alibaba / euler

A distributed graph deep learning framework.
Apache License 2.0
2.89k stars 559 forks source link

ScalableGCN运行命令 #206

Open MeliaLin opened 4 years ago

MeliaLin commented 4 years ago

采用的cora数据集,已经按成功按官网转成了二进制形式,运行命令: python -m tf_euler --data_dir cora --max_id 2707 --model scalable_gcn --mode train --num_classes 5

下面是运行错误:File "/usr/local/lib/python2.7/dist-packages/euler_gl-0.1.2-py2.7.egg/tf_euler/python/encoders.py", line 157, in call embedding = tf.concat(embeddings, -1) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/array_ops.py", line 1124, in concat return gen_array_ops.concat_v2(values=values, axis=axis, name=name) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/ops/gen_array_ops.py", line 1033, in concat_v2 "ConcatV2", values=values, axis=axis, name=name) File "/usr/local/lib/python2.7/dist-packages/tensorflow/python/framework/op_def_library.py", line 570, in _apply_op_helper (input_name, op_type_name, len(values), num_attr.minimum)) ValueError: List argument 'values' to 'ConcatV2' Op with length 0 shorter than minimum length 2.

请问:运行命令中还要加什么吗?

wenruij commented 4 years ago

跑 scalable_gcn这种有监督模型,这四个参数 好像没看到你加?

      label_idx=*,
      label_dim=*,
      feature_idx=*,
      feature_dim=*,
MeliaLin commented 4 years ago

--feature_idx,稠密特征的编号,使用稠密特征必须; --feature_dim,稠密特征维度,使用稠密特征必须; -—label_idx,label在稠密特征中的编号,监督模型必须; --label_dim,label在稠密特征中的维度,监督模型必须; graphsage的节点feature和label都放在了node_float_feature里面,如果放在node_unit64_feature里面需要在修改运行命令吗?

wenruij commented 4 years ago

建议修改一下数据格式,将lalel和feature放到float_feature中,目前label和 dense feature都是从 float_feature中获取的

MeliaLin commented 4 years ago

了解,非常感谢啦!

AI-Friend commented 4 years ago

借楼请教一下,node的label也放在了float_feature里?那么label_idx是float_feature里的一个属性id? label_dim表示label有多少类? 总结就是float_feature不仅要存 node的浮点型特征还要存,node的label?

wenruij commented 4 years ago

@Ljl-Jdsk 是的