Comments (11)
Recently I am cleaning up my code, I will try my best to find back the data and model, because they are indeed private, but maybe I can publish a tiny set of it. I will tell you if I success. Really sorry about that
from cnn-lstm-ctc.
thanks very much!can u publish a trained model?I think it wont reveal any secret,I am a undergraduate student,I wont use it for any commercial activity. I just want to study your code and re-implementation by torch with baidu warp-ctc.if u can give me a complete model and a tiny data set to test it I will accelerate my study. best wishes!
在2016年07月14日 21:44,Aaron Chen 写道:Recently I am cleaning up my code, I will try my best to find back the data and model, because they are indeed private, but maybe I can publish a tiny set of it. I will tell you if I success. Really sorry about that
—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or mute the thread.
from cnn-lstm-ctc.
I'm just curious, doesn't warp-ctc faster and more accurate than my Theano implementation? Why need my code?
from cnn-lstm-ctc.
I just started deep learning, before I have learned cnn and use keras to break captcha,so I have some knowledage about cnn,but I donot know much about lstm and ctc,I want to use your code to study it,and learn how to combine cnn lstm and ctc together
在2016年07月14日 22:13,Aaron Chen 写道:I'm just curious, doesn't warp-ctc faster and more accurate than my Theano implementation? Why need my code?
—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or mute the thread.
from cnn-lstm-ctc.
ok, I see, but my ctc implementation is a little bit complicated, due to supporting batch training, maybe I can give more comments. I need time to clean up these code.
from cnn-lstm-ctc.
can u give me your model 99.pkl file?and a tiny of data set used to test. I promise just for study,my computer's graphics card is gtx 650,its too slow to train a very large data set,thanks for your help!
在2016年07月14日 22:20,Aaron Chen 写道:ok, I see, but my ctc implementation is a little bit complicated, due to supporting batch training, maybe I can give more comments. I need time to clean up these code.
—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or mute the thread.
from cnn-lstm-ctc.
Maybe I need to train a new one for you, I loss my original model.
from cnn-lstm-ctc.
if u do this,I really appreciate u!I will wait for your good news!I am deeply grateful!
在2016年07月14日 22:26,Aaron Chen 写道:Maybe I need to train a new one for you, I loss my original model.
—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or mute the thread.
from cnn-lstm-ctc.
I saw u have uploaded a tiny of dataset,thanks a lot for that,but do u have a trained model?10k images can not train a good model
在2016年07月14日 22:26,Aaron Chen 写道:Maybe I need to train a new one for you, I loss my original model.
—You are receiving this because you authored the thread.Reply to this email directly, view it on GitHub, or mute the thread.
from cnn-lstm-ctc.
It is still training, need some time. This task is simple, I release almost 30k tiny image for training, although 0.01% of the entire dataset, enough to play I think. Once I finished, I will upload the best model.
from cnn-lstm-ctc.
I already upload 188.pkl, the best model I can find in 200 epochs with 0.01% dataset. Sorry for that I could not fetch the whole dataset at this moment. However, the code is functional well. Hope it helps.
from cnn-lstm-ctc.
Related Issues (11)
- ImportError: No module named layers.net HOT 1
- no folder 'split_tiny_images'
- Getting IndexError
- would you to implement this with Keras ? HOT 3
- Where can I find data set? HOT 1
- Training process HOT 9
- How
- how to add a uni-LSTM layer after the biLSTM layer? HOT 1
- loss: nan, iter:1/455(1, 1.076s) HOT 4
- how to generate dataset? HOT 1
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 cnn-lstm-ctc.