-
I have a language model trained on a very large corpus.
I can input any input string in the t2t_decoder interactive mode and it completely ignores it, it's not used by the model to generate text. Any…
ghego updated
7 years ago
-
```
We have image support :
[t2t.png]
and image as a link support :
[[t2t.png] as-a-link.html]
We really NEED image from link support, something like :
[http://txt2tags.sourceforge.net/img/t2tpowered…
-
### Description
Hi, guys. I have an error when using the 'translate_enzh_wmt32k_rev' problem for decoding.
It was fine during training with the reverse problem name but interrupted when decoding.
.…
-
```
Now that we have separate Python files for each target, new target configs
could be added.
Today the file extension is bound to the target nickname (that used in the
--target option). For examp…
-
### Description
I am trying to use t2t-exporter to export a specific checkpoint. I set checkpoint_path, but when i ran t2t-exporter, it complains AttributeError.
```
t2t-exporter \
--data_di…
-
I tried to reproduce results from the paper on WMT En2De, base model. In my experiments I tried both BPE and word piece model. Here are the steps I made to train models:
```
# For BPE model I used…
-
Hi, I tried to generate question for Arabic and Urdu language and it seems small model cannot fit into memory to generate question.
It runs for a long time and then runtime crashes most of the time b…
-
i installed t2t with command below:
`pip install tensor2tensor && t2t-trainer --generate_data --data_dir=~/t2t_data --output_dir=~/t2t_train/mnist --problem=image_mnist --model=shake_shake …
-
I'm relatively new to t2t and was studying leveraging it for ASR when I came across your work.
Amazing work done @mohitsshah with proper explanation over at16k. The results are pretty impressive.
I'…
-
```
Oh no :(
The current version from SVN (r578) is WAY slower than the previous 2.6
release. It's almost 2 times slower.
See a quick test:
$ targets="gwiki html lout man mgp moin pm6 sgml tex txt…