Sparsh-Bansal

This may be because the pretrained model is trained on version 1.0 and 1.1

@Affanamin do this , ckpt = ckpt + '.data-00000-of-00001' before this proc = ["python", "src/summarization.py", "--test_file", data_pattern.format(dataset), "--batch_size", str(beam_search), "--test_output", output_file, "--geneos", str(tag), "--checkpoint", ckpt]

It is trained for Giga Word Dataset

its similar like testing on other articles.............like if you want to test it on your custom dataset then it will give good results....but that doesn't means that it is trained...