-
The model was trained on flicker8k, but the results achieved only half the BLEU-4 score mentioned by the authors (about 0.14-0.15). I have not modified any parameters in train.py. May I ask why such a…
-
Hi Bsantraigi
You have provided great tutorials. I saw your transliteration.ipynb.
In this, in Input cell "In[37]", you have put the line 'pbar.set_description(f"Epoch: {_e} >> Loss: {batch_loss/(bx…
-
Hi,
I've just found that the calculation of BLEU score during the training is incorrect.
Current implementation of evaluateBLEU function in train.cc,
just compare hypothesis and target word IDs,…
-
Hello, I notice that MaKE outputs the delexicalized words sequence.
Do you compute bleu score by delexicalized words sequence or text after post-processing ?
By the way, I retrain the train_res_…
-
I tested on the IWSLT'15 English-Vietnamese data to start learning how to use NMT.
Everything worked fine with a few minor tweaks*, but I am seeing BLEU scores above 1.0 both in training and testin…
-
Hi
I'm trying to reproduce your work about YELP dataset.
However, the BLEU score that i've gotten is quite different.
I trained the model 200 times and got the score from your multi-bleu.py file…
-
Dear Authors,
Thanks for the great work and an amazing repo!
I trained an unsupervised MT model, using a pretrained MLM model that I downloaded based on the README file. After training for man…
-
### Description
I copied the Walkthrough steps exactly.After Training and decoding steps,
when I run the following codes:
t2t-bleu --translation=translation.en --reference=ref-translation.de,
the …
-
Why do I only get a CIDEr score of 0.065 on Flickr30k for bliva_vicuna7b, even if multiplied by 10 it's only 0.65? Could you tell me what might have gone wrong in this process?? Thanks.
{"test": {"Bl…
-
Hi, I am using this fantastic tool to generate tests that are better understandable than the original EvoSuite. However, I encountered an issue when using it. I also created this issue on [UTGen/UTGen…