Comments (8)
Yes, these tutorials were very helpful. Thank you for writing them!
Using pre-trained embeddings was also an issue I came across. One of the main difficulties I've had with learning to write encoder-decoder models in TF has been the rapidly changing API. I don't think that your tutorials actually use the legacy seq2seq API at any point - this is a selling point that's probably worth noting!
from tensorflow-seq2seq-tutorials.
The best tutorial about RNN in TF that I have ever read.
It would be great if you add:
- How we can assign ID numbers to the each word from corpus(the best way with TF API)
- What if would like to download pre-trained embeddings
from tensorflow-seq2seq-tutorials.
These tutorials help a lot. Thanks so much!
But I just consider that how to coding after we change the parameter time_major. It seems that the code would change a lot if we set the parameter time_major as True which means the format of the inputs is [batch_size, time_major].
And how to use MMI in seq2seq model
from tensorflow-seq2seq-tutorials.
How can I implement attention in 'Advanced dynamic seq2seq with TensorFlow'?
from tensorflow-seq2seq-tutorials.
I also want to ask the question that @pushpankar asked, it seems in 2-seq2seq-advnaced ipython notebook there aren't any implementation for attention mechanism
from tensorflow-seq2seq-tutorials.
@pushpankar
@ryh95
Yep, attention implementation is missing as of now. I was going to implement it in tutorial 2.5, but switched to pytorch soon after finishing tutorial 3.
I would definitely like to return to it at some point in the future.
from tensorflow-seq2seq-tutorials.
Matvey:
Thank you for tutorial. I have couple of questions regarding the Tutorial #1.
- Is the seq2seq trained to output [W, X, Y, Z, ] or just [X,Y,Z, ] ?
- The vocab_size is set to 10 ? why? and why we have input_embedding_size = 20 which is greater than vocab_size? I would expect that vocab_size should be far greater than embedding_size?
- Should we include less.close() at the end to close interaction session?
from tensorflow-seq2seq-tutorials.
@ematvey Thank you for your tutorial. Best one I've found yet (and I've looked at lots of them). I am implementing a multi-dimensional regression seq2seq with Mixture Density Networks. If you could explain how to implement attention mechanism then it would be perfect.
Again, thanks for the tutorial. It is awesome as is!!
from tensorflow-seq2seq-tutorials.
Related Issues (20)
- new changes in tensorflow HOT 4
- My own embedding HOT 1
- Multilayered Bidirectional seq2seq HOT 6
- Potential bug of the dynamic_decoder HOT 2
- OOM error after running some batches HOT 3
- Beam search toy example
- Using the same weights in two different positions seems wrong
- module 'tensorflow.contrib.seq2seq' has no attribute 'prepare_attention' HOT 2
- there is no input argument when loop_fn method is called. HOT 1
- Increasing max sequence length and vocab size HOT 1
- How does inference work? Always getting same predictions HOT 2
- Duplicated computation of decoder outputs?
- toy task-example1-error
- testing
- Attempted an implementation
- good job
- Logits and labes have different shapes when computing cross-entropy loss HOT 1
- decoder input
- ask for help
- How to connect multilayered encoder to decoder?
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 tensorflow-seq2seq-tutorials.