flink-issues mailing list archives

Site index · List index
Message view « Date » · « Thread »
Top « Date » · « Thread »
From GitBox <...@apache.org>
Subject [GitHub] [flink] walterddr commented on a change in pull request #8402: [FLINK-12473][ml] Add the interface of ML pipeline and ML lib
Date Fri, 17 May 2019 23:56:33 GMT
walterddr commented on a change in pull request #8402: [FLINK-12473][ml] Add the interface
of ML pipeline and ML lib
URL: https://github.com/apache/flink/pull/8402#discussion_r285317219

 File path: flink-ml/flink-ml-api/src/main/java/org/apache/flink/ml/api/core/PipelineStage.java
 @@ -0,0 +1,70 @@
+ * Licensed to the Apache Software Foundation (ASF) under one
+ * or more contributor license agreements.  See the NOTICE file
+ * distributed with this work for additional information
+ * regarding copyright ownership.  The ASF licenses this file
+ * to you under the Apache License, Version 2.0 (the
+ * "License"); you may not use this file except in compliance
+ * with the License.  You may obtain a copy of the License at
+ *
+ *     http://www.apache.org/licenses/LICENSE-2.0
+ *
+ * Unless required by applicable law or agreed to in writing, software
+ * distributed under the License is distributed on an "AS IS" BASIS,
+ * WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
+ * See the License for the specific language governing permissions and
+ * limitations under the License.
+ */
+package org.apache.flink.ml.api.core;
+import org.apache.flink.ml.api.misc.param.ParamInfo;
+import org.apache.flink.ml.api.misc.param.WithParams;
+import org.apache.flink.ml.api.misc.persist.Persistable;
+import org.apache.flink.ml.util.param.ExtractParamInfosUtil;
+import org.apache.flink.ml.util.persist.MLStageFactory;
+import com.google.gson.Gson;
+import com.google.gson.JsonObject;
+import java.io.Serializable;
+import java.util.HashMap;
+import java.util.List;
+import java.util.Map;
+ * Base class for a stage in a pipeline. The interface is only a concept, and does not have
+ * actual functionality. Its subclasses must be either Estimator or Transformer. No other
+ * should inherit this interface directly.
+ *
+ * <p>Each pipeline stage is with parameters and meanwhile persistable.
+ *
+ * @param <T> The class type of the PipelineStage implementation itself, used by {@link
+ *            org.apache.flink.ml.api.misc.param.WithParams}
+ * @see WithParams
+ */
+interface PipelineStage<T extends PipelineStage<T>> extends WithParams<T>,
+		Persistable {
+	default String toJson() {
 Review comment:
   Yes, this is also my confusion I think. let me rephrase it differently:
   1. if this is only for human readability, why do we need a `toJson` and `fromJson`? shouldn't
the `toJson` be enough? 
   2. I can see that parameter editing is a very tunable part for ML users. I am not expert
but do we usually package param literals (e.g. the values) and models together? can we load
them differently? e.g. load the param json and the model json separately. 
   3. with my personal preference, if it can be a choice for user (json or yaml, or other)
that would be even better.
   I think one large question I have is whether we use this for serving purpose. if we are
not using this for serving purpose. I am actually fine with Json or Yaml. 

This is an automated message from the Apache Git Service.
To respond to the message, please log on to GitHub and use the
URL above to go to the specific comment.
For queries about this service, please contact Infrastructure at:

With regards,
Apache Git Services

View raw message