_consolidated_solver.prototxt
from squeezenet.
@liuyipei
One more thing: I've run into a few problems with cuDNN and numerical correctness. I recommend trying a training run with cuDNN disabled, and seeing if you still get divergence.
from squeezenet.
@liuyipei
Update: We have been experimenting with solver configurations, and we have identified a configuration that converges more reliably. We just committed it to SqueezeNet-master: 0bc03d9
from squeezenet.
Related Issues (20)
- Has anyone successfully trained Squeezenet with residual connections?
- model convert HOT 1
- SqueezeNet v1.1 with Residual Connections with Dense→Sparse→Dense (DSD) Training
- Top-1 Acc=61.0% on ImageNet, without any sacrificing compared with SqueezeNet v1.1. HOT 4
- Image Width Issue HOT 1
- tensorflow- After hundreds of epochs, my total_loss stay around 0.6~0.7, and not decreased HOT 2
- 1.1 deploy.prototxt HOT 1
- SqueezeNet is slower when using GPU than when using CPU? HOT 2
- training from scratch, random seed HOT 1
- why can not get the output of the prob layer? HOT 1
- SqueezeNet training on cifar HOT 3
- The SqueezeNet deploy.caffemodel files have all 0.0 weight and bias data HOT 1
- Fine-tuning SqueezeNet HOT 2
- which label list you used HOT 1
- why not use lr_mult, decay_mult like {1, 1, 2, 0}? HOT 3
- optimization and compression
- Image normalization values HOT 1
- squeezenet v1_1 for facedetector , possible , feasable ? HOT 1
- squeezenet for speech HOT 3
- Some minor mistakes in the paper HOT 3
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 squeezenet.