search

anhaidgroup / deepmatcher

Python package for performing Entity and Text Matching using Deep Learning.
BSD 3-Clause "New" or "Revised" License
553 stars 129 forks source link

model.run_prediction Failed with Error : input.size(-1) must be equal to input_size. Expected 100, got 50 #83

Open RahulDas-dev opened 3 years ago

RahulDas-dev commented 3 years ago

I have trained two model one with embeddings glove.6B.100d and other is glove.6B.50d . The training and prediction for 100d was successful. Then I tried with 50d embeddings , there training was successful but run prediction is failed. With bellow Error.

RuntimeError: input.size(-1) must be equal to input_size. Expected 100, got 50

I use dm.data.reset_vector_cache() before the training with glove.6B.50d , bust still the error persists . No Idea how to fix this Kindly help .

my code for training the model & prediction is here with complete Error details.

while training only embeddings variable is changed to glove.6B.50d from glove.6B.100d

sidharthms commented 3 years ago

I'm not able to reproduce this issue with the described setup. Is it possible for you to create a version of the notebook that I can run on Colab, perhaps with the sample data that's provided?