public class DeepLearning extends Job.ValidatedJob
| Modifier and Type | Class and Description |
|---|---|
static class |
DeepLearning.Activation
Activation functions
|
static class |
DeepLearning.ClassSamplingMethod |
static class |
DeepLearning.InitialWeightDistribution |
static class |
DeepLearning.Loss
Loss functions
CrossEntropy is recommended
|
Job.ValidatedJob.Response2CMAdaptorJob.ChunkProgress, Job.ChunkProgressJob, Job.ColumnsJob, Job.ColumnsResJob, Job.Fail, Job.FrameJob, Job.HexJob, Job.JobCancelledException, Job.JobState, Job.List, Job.ModelJob, Job.ModelJobWithoutClassificationField, Job.Progress, Job.ProgressMonitor, Job.ValidatedJobRequest2.ColumnSelect, Request2.Dependent, Request2.DoClassBoolean, Request2.DRFCopyDataBoolean, Request2.MultiVecSelect, Request2.MultiVecSelectType, Request2.TypeaheadKey, Request2.VecClassSelect, Request2.VecSelectRequest.API, Request.Default, Request.Filter, Request.Validator<V>RequestBuilders.ArrayBuilder, RequestBuilders.ArrayHeaderRowBuilder, RequestBuilders.ArrayRowBuilder, RequestBuilders.ArrayRowElementBuilder, RequestBuilders.ArrayRowSingleColBuilder, RequestBuilders.BooleanStringBuilder, RequestBuilders.Builder, RequestBuilders.ElementBuilder, RequestBuilders.HideBuilder, RequestBuilders.KeyCellBuilder, RequestBuilders.KeyElementBuilder, RequestBuilders.KeyLinkElementBuilder, RequestBuilders.KeyMinAvgMaxBuilder, RequestBuilders.NoCaptionObjectBuilder, RequestBuilders.ObjectBuilder, RequestBuilders.PaginatedTable, RequestBuilders.PreFormattedBuilder, RequestBuilders.Response, RequestBuilders.ResponseInfo, RequestBuilders.WarningCellBuilderRequestArguments.Argument<T>, RequestArguments.Bool, RequestArguments.CaseModeSelect, RequestArguments.CaseSelect, RequestArguments.ClassifyBool, RequestArguments.DRFCopyDataBool, RequestArguments.EnumArgument<T extends java.lang.Enum<T>>, RequestArguments.ExistingFile, RequestArguments.FrameClassVec, RequestArguments.FrameKeyMultiVec, RequestArguments.FrameKeyVec, RequestArguments.GeneralFile, RequestArguments.H2OCategoryStrata, RequestArguments.H2OCategoryWeights, RequestArguments.H2OExistingKey, RequestArguments.H2OGLMModelKey, RequestArguments.H2OHexKey, RequestArguments.H2OHexKeyCol, RequestArguments.H2OIllegalArgumentException, RequestArguments.H2OKey, RequestArguments.H2OKey2, RequestArguments.H2OKMeansModelKey, RequestArguments.H2OModelKey<TM extends OldModel,TK extends TypeaheadKeysRequest>, RequestArguments.HexAllColumnSelect, RequestArguments.HexColumnSelect, RequestArguments.HexKeyClassCol, RequestArguments.HexNonClassColumnSelect, RequestArguments.HexNonConstantColumnSelect, RequestArguments.HexPCAColumnSelect, RequestArguments.InputCheckBox, RequestArguments.InputSelect<T>, RequestArguments.InputText<T>, RequestArguments.Int, RequestArguments.LongInt, RequestArguments.MultipleSelect<T>, RequestArguments.MultipleText<T>, RequestArguments.NTree, RequestArguments.NumberSequence, RequestArguments.NumberSequenceFloat, RequestArguments.Real, RequestArguments.Record<T>, RequestArguments.RFModelKey, RequestArguments.RSeq, RequestArguments.RSeqFloat, RequestArguments.Str, RequestArguments.StringList, RequestArguments.TypeaheadInputText<T>RequestStatics.RequestTypeConstants.Extensions, Constants.Schemes, Constants.Suffixes| Modifier and Type | Field and Description |
|---|---|
DeepLearning.Activation |
activation
The activation function (non-linearity) to be used the neurons in the hidden layers.
|
long |
actual_train_samples_per_iteration |
boolean |
adaptive_rate
The implemented adaptive learning rate algorithm (ADADELTA) automatically
combines the benefits of learning rate annealing and momentum
training to avoid slow convergence.
|
boolean |
balance_classes
For imbalanced data, balance training data class counts via
over/under-sampling.
|
Key |
best_model_key
If given, store the best model so far under this key.
|
Key |
checkpoint
A model key associated with a previously trained Deep Learning
model.
|
double |
classification_stop
The stopping criteria in terms of classification error (1-accuracy) on the
training data scoring dataset.
|
boolean |
col_major |
boolean |
diagnostics
Gather diagnostics for hidden layers, such as mean and RMS values of learning
rate, momentum, weights and biases.
|
static DocGen.FieldDoc[] |
DOC_FIELDS |
static java.lang.String |
DOC_GET |
double |
epochs
The number of passes over the training dataset to be carried out.
|
double |
epsilon
The second of two hyper parameters for adaptive learning rate (ADADELTA).
|
boolean |
expert_mode
Unlock expert mode parameters than can affect model building speed,
predictive accuracy and scoring.
|
boolean |
fast_mode
Enable fast mode (minor approximation in back-propagation), should not affect results significantly.
|
boolean |
force_load_balance
Increase training speed on small datasets by splitting it into many chunks
to allow utilization of all cores.
|
int[] |
hidden
The number and size of each hidden layer in the model.
|
double[] |
hidden_dropout_ratios
A fraction of the inputs for each hidden layer to be omitted from training in order
to improve generalization.
|
boolean |
ignore_const_cols
Ignore constant training columns (no information can be gained anyway).
|
DeepLearning.InitialWeightDistribution |
initial_weight_distribution
The distribution from which initial weights are to be drawn.
|
double |
initial_weight_scale
The scale of the distribution function for Uniform or Normal distributions.
|
double |
input_dropout_ratio
A fraction of the features for each training row to be omitted from training in order
to improve generalization (dimension sampling).
|
double |
l1
A regularization method that constrains the absolute value of the weights and
has the net effect of dropping some weights (setting them to zero) from a model
to reduce complexity and avoid overfitting.
|
double |
l2
A regularization method that constrdains the sum of the squared
weights.
|
DeepLearning.Loss |
loss
The loss (error) function to be minimized by the model.
|
float |
max_after_balance_size
When classes are balanced, limit the resulting dataset size to the
specified multiple of the original dataset size.
|
int |
max_confusion_matrix_size
For classification models, the maximum size (in terms of classes) of the
confusion matrix for it to be printed.
|
int |
max_hit_ratio_k
The maximum number (top K) of predictions to use for hit ratio computation (for multi-class only, 0 to disable)
|
float |
max_w2
A maximum on the sum of the squared incoming weights into
any one neuron.
|
double |
momentum_ramp
The momentum_ramp parameter controls the amount of learning for which momentum increases
(assuming momentum_stable is larger than momentum_start).
|
double |
momentum_stable
The momentum_stable parameter controls the final momentum value reached after momentum_ramp training samples.
|
double |
momentum_start
The momentum_start parameter controls the amount of momentum at the beginning of training.
|
boolean |
nesterov_accelerated_gradient
The Nesterov accelerated gradient descent method is a modification to
traditional gradient descent for convex functions.
|
boolean |
quiet_mode
Enable quiet mode for less output to standard output.
|
double |
rate
When adaptive learning rate is disabled, the magnitude of the weight
updates are determined by the user specified learning rate
(potentially annealed), and are a function of the difference
between the predicted value and the target value.
|
double |
rate_annealing
Learning rate annealing reduces the learning rate to "freeze" into
local minima in the optimization landscape.
|
double |
rate_decay
The learning rate decay parameter controls the change of learning rate across layers.
|
double |
regression_stop
The stopping criteria in terms of regression error (MSE) on the training
data scoring dataset.
|
boolean |
replicate_training_data
Replicate the entire training dataset onto every node for faster training on small datasets.
|
double |
rho
The first of two hyper parameters for adaptive learning rate (ADADELTA).
|
double |
score_duty_cycle
Maximum fraction of wall clock time spent on model scoring on training and validation samples,
and on diagnostics such as computation of feature importances (i.e., not on training).
|
double |
score_interval
The minimum time (in seconds) to elapse between model scoring.
|
long |
score_training_samples
The number of training dataset points to be used for scoring.
|
long |
score_validation_samples
The number of validation dataset points to be used for scoring.
|
DeepLearning.ClassSamplingMethod |
score_validation_sampling
Method used to sample the validation dataset for scoring, see Score Validation Samples above.
|
long |
seed
The random seed controls sampling and initialization.
|
boolean |
shuffle_training_data
Enable shuffling of training data (on each node).
|
boolean |
single_node_mode
Run on a single node for fine-tuning of model parameters.
|
boolean |
sparse |
long |
train_samples_per_iteration
The number of training data rows to be processed per iteration.
|
boolean |
variable_importances
Whether to compute variable importances for input features.
|
_cmDomain, _names, _responseName, _sourceResponseDomain, _train, _valid, _validResponse, _validResponseDomain, validationclassificationresponsecols, ignored_cols, ignored_cols_by_namesource_fjtask, description, destination_key, end_time, exception, job_key, LIST, start_time, state_parms, response_info_requestHelp, SUPPORTS_ONLY_V1, SUPPORTS_ONLY_V2, SUPPORTS_V1_V2ARRAY_BUILDER, ARRAY_HEADER_ROW_BUILDER, ARRAY_ROW_BUILDER, ARRAY_ROW_ELEMENT_BUILDER, ARRAY_ROW_SINGLECOL_BUILDER, ELEMENT_BUILDER, GSON_BUILDER, OBJECT_BUILDER, ROOT_OBJECT_queryHtml_argumentsALPHA, ARGUMENTS, AUC, BASE, BEST_THRESHOLD, BETA_EPS, BIN_LIMIT, BROWSE, BUCKET, BUILT_IN_KEY_JOBS, CANCELLED, CARDINALITY, CASE, CASE_MODE, CHUNK, CLASS, CLOUD_HEALTH, CLOUD_NAME, CLOUD_SIZE, CLOUD_UPTIME_MILLIS, CLUSTERS, COEFFICIENTS, COL_INDEX, COLS, COLUMN_NAME, COLUMNS_DISPLAY, CONSENSUS, CONTENTS, COUNT, DATA_KEY, DEPTH, DESCRIPTION, DEST_KEY, DTHRESHOLDS, ELAPSED, END_TIME, ENUM_DOMAIN_SIZE, ERROR, ESCAPE_NAN, EXCLUSIVE_SPLIT_LIMIT, EXPRESSION, FAILED, FAMILY, FEATURES, FILE, FILES, FILTER, FIRST_CHUNK, FJ_QUEUE_HI, FJ_QUEUE_LO, FJ_THREADS_HI, FJ_THREADS_LO, FREE_DISK, FREE_MEM, HEADER, HEIGHT, HELP, IGNORE, ITEMS, ITERATIVE_CM, JOB, JOB_KEY, JOBS, JSON_H2O, KEY, KEYS, LAMBDA, LAST_CONTACT, LIMIT, LINK, LOCKED, MAX, MAX_DISK, MAX_ITER, MAX_MEM, MAX_ROWS, MEAN, MIN, MODEL_KEY, MODELS, MORE, MTRY, MTRY_NODES, NAME, NEG_X, NO_CM, NODE, NODE_HEALTH, NODE_NAME, NODES, NORMALIZE, NUM_COLS, NUM_CPUS, NUM_FAILED, NUM_KEYS, NUM_MISSING_VALUES, NUM_ROWS, NUM_SUCCEEDED, NUM_TREES, OBJECT, OFFSET, OOBEE, PARALLEL, PARSER_TYPE, PATH, PREVIEW, PREVIOUS_MODEL_KEY, PRIOR, PROGRESS, PROGRESS_KEY, PROGRESS_TOTAL, REDIRECT, REDIRECT_ARGS, REPLICATION_FACTOR, REQUEST_TIME, RESPONSE, RHO, ROW, ROW_SIZE, ROWS, RPCS, SAMPLE, SAMPLING_STRATEGY, SCALE, SEED, SENT_ROWS, SEPARATOR, SIZE, SOURCE_KEY, STACK_TRACES, START_TIME, STAT_TYPE, STATUS, STEP, STRATA_SAMPLES, SUCCEEDED, SYSTEM_LOAD, TASK_KEY, TCPS_ACTIVE, TCPS_DUTY, TIME, TO_ENUM, TOT_MEM, TREE_COUNT, TREE_DEPTH, TREE_LEAVES, TREE_NUM, TREES, TWEEDIE_POWER, TYPE, URL, USE_NON_LOCAL_DATA, VALUE, VALUE_SIZE, VALUE_TYPE, VARIANCE, VERSION, VIEW, WARNINGS, WEIGHT, WEIGHTS, WIDTH, X, XVAL, Y| Constructor and Description |
|---|
DeepLearning() |
| Modifier and Type | Method and Description |
|---|---|
void |
delete()
Delete job related keys
|
void |
execImpl()
Train a Deep Learning model, assumes that all members are populated
|
DeepLearningModel |
initModel()
Create an initial Deep Learning model, typically to be trained by trainModel(model)
|
static java.lang.String |
link(Key k,
java.lang.String content)
Return a query link to this page
|
static java.lang.String |
link(Key k,
java.lang.String content,
Key cp,
java.lang.String response,
Key val)
Return a query link to this page
|
float |
progress()
Report the relative progress of building a Deep Learning model (measured by how many epochs are done)
|
protected void |
queryArgumentValueSet(RequestArguments.Argument arg,
java.util.Properties inputArgs)
Helper to handle arguments based on existing input values
|
protected RequestBuilders.Response |
redirect()
Redirect to the model page for that model that is trained by this job
|
protected void |
registered(RequestServer.API_VERSION ver)
Helper to specify which arguments trigger a refresh on change
|
boolean |
toHTML(java.lang.StringBuilder sb)
Print model parameters as JSON
|
DeepLearningModel |
trainModel(DeepLearningModel model)
Train a Deep Learning neural net model
|
getCMDomain, getOrigValidation, getValidAdaptor, getValidation, getVectorDomain, hasValidation, init, prepareValidationWithModel, toJSONselectFrame, selectVecsall, cancel, cancel, cancel, checkIdx, defaultDestKey, defaultJobKey, dest, findJob, findJobByDest, fork, get, getState, gridParallelism, invoke, isCancelledOrCrashed, isCancelledXX, isCrashed, isDone, isEnded, isRunning, isRunning, onCancelled, remove, runTimeMs, self, serve, speedDescription, speedValue, start, waitUntilJobEnded, waitUntilJobEndedcreate, fillResponseInfo, filterNaCols, input, logStart, makeJsonBox, serveGrid, set, split, superServeGrid, supportedVersions, toStringaddToNavbar, addToNavbar, addToNavbar, DocExampleFail, DocExampleSucc, href, href, hrefType, HTMLHelp, htmlTemplate, initializeNavBar, log, mapTypeahead, ReSTHelp, serve, serveJava, serveResponse, toDocGET, toJava, wrap, wrap, wrap, writeJSONFieldsbuild, buildJSONResponseBox, buildResponseHeader, namebuildQuery, checkArgumentsarguments, argumentsToJson, frameColumnNameToIndex, vaCategoryNames, vaCategoryNames, vaColumnNameToIndexcheckJsonName, encodeRedirectArgs, JSON2HTML, jsonError, requestName, Str2JSONclone, frozenType, init, newInstance, read, toDocField, write, writeJSONpublic static DocGen.FieldDoc[] DOC_FIELDS
public static final java.lang.String DOC_GET
@Request.API(help="Model checkpoint to resume training with", filter=Request.Default.class, json=true) public Key checkpoint
@Request.API(help="Key to store the always-best model under", filter=Request.Default.class, json=true) public Key best_model_key
@Request.API(help="Enable expert mode (to access all options from GUI)", filter=Request.Default.class, json=true) public boolean expert_mode
@Request.API(help="Activation function", filter=Request.Default.class, json=true, importance=CRITICAL) public DeepLearning.Activation activation
@Request.API(help="Hidden layer sizes (e.g. 100,100). Grid search: (10,10), (20,20,20)", filter=Request.Default.class, json=true, importance=CRITICAL) public int[] hidden
@Request.API(help="How many times the dataset should be iterated (streamed), can be fractional", filter=Request.Default.class, dmin=0.001, json=true, importance=CRITICAL) public double epochs
@Request.API(help="Number of training samples (globally) per MapReduce iteration. Special values are 0: one epoch, -1: all available data (e.g., replicated training data)", filter=Request.Default.class, lmin=-1L, json=true, importance=SECONDARY) public long train_samples_per_iteration
public long actual_train_samples_per_iteration
@Request.API(help="Seed for random numbers (affects sampling) - Note: only reproducible when running single threaded", filter=Request.Default.class, json=true) public long seed
@Request.API(help="Adaptive learning rate (ADADELTA)", filter=Request.Default.class, json=true, importance=SECONDARY) public boolean adaptive_rate
@Request.API(help="Adaptive learning rate time decay factor (similarity to prior updates)", filter=Request.Default.class, dmin=0.01, dmax=1.0, json=true, importance=SECONDARY) public double rho
@Request.API(help="Adaptive learning rate smoothing factor (to avoid divisions by zero and allow progress)", filter=Request.Default.class, dmin=1.0E-15, dmax=1.0, json=true, importance=SECONDARY) public double epsilon
@Request.API(help="Learning rate (higher => less stable, lower => slower convergence)", filter=Request.Default.class, dmin=1.0E-10, dmax=1.0, json=true, importance=SECONDARY) public double rate
@Request.API(help="Learning rate annealing: rate / (1 + rate_annealing * samples)", filter=Request.Default.class, dmin=0.0, dmax=1.0, json=true, importance=SECONDARY) public double rate_annealing
@Request.API(help="Learning rate decay factor between layers (N-th layer: rate*alpha^(N-1))", filter=Request.Default.class, dmin=0.0, json=true, importance=EXPERT) public double rate_decay
@Request.API(help="Initial momentum at the beginning of training (try 0.5)", filter=Request.Default.class, dmin=0.0, dmax=0.9999999999, json=true, importance=SECONDARY) public double momentum_start
@Request.API(help="Number of training samples for which momentum increases", filter=Request.Default.class, dmin=1.0, json=true, importance=SECONDARY) public double momentum_ramp
@Request.API(help="Final momentum after the ramp is over (try 0.99)", filter=Request.Default.class, dmin=0.0, dmax=0.9999999999, json=true, importance=SECONDARY) public double momentum_stable
@Request.API(help="Use Nesterov accelerated gradient (recommended)", filter=Request.Default.class, json=true, importance=SECONDARY) public boolean nesterov_accelerated_gradient
@Request.API(help="Input layer dropout ratio (can improve generalization, try 0.1 or 0.2)", filter=Request.Default.class, dmin=0.0, dmax=1.0, json=true, importance=SECONDARY) public double input_dropout_ratio
@Request.API(help="Hidden layer dropout ratios (can improve generalization), specify one value per hidden layer, defaults to 0.5", filter=Request.Default.class, dmin=0.0, dmax=1.0, json=true, importance=SECONDARY) public double[] hidden_dropout_ratios
@Request.API(help="L1 regularization (can add stability and improve generalization, causes many weights to become 0)", filter=Request.Default.class, dmin=0.0, dmax=1.0, json=true, importance=SECONDARY) public double l1
@Request.API(help="L2 regularization (can add stability and improve generalization, causes many weights to be small", filter=Request.Default.class, dmin=0.0, dmax=1.0, json=true, importance=SECONDARY) public double l2
@Request.API(help="Constraint for squared sum of incoming weights per unit (e.g. for Rectifier)", filter=Request.Default.class, dmin=1.0E-10, json=true, importance=EXPERT) public float max_w2
@Request.API(help="Initial Weight Distribution", filter=Request.Default.class, json=true, importance=EXPERT) public DeepLearning.InitialWeightDistribution initial_weight_distribution
@Request.API(help="Uniform: -value...value, Normal: stddev)", filter=Request.Default.class, dmin=0.0, json=true, importance=EXPERT) public double initial_weight_scale
@Request.API(help="Loss function", filter=Request.Default.class, json=true, importance=EXPERT) public DeepLearning.Loss loss
@Request.API(help="Shortest time interval (in secs) between model scoring", filter=Request.Default.class, dmin=0.0, json=true, importance=SECONDARY) public double score_interval
@Request.API(help="Number of training set samples for scoring (0 for all)", filter=Request.Default.class, lmin=0L, json=true, importance=EXPERT) public long score_training_samples
@Request.API(help="Number of validation set samples for scoring (0 for all)", filter=Request.Default.class, lmin=0L, json=true, importance=EXPERT) public long score_validation_samples
@Request.API(help="Maximum duty cycle fraction for scoring (lower: more training, higher: more scoring).", filter=Request.Default.class, dmin=0.0, dmax=1.0, json=true, importance=EXPERT) public double score_duty_cycle
@Request.API(help="Stopping criterion for classification error fraction on training data (-1 to disable)", filter=Request.Default.class, dmin=-1.0, dmax=1.0, json=true, importance=EXPERT) public double classification_stop
@Request.API(help="Stopping criterion for regression error (MSE) on training data (-1 to disable)", filter=Request.Default.class, dmin=-1.0, json=true, importance=EXPERT) public double regression_stop
@Request.API(help="Enable quiet mode for less output to standard output", filter=Request.Default.class, json=true) public boolean quiet_mode
@Request.API(help="Max. size (number of classes) for confusion matrices to be shown", filter=Request.Default.class, json=true) public int max_confusion_matrix_size
@Request.API(help="Max. number (top K) of predictions to use for hit ratio computation (for multi-class only, 0 to disable)", filter=Request.Default.class, lmin=0L, json=true, importance=EXPERT) public int max_hit_ratio_k
@Request.API(help="Balance training data class counts via over/under-sampling (for imbalanced data)", filter=Request.Default.class, json=true, importance=EXPERT) public boolean balance_classes
@Request.API(help="Maximum relative size of the training data after balancing class counts (can be less than 1.0)", filter=Request.Default.class, json=true, dmin=0.001, importance=EXPERT) public float max_after_balance_size
@Request.API(help="Method used to sample validation dataset for scoring", filter=Request.Default.class, json=true, importance=EXPERT) public DeepLearning.ClassSamplingMethod score_validation_sampling
@Request.API(help="Enable diagnostics for hidden layers", filter=Request.Default.class, json=true) public boolean diagnostics
@Request.API(help="Compute variable importances for input features (Gedeon method) - can be slow for large networks", filter=Request.Default.class, json=true) public boolean variable_importances
@Request.API(help="Enable fast mode (minor approximation in back-propagation)", filter=Request.Default.class, json=true, importance=EXPERT) public boolean fast_mode
@Request.API(help="Ignore constant training columns (no information can be gained anyway)", filter=Request.Default.class, json=true, importance=EXPERT) public boolean ignore_const_cols
@Request.API(help="Force extra load balancing to increase training speed for small datasets (to keep all cores busy)", filter=Request.Default.class, json=true) public boolean force_load_balance
@Request.API(help="Replicate the entire training dataset onto every node for faster training on small datasets", filter=Request.Default.class, json=true, importance=EXPERT) public boolean replicate_training_data
@Request.API(help="Run on a single node for fine-tuning of model parameters", filter=Request.Default.class, json=true) public boolean single_node_mode
@Request.API(help="Enable shuffling of training data (recommended if training data is replicated and train_samples_per_iteration is close to #nodes x #rows)", filter=Request.Default.class, json=true, importance=EXPERT) public boolean shuffle_training_data
@Request.API(help="Sparse data handling (Experimental).", filter=Request.Default.class, json=true, importance=EXPERT) public boolean sparse
@Request.API(help="Use a column major weight matrix for input layer. Can speed up forward propagation, but might slow down backpropagation (Experimental).", filter=Request.Default.class, json=true, importance=EXPERT) public boolean col_major
protected void registered(RequestServer.API_VERSION ver)
registered in class Job.ColumnsResJobver - protected void queryArgumentValueSet(RequestArguments.Argument arg, java.util.Properties inputArgs)
queryArgumentValueSet in class RequestQueriesarg - inputArgs - public boolean toHTML(java.lang.StringBuilder sb)
public static java.lang.String link(Key k, java.lang.String content)
k - Model Keycontent - Link textpublic static java.lang.String link(Key k, java.lang.String content, Key cp, java.lang.String response, Key val)
k - Model Keycontent - Link textcp - Key to checkpoint to continue training with (optional)response - Responseval - Validation data set keypublic float progress()
public final void execImpl()
protected RequestBuilders.Response redirect()
public final DeepLearningModel initModel()
public final DeepLearningModel trainModel(DeepLearningModel model)
model - Input model (e.g., from initModel(), or from a previous training run)public void delete()