-
Tensorflow tutorial Seq2Seq is using the [WMT 2015 datasets](http://www.statmt.org/wmt15/translation-task.html) Following the step by step guide in the OpenNMT readme, will it work switching to those …
-
Hi!
I'm running the last version of Nematus in a machine with Ubuntu16.04, cuda-8.0 and Theano 0.9, but when I run test_train.sh the cost of the iteration is 0.0.
The same thing happens when runn…
-
Hi all,
I've run a training on a private small dataset (~200K parallel sentences) trying to use the same hyperparameters for both OpenNMT-py and Sockeye, as far as it was possible.
After 50 epoc…
-
I get `403 Forbidden` when accessing `http://demo.clab.cs.cmu.edu/cdyer/iwslt-deen-prepared.tar.gz`.
-
The cond. GRU outperforms LSTM and LSTM outperforms GRU (as several people mentioned on IWSLT).
-
Hi,
I've tried to run a training on iwslt data en-fr. The first train epoch finished with loss: nan, but this may be due to my choice of parameters. The problem is, when it started the validation I…
-
Hi,
I am trying to write a callback that can roll back to the best model when the evaluation metric keeps going down. However I observed that it cannot successfully modify the params.
For example, t…
-
I couldn't figure out exactly who to send this to as a lot of people interact with the TEDLIUM data set, so I'm just putting it as an issue on github.
I'm working on some scripts to resegment data, a…
-
Hullo -- is it possible to have KenLM support added? It can produce sentence-level perplexity scores (previous versions did not), and is much more efficient.
-
1. I'm training the session3/nmt.py with attention on the europarl corpus with 5000-term vocabulary, 250 word vector dimension, 500 internal representation dimension. It takes a couple of days for a f…
jli05 updated
8 years ago