[Roadmap] Graph4NLP v0.5.5 plan

graph4ai / graph4nlp

Graph4nlp is the library for the easy use of Graph Neural Networks for NLP. Welcome to visit our DLG4NLP website (https://dlg4nlp.github.io/index.html) for various learning resources!

Apache License 2.0

1.67k stars 201 forks source link

[Roadmap] Graph4NLP v0.5.5 plan #429

Closed AlanSwift closed 2 years ago

AlanSwift commented 3 years ago

Task List

[New Feature] Develop the inference wrapper: InferenceWrapper. Users can easily generate results by the wrapper with raw inputs. For example:
```
model = load_model(...)
wrapper = InferenceWrapper(model, ....)
wrapper.predict(["Hello world"])
```
[Feature] Develop StandfordNLP and tokenize template parameters.
[Feature] Separate Graph topology and graph embeddings.

sandeepKondaveeti commented 3 years ago

Please confirm the new features on the roadmap would support all the applications of Graph4NLP and also ETA for the completion of the above mentioned features.

For Example, in case of KGC can we use the same InferenceWrapper for Seperate Graph topology and Graph Embeddings.

AlanSwift commented 3 years ago

Please confirm the new features on the roadmap would support all the applications of Graph4NLP and also ETA for the completion of the above mentioned features.

For Example, in case of KGC can we use the same InferenceWrapper for Seperate Graph topology and Graph Embeddings.

Yes, finally we will support it.

nashid commented 2 years ago

@AlanSwift when these features will be delivered? Any ETA?

AlanSwift commented 2 years ago

@AlanSwift when these features will be delivered? Any ETA?

Yes, please check 0.5.5.

AlanSwift commented 2 years ago

This issue is closed.