Hi,
I am new to pytorch. I have trained as you mentioned using flickr8k dataset. After training how to run eval.py? I am running the code in google colab. on running the eval.py i am getting the following error
Traceback (most recent call last):
File "/content/gdrive/My Drive/B.Tech Project/Attention/eval.py", line 179, in
print("\nBLEU-4 score @ beam size of %d is %.4f." % (beam_size, evaluate(beam_size)))
File "/content/gdrive/My Drive/B.Tech Project/Attention/eval.py", line 131, in evaluate
incomplete_inds = [ind for ind, next_word in enumerate(next_word_inds) if
File "/content/gdrive/My Drive/B.Tech Project/Attention/eval.py", line 132, in
next_word != word_map['']]
RuntimeError: CUDA error: device-side assert triggered
Hi, I am new to pytorch. I have trained as you mentioned using flickr8k dataset. After training how to run eval.py? I am running the code in google colab. on running the eval.py i am getting the following error
Traceback (most recent call last): File "/content/gdrive/My Drive/B.Tech Project/Attention/eval.py", line 179, in
print("\nBLEU-4 score @ beam size of %d is %.4f." % (beam_size, evaluate(beam_size)))
File "/content/gdrive/My Drive/B.Tech Project/Attention/eval.py", line 131, in evaluate
incomplete_inds = [ind for ind, next_word in enumerate(next_word_inds) if
File "/content/gdrive/My Drive/B.Tech Project/Attention/eval.py", line 132, in
next_word != word_map['']]
RuntimeError: CUDA error: device-side assert triggered