Beronx86

Hi, I want to draw the rectangles. It seems that I needs the return values of ZoomableAttentiuonWindow.nn2att (center_y, center_x, delta, sigma, gamma). So what's the best way to monitor these...

support recurrent with no states.

5

The recurrent wrapper does not support loop with no states. But this kind of loop may be useful. So I modified the codes. Fixes #1112

Bug in nested recurrent model

2

Multiple `top_bricks` are found while building the nested recurrent model. But there is only one in fact. So errors occurs. Besides, the `theano.function` of the nested recurrent method works well....

I am writing a `recurrent` method. It only takes `sequences` and `contexts` as inputs. And there is no recurrent states. But it seems that the `recurrent` wrapper still nees the...

LSTM.apply interface. recurrent.states and recurrent.outputs

1

LSTM.apply use 'states', 'cells' as recurrent states and return them as output. I think returning 'cells' is unnecessary. It makes the LSTM interface different from that of GRU. And LSTM...

Convenient way to monitor many variables with the same names

5

My program raise NaN error after several iterations. I want to monitor the automatically added auxiliary variables (`W_norm` and `b_norm`). I use `TrainingDataMonitoring(fg.auxiliary_variables, after_batch=True)`. But the program raise `ValueError: variables...

draw align map in machine translation

2

Add some code to draw align heatmap when sampling.

When collecting vocabulary, vocab = {'UNK': 1, '': 0, '': 0} bos, eos tag are all set to `0`, then vocab would collect `vocab_size - 2` normal words. But in...

not understand "log probability in the center of the bin"

2

Could anybody explain the default values to be used in extreme cases. ``` 68 # log probability in the center of the bin, to be used in extreme cases 69...

I want to add a valid set in my code. So I created 2 AudioReader in train.py. But the code locks up. if there is only 1 AudioReader and`n_threads` argument...