/*
* 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 hivemall;
import static org.apache.hadoop.hive.serde2.objectinspector.primitive.PrimitiveObjectInspectorFactory.writableFloatObjectInspector;
import hivemall.mix.MixMessage.MixEventName;
import hivemall.mix.client.MixClient;
import hivemall.model.DenseModel;
import hivemall.model.PredictionModel;
import hivemall.model.SpaceEfficientDenseModel;
import hivemall.model.SparseModel;
import hivemall.model.SynchronizedModelWrapper;
import hivemall.model.WeightValue;
import hivemall.model.WeightValue.WeightValueWithCovar;
import hivemall.utils.datetime.StopWatch;
import hivemall.utils.hadoop.HadoopUtils;
import hivemall.utils.hadoop.HiveUtils;
import hivemall.utils.io.IOUtils;
import hivemall.utils.lang.Primitives;
import java.io.BufferedReader;
import java.io.File;
import java.io.IOException;
import java.util.List;
import javax.annotation.Nonnull;
import javax.annotation.Nullable;
import org.apache.commons.cli.CommandLine;
import org.apache.commons.cli.Options;
import org.apache.commons.logging.Log;
import org.apache.commons.logging.LogFactory;
import org.apache.hadoop.hive.ql.exec.UDFArgumentException;
import org.apache.hadoop.hive.ql.metadata.HiveException;
import org.apache.hadoop.hive.serde2.SerDeException;
import org.apache.hadoop.hive.serde2.lazy.LazySimpleSerDe;
import org.apache.hadoop.hive.serde2.objectinspector.ObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.PrimitiveObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.StructField;
import org.apache.hadoop.hive.serde2.objectinspector.StructObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.FloatObjectInspector;
import org.apache.hadoop.hive.serde2.objectinspector.primitive.WritableFloatObjectInspector;
import org.apache.hadoop.io.Text;
public abstract class LearnerBaseUDTF extends UDTFWithOptions {
private static final Log logger = LogFactory.getLog(LearnerBaseUDTF.class);
protected String preloadedModelFile;
protected boolean dense_model;
protected int model_dims;
protected boolean disable_halffloat;
protected boolean is_mini_batch;
protected int mini_batch_size;
protected String mixConnectInfo;
protected String mixSessionName;
protected int mixThreshold;
protected boolean mixCancel;
protected boolean ssl;
protected MixClient mixClient;
public LearnerBaseUDTF() {}
protected boolean useCovariance() {
return false;
}
@Override
protected Options getOptions() {
Options opts = new Options();
opts.addOption("loadmodel", true, "Model file name in the distributed cache");
opts.addOption("dense", "densemodel", false, "Use dense model or not");
opts.addOption("dims", "feature_dimensions", true,
"The dimension of model [default: 16777216 (2^24)]");
opts.addOption("disable_halffloat", false,
"Toggle this option to disable the use of SpaceEfficientDenseModel");
opts.addOption("mini_batch", "mini_batch_size", true,
"Mini batch size [default: 1]. Expecting the value in range [1,100] or so.");
opts.addOption("mix", "mix_servers", true, "Comma separated list of MIX servers");
opts.addOption("mix_session", "mix_session_name", true,
"Mix session name [default: ${mapred.job.id}]");
opts.addOption("mix_threshold", true,
"Threshold to mix local updates in range (0,127] [default: 3]");
opts.addOption("mix_cancel", "enable_mix_canceling", false, "Enable mix cancel requests");
opts.addOption("ssl", false, "Use SSL for the communication with mix servers");
return opts;
}
@Nullable
@Override
protected CommandLine processOptions(@Nonnull ObjectInspector[] argOIs)
throws UDFArgumentException {
String modelfile = null;
boolean denseModel = false;
int modelDims = -1;
boolean disableHalfFloat = false;
int miniBatchSize = 1;
String mixConnectInfo = null;
String mixSessionName = null;
int mixThreshold = -1;
boolean mixCancel = false;
boolean ssl = false;
CommandLine cl = null;
if (argOIs.length >= 3) {
String rawArgs = HiveUtils.getConstString(argOIs[2]);
cl = parseOptions(rawArgs);
modelfile = cl.getOptionValue("loadmodel");
denseModel = cl.hasOption("dense");
if (denseModel) {
modelDims = Primitives.parseInt(cl.getOptionValue("dims"), 16777216);
}
disableHalfFloat = cl.hasOption("disable_halffloat");
miniBatchSize = Primitives.parseInt(cl.getOptionValue("mini_batch_size"), miniBatchSize);
if (miniBatchSize <= 0) {
throw new UDFArgumentException("mini_batch_size must be greater than 0: "
+ miniBatchSize);
}
mixConnectInfo = cl.getOptionValue("mix");
mixSessionName = cl.getOptionValue("mix_session");
mixThreshold = Primitives.parseInt(cl.getOptionValue("mix_threshold"), 3);
if (mixThreshold > Byte.MAX_VALUE) {
throw new UDFArgumentException("mix_threshold must be in range (0,127]: "
+ mixThreshold);
}
mixCancel = cl.hasOption("mix_cancel");
ssl = cl.hasOption("ssl");
}
this.preloadedModelFile = modelfile;
this.dense_model = denseModel;
this.model_dims = modelDims;
this.disable_halffloat = disableHalfFloat;
this.is_mini_batch = miniBatchSize > 1;
this.mini_batch_size = miniBatchSize;
this.mixConnectInfo = mixConnectInfo;
this.mixSessionName = mixSessionName;
this.mixThreshold = mixThreshold;
this.mixCancel = mixCancel;
this.ssl = ssl;
return cl;
}
protected PredictionModel createModel() {
return createModel(null);
}
protected PredictionModel createModel(String label) {
PredictionModel model;
final boolean useCovar = useCovariance();
if (dense_model) {
if (disable_halffloat == false && model_dims > 16777216) {
logger.info("Build a space efficient dense model with " + model_dims
+ " initial dimensions" + (useCovar ? " w/ covariances" : ""));
model = new SpaceEfficientDenseModel(model_dims, useCovar);
} else {
logger.info("Build a dense model with initial with " + model_dims
+ " initial dimensions" + (useCovar ? " w/ covariances" : ""));
model = new DenseModel(model_dims, useCovar);
}
} else {
int initModelSize = getInitialModelSize();
logger.info("Build a sparse model with initial with " + initModelSize
+ " initial dimensions");
model = new SparseModel(initModelSize, useCovar);
}
if (mixConnectInfo != null) {
model.configureClock();
model = new SynchronizedModelWrapper(model);
MixClient client = configureMixClient(mixConnectInfo, label, model);
model.configureMix(client, mixCancel);
this.mixClient = client;
}
assert (model != null);
return model;
}
protected MixClient configureMixClient(String connectURIs, String label, PredictionModel model) {
assert (connectURIs != null);
assert (model != null);
String jobId = (mixSessionName == null) ? MixClient.DUMMY_JOB_ID : mixSessionName;
if (label != null) {
jobId = jobId + '-' + label;
}
MixEventName event = useCovariance() ? MixEventName.argminKLD : MixEventName.average;
MixClient client = new MixClient(event, jobId, connectURIs, ssl, mixThreshold, model);
logger.info("Successfully configured mix client: " + connectURIs);
return client;
}
protected int getInitialModelSize() {
return 16384;
}
protected void loadPredictionModel(PredictionModel model, String filename,
PrimitiveObjectInspector keyOI) {
final StopWatch elapsed = new StopWatch();
final long lines;
try {
if (useCovariance()) {
lines = loadPredictionModel(model, new File(filename), keyOI,
writableFloatObjectInspector, writableFloatObjectInspector);
} else {
lines = loadPredictionModel(model, new File(filename), keyOI,
writableFloatObjectInspector);
}
} catch (IOException e) {
throw new RuntimeException("Failed to load a model: " + filename, e);
} catch (SerDeException e) {
throw new RuntimeException("Failed to load a model: " + filename, e);
}
if (model.size() > 0) {
logger.info("Loaded " + model.size() + " features from distributed cache '" + filename
+ "' (" + lines + " lines) in " + elapsed);
}
}
private static long loadPredictionModel(PredictionModel model, File file,
PrimitiveObjectInspector keyOI, WritableFloatObjectInspector valueOI)
throws IOException, SerDeException {
long count = 0L;
if (!file.exists()) {
return count;
}
if (!file.getName().endsWith(".crc")) {
if (file.isDirectory()) {
for (File f : file.listFiles()) {
count += loadPredictionModel(model, f, keyOI, valueOI);
}
} else {
LazySimpleSerDe serde = HiveUtils.getKeyValueLineSerde(keyOI, valueOI);
StructObjectInspector lineOI = (StructObjectInspector) serde.getObjectInspector();
StructField keyRef = lineOI.getStructFieldRef("key");
StructField valueRef = lineOI.getStructFieldRef("value");
PrimitiveObjectInspector keyRefOI = (PrimitiveObjectInspector) keyRef.getFieldObjectInspector();
FloatObjectInspector varRefOI = (FloatObjectInspector) valueRef.getFieldObjectInspector();
BufferedReader reader = null;
try {
reader = HadoopUtils.getBufferedReader(file);
String line;
while ((line = reader.readLine()) != null) {
count++;
Text lineText = new Text(line);
Object lineObj = serde.deserialize(lineText);
List<Object> fields = lineOI.getStructFieldsDataAsList(lineObj);
Object f0 = fields.get(0);
Object f1 = fields.get(1);
if (f0 == null || f1 == null) {
continue; // avoid the case that key or value is null
}
Object k = keyRefOI.getPrimitiveWritableObject(keyRefOI.copyObject(f0));
float v = varRefOI.get(f1);
model.set(k, new WeightValue(v, false));
}
} finally {
IOUtils.closeQuietly(reader);
}
}
}
return count;
}
private static long loadPredictionModel(PredictionModel model, File file,
PrimitiveObjectInspector featureOI, WritableFloatObjectInspector weightOI,
WritableFloatObjectInspector covarOI) throws IOException, SerDeException {
long count = 0L;
if (!file.exists()) {
return count;
}
if (!file.getName().endsWith(".crc")) {
if (file.isDirectory()) {
for (File f : file.listFiles()) {
count += loadPredictionModel(model, f, featureOI, weightOI, covarOI);
}
} else {
LazySimpleSerDe serde = HiveUtils.getLineSerde(featureOI, weightOI, covarOI);
StructObjectInspector lineOI = (StructObjectInspector) serde.getObjectInspector();
StructField c1ref = lineOI.getStructFieldRef("c1");
StructField c2ref = lineOI.getStructFieldRef("c2");
StructField c3ref = lineOI.getStructFieldRef("c3");
PrimitiveObjectInspector c1oi = (PrimitiveObjectInspector) c1ref.getFieldObjectInspector();
FloatObjectInspector c2oi = (FloatObjectInspector) c2ref.getFieldObjectInspector();
FloatObjectInspector c3oi = (FloatObjectInspector) c3ref.getFieldObjectInspector();
BufferedReader reader = null;
try {
reader = HadoopUtils.getBufferedReader(file);
String line;
while ((line = reader.readLine()) != null) {
count++;
Text lineText = new Text(line);
Object lineObj = serde.deserialize(lineText);
List<Object> fields = lineOI.getStructFieldsDataAsList(lineObj);
Object f0 = fields.get(0);
Object f1 = fields.get(1);
Object f2 = fields.get(2);
if (f0 == null || f1 == null) {
continue; // avoid unexpected case
}
Object k = c1oi.getPrimitiveWritableObject(c1oi.copyObject(f0));
float v = c2oi.get(f1);
float cov = (f2 == null) ? WeightValueWithCovar.DEFAULT_COVAR
: c3oi.get(f2);
model.set(k, new WeightValueWithCovar(v, cov, false));
}
} finally {
IOUtils.closeQuietly(reader);
}
}
}
return count;
}
@Override
public void close() throws HiveException {
if (mixClient != null) {
IOUtils.closeQuietly(mixClient);
this.mixClient = null;
}
}
}