AIPHES / emnlp19-moverscore

MoverScore: Text Generation Evaluating with Contextualized Embeddings and Earth Mover Distance
MIT License
192 stars 31 forks source link

CUDA problem #10

Open JessedeDoes opened 4 years ago

JessedeDoes commented 4 years ago

I get this error:

Traceback (most recent call last): File "test.py", line 4, in plot_example(True, reference, translation) File "/home/jesse/.local/lib/python3.7/site-packages/moverscore_v2.py", line 194, in plot_example distance_matrix = masks * distance_matrix
RuntimeError: expected backend CUDA and dtype Float but got backend CUDA and dtype Long

(The line in question is distance_matrix = masks * distance_matrix,

when i change it to distance_matrix = masks.double() * distance_matrix.double(),

the problem seems to disappear)