rebel

rebel copied to clipboard

Error when training on DocRED: expected str, bytes or os.PathLike object, not DataFilesList

1

When evaluating on the DocRED dataset, I'm receiving the following error with traceback. I tried replacing the model variable with the path string but made no difference. I would be...

Iqra840

Seems that when using a spacy pipeline, a new line at the end of the input text causes the model to return a head and tail that are not in...

sean-simbolo

Hi, great paper! In your paper you state that you train on 220 relations but relation_counts.tsv has 230 relations. Is that a typo in the paper or is the relations...

bablf

hello, I would like to ask how many entity types and relationship types are there in the SREDFM Chinese data set, and which ones are included? Where can I check...

GitHubwwZ

Hi @LittlePea13, I have a question about the details of the experimental evaluation on DocRED. For Docred, the dataset includes 96 relation classes and class of 'no_relation'. So, I want...

gxxxzhang

Hi there! I tried to extract financial information from a bunch of financial articles summaries, but I didn't get expected results (such as funding info, investment etc). Does your model...

vmehera123

I am not able to understand from where we are getting the prediction and gold file. The batch is used many times in the code how is this batch getting...

mayanku

Problem with negative samples

3

Hi @LittlePea13 , I was just experimenting with REBEL, trying to see how it responds. Following are a few negative samples and relations extracted: ```python # Negative Samples negative_samples =...

sky-2002

Fine tuning for person to person entity relationship extraction

1

Please how cam I fine tune this model on my dataset to predict triplets of only person to person entities and their relationships. I am using CustFRE dataset [1-s2.0-S2352340922001913-mmc3.xlsx](https://github.com/Babelscape/rebel/files/12140176/1-s2.0-S2352340922001913-mmc3.xlsx) [custFRE...

Theosagie

Can't find factory for 'rebel' for language English (en).

1

I'm trying to run the Spacy example but I'm getting this error: `ValueError: [E002] Can't find factory for 'rebel' for language English (en). This usually happens when spaCy calls `nlp.create_pipe`...

flamingrickpat