How do you load the saved model ? about jointre HOT 1 CLOSED

The current uploaded code automatically evaluates the performance on the dev set, after each training epoch, without saving the learned parameters.

To save the model parameters:

• First, you add the following two code lines before code line 79:

  with open("PATH-TO-VOCAB-FILE", 'w') as paramsfp:
    pickle.dump((words, nertags, rels, w2i, c2i, options), paramsfp)

  Then you uncomment the code line 112 to save learned parameters:

  parser.Save("PATH-TO-LEARNED-MODEL-FILE")

• Then for loading a trained model, you can use:

  words, nertags, rels, w2i, c2i, options = pickle.load("PATH-TO-VOCAB-FILE")
trained_model = learner.jNeRE(words, nertags, rels, w2i, c2i, options)
trained_model.Load("PATH-TO-LEARNED-MODEL-FILE")
predTest, relsTest = trained_model.Predict(test_data)

from jointre.