-
## description
The aim of this issue is to list all the ideas about the structure of the thesis and its content.
## how to compile
The [thesis](https://github.com/dmsgnn/gnn-acceleration-mas…
-
/usr/local/bin/python3.10 /Users/mac/PycharmProjects/gnnFakeNews/gnn_model/gnn.py
dataset>> politifact(314)
args>> Namespace(seed=777, device='cpu', dataset='politifact', batch_size=128, lr=0.01,…
-
Trying to use py-nufft in my ada-lanczos back-propagation step. Please, could you assist?
Implementations of ![arXiv:1808.04580](https://arxiv.org/abs/1808.04580) and ![arXiv:1901.01484](https://arx…
-
I want to print the trained embeddings of GNN nodes in the node_set after training is complte. I cannot print it using intermidiate layers as the intermadiate layers are graph_update layers and only h…
-
-
## description
Since Torch-MLIR most tested path down to Torch MLIR Dialect is the TorchScript one, the aim of this issue is to change the export and load model from `state_dict` to TorchScript.
-
Using the cofiguration resolution -- 0.25°, mesh_size -- 6, latent_size -- 512, and gnn_msg_steps --16, an error message is displayed when dataset_source-era5_date-2022-01-01_res-0.25_levels-13_step…
-
ubuntu 20.04, rocm 5.x
```
git remote -v
origin https://github.com/pytorch/benchmark.git (fetch)
origin https://github.com/pytorch/benchmark.git (push)
git branch
* main
python3 install.…
-
## 背景
众所周知,Paddle 是一个历史悠久的框架,使得 Paddle 能够久经考验,应对各种场景稳定运行。但历史的沉淀同样带来一个严重的问题,就是框架内 API 语义不清晰,多种 API 能够做同样或者类似的事情。得益于我们的公开 API 审查机制和 fluid 清理,公开 API 中类似问题较少,但框架内部仍存在大量历史遗留的内部 API 的使用,这些 API 的存在导致框架内部需…
-
When running the PA dataset and preparing to test the model's correctness, I encountered some issues. I stored the model at the end of the entry function in the run_ducati.py file. Subsequently, I tes…