-
Hey everyone. First, thanks for the awesome work with this package!
As I am working more and more with it, I started wondering if:
1 - would it be possible to extend the main STAN model to incor…
-
Hello, I am very confused of the part of context model in your code, what's the output of the context model represents? And what's the meaning of pc_loss = beta * tf.maximum(H_soft - H_target, 0), th…
-
There were additional topics discussed that could not be turned into coherent notes after the fact.
## Action items
* TensorFlow team to create a block diagram and then go deeper. Goal is to h…
-
I can now successfully run a submission as long as I check the "Do not apply Bayesian smoothing" option. But if I select one of the other values (at least the first of them, "car_r_procedure: Conditio…
-
I trained a model with millions of sentence pairs, and tested the model with thousands of testing data. Then I happened to continue the training with the testing data (sort of over-fitting the model f…
lkluo updated
6 years ago
-
As noted in https://github.com/gcp/leela-zero/issues/1417#issuecomment-392674148, leelaz has played some unexpected moves against Haylee from the last few games that resulted in her being curious/inte…
-
I really hate to be *that guy*, but I'm still having some issues with sampling on more challenging models. I've seen CPU usage shoot up, and the sampler also crashes sometimes, without good reason. Th…
-
i.e. `log(x(t)) = a log(x(t-1)) + b + omega`
@lkapelevich you did this with JADE inflows yeah? @Thuener is looking at doing it for price.
It would be nice if this was implemented nicely.
-
in the train.py, you get predicted wav by "ypred = sess.run(yhat, feed_dict = {x:labels, y:wavs})". I can't understand tts's input is text and output is a wav, your project do a tts mission?
-
I want to implement a text classification problem using tensor2tensor. My hparams is
`p.target_modality = (registry.Modalities.CLASS_LABEL, target_vocab_size)`.
However, the `targets_bottom` metho…