FaezeBr

Getting "in complete assert(isLoaded) AssertionError" when running complete.py on my own checkpoints directory

1

Hello Guys, I have trained the model and saved the .index, .meta and .data file in a different directory. However; when running complete.py I am getting this error: File "/home/hannahbrahman/ADV...

I wonder if there is any strategy to empty cache after loading (Restoring) gpt2 parameters into our model? I am getting OOM error for batch_size and I wonder how if...

Beam search decoding during inference doesn't generate good text.

4

Hi, I have trained a model using Reinforcement learning. When I use **"beam search"** to generate text, it generates all ``` "raeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraeraera" ``` However, when I use **greedy** or **topk...

Using reinforce_loss

19

Hi, I was trying to write a function for computing reinforce loss (as below) when I realized you have this [here](https://github.com/asyml/texar/blob/caa931a276ec215ef85bb1cfc1d486cdf877b1fd/texar/tf/losses/rl_losses.py#L27). In this regard, how I can use the [TransformerDecoder](https://github.com/asyml/texar/blob/caa931a276ec215ef85bb1cfc1d486cdf877b1fd/texar/tf/modules/decoders/transformer_decoders.py#L301)...

Hello I checked #36 and I wonder how I can retrain the LM on new dataset. Any guide would be appreciated.

Is it possible to do the auxilary task multiple times for each instance?

3

Hi, Is it possible to define multiple mc_token for each instance and do an auxilary classification for each of them? If yes, can you guide me? Thanks

Hi, I was able to run the first two steps of annotating a corpus using pretrained model, i.e. target id and frame id identification. But when I run the code...

Does the multi-label classifier optimize the entire model or it freezes BERT?

5

I wonder if the default code optimizes the entire model end-to-end or just the additional classifier layer parameters get updated? Thanks

TypeError: Input 'input_sizes' of 'Conv2DBackpropInput' Op has type float32 that does not match expected type of int32.

2

Hello, When running train.py I got this error which raise from deconv2d: During handling of the above exception, another exception occurred: Traceback (most recent call last): File "train.py", line 128,...

Longer sequences is being generated than expected.

10

Hi, I trained a model using your code on different data and I realized for all the models that I trained, the model is generating the output sequence that is...