In Text-Style-Transfer Example, how do you use the model on a chosen sentence?

[ghost]

Hi,
I tried running the code for the text style transfer example after reading the related paper(Unsupervised Text Style Transfer using Language Models as Discriminators) and I have a doubt regarding its use.
I am running the code main.py and it is training on the sample yelp datasets given. Training the model on my custom dataset seems simple enough, I just change the path in the config.py file, but after training, if I want to manually use the model to perform a text style transfer on certain sentences I've chosen and read the outputs, how would I go about doing that? Is there a function I can run with a string as an input that gives the transformed string as the output?
I apologize if this is too trivial, I'm very much beginner in this subject.
Thanks in advance for your help

[BJavid]

Hi,
Excuse me
I want to run this example but got an error in main.py. This is terminal error. Can you help me?

` File "main.py", line 179, in main
eval_epoch(sess, gamma, lambda_g, epoch, 'val')

File "main.py", line 133, in _eval_epoch
avg_meters.add(vals, weight=batch_size)

File "C:\Users\Bahar\Anaconda3\envs\envm\lib\site-packages\texar\tf\utils\average_recorder.py", line 222, in add
self._recorders[name].add(val, weight=weight)

File "C:\Users\Bahar\Anaconda3\envs\envm\lib\site-packages\texar\tf\utils\average_recorder.py", line 65, in add
self._sum += record * w
TypeError: unsupported operand type(s) for +=: 'float' and 'list'
`

[gpengzhi]

Hi @ZhitingHu , could you help answer this question?

[cece95]

Hi, i'm trying to use the model on new examples, was there any development on this issue?

Thanks in advance
Cece

[ZhitingHu]

Hi,
I tried running the code for the text style transfer example after reading the related paper(Unsupervised Text Style Transfer using Language Models as Discriminators) and I have a doubt regarding its use.
I am running the code main.py and it is training on the sample yelp datasets given. Training the model on my custom dataset seems simple enough, I just change the path in the config.py file, but after training, if I want to manually use the model to perform a text style transfer on certain sentences I've chosen and read the outputs, how would I go about doing that? Is there a function I can run with a string as an input that gives the transformed string as the output?
I apologize if this is too trivial, I'm very much beginner in this subject.
Thanks in advance for your help

The current code does not provide such interactive use. It only processes inputs from files (e.g., ./data/yelp/sentiment.test.text). You may want to adapt the code to add the function, or to use the current code you may write the sentences in a file in the same format as ./data/yelp/sentiment.test.text.

[ZhitingHu]

Hi,
Excuse me
I want to run this example but got an error in main.py. This is terminal error. Can you help me?

` File "main.py", line 179, in main eval_epoch(sess, gamma, lambda_g, epoch, 'val')

File "main.py", line 133, in _eval_epoch
avg_meters.add(vals, weight=batch_size)

File "C:\Users\Bahar\Anaconda3\envs\envm\lib\site-packages\texar\tf\utils\average_recorder.py", line 222, in add
self._recorders[name].add(val, weight=weight)

File "C:\Users\Bahar\Anaconda3\envs\envm\lib\site-packages\texar\tf\utils\average_recorder.py", line 65, in add
self._sum += record * w
TypeError: unsupported operand type(s) for +=: 'float' and 'list'
`

The code works well on my side. Did you change any configs?

Can you print vals before L.333 and see if any of the values in the dict is list?