Environment issues

[Harley-ZP]

I am using window 10 OS,
and I followed the markdown instruction, installed packages including dgl-cuda 10.1,
but showing up 'no module named dgl' while executing "import dgl",
it seems that the dgl-cuda could not be correctly imported.

Any one facing same problem?

Chinese version:

按.md装了包之后dgl cuda没法正常被import..
import dgl报错。

[MrLiuCC]

应该是环境没有配好

[MrLiuCC]

I use following codes to create environments successfully:
`
conda create -n renet python=3.6 numpy
conda activate renet
conda install pytorch==1.6.0 torchvision==0.7.0 cudatoolkit=10.1 -c python
conda install -c dglteam dgl-cuda10.1
conda install scikit-learn

`

[Harley-ZP]

Thanks a lot! I may try this solution later !

…

---Original--- From: ***@***.***> Date: Thu, Mar 31, 2022 14:22 PM To: ***@***.***>; Cc: ***@***.******@***.***>; Subject: Re: [INK-USC/RE-Net] Environment issues (Issue #60) I use following codes to create environments successfully: ` conda create -n renet python=3.6 numpy conda activate renet conda install pytorch==1.6.0 torchvision==0.7.0 cudatoolkit=10.1 -c python conda install -c dglteam dgl-cuda10.1 conda install scikit-learn ` — Reply to this email directly, view it on GitHub, or unsubscribe. You are receiving this because you authored the thread.Message ID: ***@***.***>

[ZhengJialin1000]

报错
E:\anaconda3\lib\site-packages\torch_tensor.py:575: UserWarning: floor_divide is deprecated, and will be removed in a future version of pytorch. It currently rounds toward 0 (like the
'trunc' function NOT 'floor'). This results in incorrect rounding for negative values.
To keep the current behavior, use torch.div(a, b, rounding_mode='trunc'), or for actual floor division, use torch.div(a, b, rounding_mode='floor'). (Triggered internally at ..\aten\sr
c\ATen\native\BinaryOps.cpp:467.)
return torch.floor_divide(self, other)
Traceback (most recent call last):
File "pretrain.py", line 139, in
train(args)
File "pretrain.py", line 83, in train
loss = model(batch_data, true_s, true_o, graph_dict)
File "E:\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "F:\code\RE-Net-master\global_model.py", line 47, in forward
packed_input = self.aggregator(sorted_t, self.ent_embeds, graph_dict, reverse=reverse)
File "E:\anaconda3\lib\site-packages\torch\nn\modules\module.py", line 1051, in _call_impl
return forward_call(*input, **kwargs)
File "F:\code\RE-Net-master\Aggregator.py", line 55, in forward
batched_graph.ndata['h'] = ent_embeds[batched_graph.ndata['id']].view(-1, ent_embeds.shape[1])
File "E:\anaconda3\lib\site-packages\dgl\view.py", line 84, in setitem
self._graph._set_n_repr(self._ntid, self._nodes, {key : val})
File "E:\anaconda3\lib\site-packages\dgl\heterograph.py", line 4124, in _set_n_repr
' same device.'.format(key, F.context(val), self.device))
dgl._ffi.base.DGLError: Cannot assign node feature "h" on device cuda:0 to a graph on device cpu. Call DGLGraph.to() to copy the graph to the same device.
在aggregator 55行前添加了batched_graph = dgl.batch(g_list)改为 local variable 'batch_data' referenced before assignment
请问应该怎么修改，应该是dgl版本问题