Steven Huang
Steven Huang
#### Describe the bug As titled, the minimal example for codeAct agent provided in https://github.com/OpenDevin/OpenDevin/tree/main/agenthub/codeact_agent could not be reproduced. Agent constantly output `I don't understand your input`. Is this feature...
In this PR, we 1. Support the saving and loading of raw text in GLI datasets. In metadata, raw text will occupy an entry `RawText` in parallel with `Node`, `Edge`...
**Is your feature request related to a problem? Please describe.** Task 1: Add raw text loading support for GLI, which should cover cora, pubmed, cite seer, ogbn-product and ogbn-arxiv. Task...
As titled. Provide `mse` and `mae` as metrics and loss function. `train.py` can run successfully on `cora` (pretending it is a node regression task). TODO: update `train_minibatch.py` and `train_gbdt.py`
**Is your feature request related to a problem? Please describe.** Using lightning can simplify our implementation of models, training, etc. The idea is from @xingjian-zhang. I am still going through...
**Is your feature request related to a problem? Please describe.** In our implementation, some APIs are not clear. For example: 1. We need access node feature by `g.ndata["NodeFeature"]`, which is...
**Is your feature request related to a problem? Please describe.** When we load a set of graphs, we have no information on the difference between datasets. For example,  where...
**Is your feature request related to a problem? Please describe.** Now we have three evaluation function for both node classification task and graph classification task. In the future we may...
In graph classification, there are two ways of adding self loop to all graphs (it is common to add self loop when using GCN, and on some datasets). The first...