Modifier and Type | Method and Description |
---|---|
ByteBuffer |
BasicGraphExecutioner.convertToFlatBuffers(SameDiff diff,
ExecutorConfiguration configuration) |
ByteBuffer |
GraphExecutioner.convertToFlatBuffers(SameDiff diff,
ExecutorConfiguration configuration)
This method converts given SameDiff instance to FlatBuffers representation
|
INDArray[] |
BasicGraphExecutioner.executeGraph(SameDiff sd)
This method executes given graph and returns results
PLEASE NOTE: Default configuration is used
|
INDArray[] |
GraphExecutioner.executeGraph(SameDiff graph) |
INDArray[] |
BasicGraphExecutioner.executeGraph(SameDiff graph,
ExecutorConfiguration configuration)
This method executes given graph and returns results
|
INDArray[] |
GraphExecutioner.executeGraph(SameDiff graph,
ExecutorConfiguration configuration)
This method executes given graph and returns results
|
int |
BasicGraphExecutioner.registerGraph(SameDiff graph)
This method stores given graph for future execution
|
int |
GraphExecutioner.registerGraph(SameDiff graph)
This method stores given graph for future execution
|
INDArray[] |
BasicGraphExecutioner.reuseGraph(SameDiff graph,
Map<Integer,INDArray> inputs) |
INDArray[] |
GraphExecutioner.reuseGraph(SameDiff graph,
Map<Integer,INDArray> inputs) |
Modifier and Type | Field and Description |
---|---|
protected SameDiff |
DifferentialFunction.sameDiff |
protected SameDiff |
DifferentialFunctionFactory.sameDiff |
Modifier and Type | Method and Description |
---|---|
SameDiff |
DifferentialFunctionFactory.sameDiff() |
Modifier and Type | Method and Description |
---|---|
abstract void |
DifferentialFunction.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph)
Iniitialize the function from the given
Onnx.NodeProto |
abstract void |
DifferentialFunction.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph)
Initialize the function from the given
NodeDef |
Constructor and Description |
---|
DifferentialFunction(SameDiff sameDiff,
boolean inPlace,
Object[] extraArgs) |
DifferentialFunction(SameDiff sameDiff,
boolean inPlace,
SDVariable[] args)
Add the various arguments for
this function
|
DifferentialFunction(SameDiff sameDiff,
NodeDef nodeDef,
Map<String,AttrValue> attributesForNode,
GraphDef graph)
Initialize the function from the given
NodeDef |
DifferentialFunction(SameDiff sameDiff,
Object[] extraArgs) |
DifferentialFunction(SameDiff sameDiff,
Onnx.NodeProto node,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph)
Initialize the function from the given
Onnx.NodeProto |
DifferentialFunction(SameDiff sameDiff,
SDVariable[] args) |
DifferentialFunctionFactory(SameDiff sameDiff) |
Modifier and Type | Method and Description |
---|---|
void |
BaseListener.activationAvailable(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation) |
void |
Listener.activationAvailable(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation)
Called when any activation becomes available.
|
void |
BaseEvaluationListener.activationAvailable(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation) |
void |
BaseEvaluationListener.activationAvailableEvaluations(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
String varName,
INDArray activation)
|
ListenerResponse |
BaseListener.epochEnd(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis) |
ListenerResponse |
Listener.epochEnd(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis)
Called at the end of every epoch, when fitting from an iterator
|
ListenerResponse |
BaseEvaluationListener.epochEnd(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis) |
ListenerResponse |
BaseEvaluationListener.epochEndEvaluations(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis,
EvaluationRecord evaluations)
See
Listener.epochEnd(SameDiff, At, LossCurve, long) , also provided the requested evaluations |
void |
BaseListener.epochStart(SameDiff sd,
At at) |
void |
Listener.epochStart(SameDiff sd,
At at)
Called at the start of every epoch, when fitting from an iterator
|
void |
BaseEvaluationListener.epochStart(SameDiff sd,
At at) |
void |
BaseEvaluationListener.epochStartEvaluations(SameDiff sd,
At at)
|
void |
BaseListener.iterationDone(SameDiff sd,
At at,
MultiDataSet dataSet,
Loss loss) |
void |
Listener.iterationDone(SameDiff sd,
At at,
MultiDataSet dataSet,
Loss loss)
Called at the end of every iteration, after all operations (including updating parameters) has been completed
|
void |
BaseListener.iterationStart(SameDiff sd,
At at,
MultiDataSet data,
long etlMs) |
void |
Listener.iterationStart(SameDiff sd,
At at,
MultiDataSet data,
long etlTimeMs)
Called at the start of every iteration (minibatch), before any operations have been executed
|
void |
BaseListener.operationEnd(SameDiff sd,
Operation op) |
void |
Listener.operationEnd(SameDiff sd,
Operation op)
Called at the end of an operation, e.g.
|
void |
BaseListener.operationStart(SameDiff sd,
Operation op) |
void |
Listener.operationStart(SameDiff sd,
Operation op)
Called at the start of an operation, e.g.
|
void |
BaseListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs) |
void |
Listener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs)
Called at the end of each operation execution
|
ListenerVariables |
BaseEvaluationListener.otherRequiredVariables(SameDiff sd)
Return any requested variables that are not part of the evaluations
|
void |
BaseListener.preOpExecution(SameDiff sd,
At at,
SameDiffOp op) |
void |
Listener.preOpExecution(SameDiff sd,
At at,
SameDiffOp op)
Called just before each operation is executed (native code called, etc) - after all inputs etc have been set
|
void |
BaseListener.preUpdate(SameDiff sd,
At at,
Variable v,
INDArray update) |
void |
Listener.preUpdate(SameDiff sd,
At at,
Variable v,
INDArray update)
Called just before each parameter is to be updated - i.e., just before each parameter is modified.
|
ListenerVariables |
BaseListener.requiredVariables(SameDiff sd) |
ListenerVariables |
Listener.requiredVariables(SameDiff sd)
Required variables for this listener.
|
ListenerVariables |
BaseEvaluationListener.requiredVariables(SameDiff sd) |
ListenerResponse |
BaseListener.validationDone(SameDiff sd,
At at,
long validationTimeMillis) |
ListenerResponse |
Listener.validationDone(SameDiff sd,
At at,
long validationTimeMillis)
Called after the end of every epoch, once validation evaluation is done, when training
|
ListenerResponse |
BaseEvaluationListener.validationDone(SameDiff sd,
At at,
long validationTimeMillis) |
ListenerResponse |
BaseEvaluationListener.validationDoneEvaluations(SameDiff sd,
At at,
long validationTimeMillis,
EvaluationRecord evaluations)
See
Listener.validationDone(SameDiff, At, long) , also provided the requested evaluations |
Modifier and Type | Method and Description |
---|---|
static SameDiff |
CheckpointListener.loadCheckpoint(File rootDir,
int checkpointNum,
boolean loadUpdaterState)
Load a SameDiff instance for the given checkpoint that resides in the specified root directory
|
SameDiff |
CheckpointListener.loadCheckpoint(int checkpointNum,
boolean loadUpdaterState)
Load a given checkpoint number
|
static SameDiff |
CheckpointListener.loadLastCheckpoint(File rootDir,
boolean loadUpdaterState)
Load the last (most recent) checkpoint from the specified root directory
|
Modifier and Type | Method and Description |
---|---|
ListenerResponse |
CheckpointListener.epochEnd(SameDiff sameDiff,
At at,
LossCurve lossCurve,
long epochTimeMillis) |
void |
CheckpointListener.iterationDone(SameDiff sd,
At at,
MultiDataSet dataSet,
Loss loss) |
Modifier and Type | Method and Description |
---|---|
void |
OpBenchmarkListener.operationEnd(SameDiff sd,
Operation op) |
void |
OpBenchmarkListener.operationStart(SameDiff sd,
Operation op) |
void |
OpBenchmarkListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs) |
void |
ArraySavingListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs) |
void |
OpBenchmarkListener.preOpExecution(SameDiff sd,
At at,
SameDiffOp op) |
void |
ExecDebuggingListener.preOpExecution(SameDiff sd,
At at,
SameDiffOp op) |
Modifier and Type | Method and Description |
---|---|
protected void |
UIListener.checkStructureForRestore(SameDiff sd) |
ListenerResponse |
UIListener.epochEnd(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis) |
ListenerResponse |
ScoreListener.epochEnd(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis) |
ListenerResponse |
HistoryListener.epochEndEvaluations(SameDiff sd,
At at,
LossCurve lossCurve,
long epochTimeMillis,
EvaluationRecord evaluations) |
void |
UIListener.epochStart(SameDiff sd,
At at) |
void |
ScoreListener.epochStart(SameDiff sd,
At at) |
protected void |
UIListener.initalizeWriter(SameDiff sd) |
protected void |
UIListener.initializeHelper(SameDiff sd) |
void |
UIListener.iterationDone(SameDiff sd,
At at,
MultiDataSet dataSet,
Loss loss) |
void |
ScoreListener.iterationDone(SameDiff sd,
At at,
MultiDataSet dataSet,
Loss loss) |
void |
UIListener.iterationStart(SameDiff sd,
At at,
MultiDataSet data,
long etlMs) |
void |
ScoreListener.iterationStart(SameDiff sd,
At at,
MultiDataSet data,
long etlMs) |
void |
HistoryListener.operationEnd(SameDiff sd,
Operation op) |
void |
HistoryListener.operationStart(SameDiff sd,
Operation op) |
void |
UIListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs) |
void |
UIListener.preUpdate(SameDiff sd,
At at,
Variable v,
INDArray update) |
ListenerResponse |
HistoryListener.validationDoneEvaluations(SameDiff sd,
At at,
long validationTimeMillis,
EvaluationRecord evaluations) |
Modifier and Type | Field and Description |
---|---|
protected SameDiff |
SDVariable.sameDiff |
Modifier and Type | Method and Description |
---|---|
static SameDiff |
SameDiff.create()
Create a new (empty) SameDiff instance without any functions or variables
|
static SameDiff |
SameDiff.create(SameDiff originalSameDiff)
Create a new SameDiff instance from an existing instance.
|
SameDiff |
SameDiff.defineFunction(String function,
SameDiffFunctionDefinition functionDefinition,
SDVariable[] variables) |
SameDiff |
SameDiff.disableDebugging()
Clears debugging state and disables debug mode.
|
SameDiff |
SameDiff.dup()
Clone/duplicate the SameDiff instance, including arrays etc.
|
SameDiff |
SameDiff.enableDebugMode()
Enables tracing of graphs automatically.
|
static SameDiff |
SameDiff.fromFlatBuffers(ByteBuffer bbIn)
Create a
SameDiff
instance from a byte buffers
instance. |
static SameDiff |
SameDiff.fromFlatBuffers(ByteBuffer bbIn,
boolean loadUpdaterState)
Create a
SameDiff
instance from a byte buffers
instance. |
static SameDiff |
SameDiff.fromFlatFile(File file)
Create a
SameDiff instance from a file, including the updater state
The method to save the file is save(File, boolean) |
static SameDiff |
SameDiff.fromFlatFile(File file,
boolean loadUpdaterState)
Create a
SameDiff instance from a file, optionally also loading the updater state
The method to save the file is save(File, boolean) |
SameDiff |
SameDiff.getFunction(String functionName)
Get a SameDiff function instance given the name of the function
|
static SameDiff |
SameDiff.importFrozenTF(File graphFile)
Import a frozen Tensorflow graph to a new SameDiff graph.
|
static SameDiff |
SameDiff.importFrozenTF(GraphDef graphDef)
|
static SameDiff |
SameDiff.importFrozenTF(InputStream graph)
|
static SameDiff |
SameDiff.load(File file,
boolean loadUpdaterState)
Load the SameDiff instance previously saved with
save(File, boolean) |
static SameDiff |
SameDiff.load(InputStream is,
boolean loadUpdaterState)
As per
load(File, boolean) but the SameDiff instance |
protected SameDiff |
SameDiff.sd() |
Modifier and Type | Method and Description |
---|---|
protected int |
SameDiff.asFlatNode(String name,
SameDiff scope,
com.google.flatbuffers.FlatBufferBuilder bufferBuilder) |
SDVariable |
SDVariable.clone(SameDiff sd) |
static SameDiff |
SameDiff.create(SameDiff originalSameDiff)
Create a new SameDiff instance from an existing instance.
|
SDVariable |
SameDiffNoArgSingleLambda.define(SameDiff sameDiff) |
SDVariable[] |
SameDiffFunctionDefinition.define(SameDiff sameDiff,
Map<String,INDArray> inputs,
SDVariable[] variableInputs) |
SDVariable |
SameDiffSingleLambda.define(SameDiff sameDiff,
SDVariable[] inputs) |
SDVariable[] |
SameDiffLambda.define(SameDiff sameDiff,
SDVariable[] inputs) |
SDVariable |
SameDiffConditional.eval(SameDiff context,
SameDiffFunctionDefinition body,
SDVariable[] inputVars) |
SDVariable |
SameDiff.invokeFunctionOn(String functionName,
SameDiff with) |
SDVariable |
SameDiff.invokeGraphOn(SameDiff sameDiff) |
void |
SameDiff.putSubFunction(String name,
SameDiff nameSpace)
Associate a
SameDiff namespace as a sub function. |
Constructor and Description |
---|
NameScope(SameDiff sameDiff,
String name) |
SDVariable(String varName,
VariableType varType,
SameDiff sameDiff,
long[] shape,
DataType dataType) |
Constructor and Description |
---|
BatchOutputConfig(SameDiff sd) |
EvaluationConfig(SameDiff sd) |
FitConfig(SameDiff sd) |
OutputConfig(SameDiff sd) |
Modifier and Type | Method and Description |
---|---|
SDVariable |
DefaultSameDiffConditional.eval(SameDiff context,
SameDiffFunctionDefinition body,
SDVariable[] inputVars) |
Modifier and Type | Field and Description |
---|---|
protected SameDiff |
AbstractSession.sameDiff |
Constructor and Description |
---|
AbstractSession(SameDiff sameDiff) |
InferenceSession(SameDiff sameDiff) |
TrainingSession(SameDiff sameDiff) |
Constructor and Description |
---|
CloseValidationMemoryMgr(SameDiff sd,
SessionMemMgr underlying) |
Modifier and Type | Field and Description |
---|---|
protected SameDiff |
SDOps.sd |
Modifier and Type | Method and Description |
---|---|
protected abstract SameDiff |
SDBaseOps.sd() |
Constructor and Description |
---|
SDBitwise(SameDiff sameDiff) |
SDCNN(SameDiff sameDiff) |
SDImage(SameDiff sameDiff) |
SDLoss(SameDiff sameDiff) |
SDMath(SameDiff sameDiff) |
SDNN(SameDiff sameDiff) |
SDOps(SameDiff sameDiff) |
SDRandom(SameDiff sd) |
SDRNN(SameDiff sameDiff) |
Modifier and Type | Method and Description |
---|---|
static int |
FlatBuffersMapper.asFlatNode(SameDiff sameDiff,
DifferentialFunction node,
com.google.flatbuffers.FlatBufferBuilder bufferBuilder,
List<SDVariable> variables,
Map<String,Integer> reverseMap,
Map<String,Integer> forwardMap,
Map<String,Integer> framesMap,
AtomicInteger idCounter,
Integer id) |
static DifferentialFunction |
FlatBuffersMapper.cloneViaSerialize(SameDiff sd,
DifferentialFunction df) |
static DifferentialFunction |
FlatBuffersMapper.cloneViaSerialize(SameDiff sd,
DifferentialFunction df,
Map<String,Integer> nameToIdxMap) |
Modifier and Type | Field and Description |
---|---|
protected SameDiff |
SubGraph.sameDiff |
Modifier and Type | Method and Description |
---|---|
static SameDiff |
GraphTransformUtil.replaceSubgraphsMatching(SameDiff sd,
SubGraphPredicate p,
SubGraphProcessor processor)
Find all of the subgraphs that match the specified SubGraphPredicate and then replace them with a different subgraph.
Note that the original SameDiff instance is not modified; a copy is made, which is then modified and returned. |
Modifier and Type | Method and Description |
---|---|
SubGraph |
SubGraphPredicate.getSubGraph(SameDiff sd,
DifferentialFunction rootFn)
Get the SubGraph that matches the predicate
|
static List<SubGraph> |
GraphTransformUtil.getSubgraphsMatching(SameDiff sd,
SubGraphPredicate p)
Get a list of all the subgraphs that match the specified predicate
|
boolean |
SubGraphPredicate.matches(SameDiff sameDiff,
DifferentialFunction rootFn)
Determine if the subgraph, starting with the root function, matches the predicate
|
abstract boolean |
OpPredicate.matches(SameDiff sameDiff,
DifferentialFunction function) |
List<SDVariable> |
SubGraphProcessor.processSubgraph(SameDiff sd,
SubGraph subGraph)
Replace the subgraph, and return the new outputs that should replace the old outputs.
Note that the order of the outputs you return matters! If the original outputs are [A,B,C] and you return output variables [X,Y,Z], then anywhere "A" was used as input will now use "X"; similarly Y replaces B, and Z replaces C. |
static SameDiff |
GraphTransformUtil.replaceSubgraphsMatching(SameDiff sd,
SubGraphPredicate p,
SubGraphProcessor processor)
Find all of the subgraphs that match the specified SubGraphPredicate and then replace them with a different subgraph.
Note that the original SameDiff instance is not modified; a copy is made, which is then modified and returned. |
Modifier and Type | Method and Description |
---|---|
static void |
OpValidation.checkDeserializedEquality(SameDiff original,
ByteBuffer bbSerialized,
TestCase tc) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
boolean print,
boolean exitOnFirstFailure) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure,
boolean skipValidation,
boolean debugMode,
Set<String> skipVariables,
Map<String,INDArray> gradCheckMask) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
double eps,
double maxRelError,
double minAbsError,
boolean print,
boolean exitOnFirstFailure,
boolean skipValidation,
boolean debugMode,
Set<String> skipVariables,
Map<String,INDArray> gradCheckMask,
int maxPerParam,
GradCheckUtil.Subset subset) |
static boolean |
GradCheckUtil.checkGradients(SameDiff sd,
Map<String,INDArray> placeholderValues,
String... skipVariables) |
static void |
OpValidation.collectTensorflowImportCoverage(SameDiff graph) |
void |
ActivationGradientCheckListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs) |
static void |
GradCheckUtil.validateInternalState(SameDiff sd,
boolean generateAndCheckGradFn) |
Constructor and Description |
---|
TestCase(SameDiff sameDiff) |
Modifier and Type | Method and Description |
---|---|
void |
NonInplaceValidationListener.opExecution(SameDiff sd,
At at,
MultiDataSet batch,
SameDiffOp op,
INDArray[] outputs) |
void |
NonInplaceValidationListener.preOpExecution(SameDiff sd,
At at,
SameDiffOp op) |
Modifier and Type | Method and Description |
---|---|
protected int |
LogFileWriter.encodeGraphStructure(com.google.flatbuffers.FlatBufferBuilder fbb,
SameDiff sd) |
long |
LogFileWriter.writeGraphStructure(SameDiff sd)
Write the graph structure
|
Modifier and Type | Method and Description |
---|---|
List<SDVariable> |
OpImportOverride.initFromTensorFlow(List<SDVariable> inputs,
List<SDVariable> controlDepInputs,
NODE_TYPE nodeDef,
SameDiff initWith,
Map<String,ATTR_TYPE> attributesForNode,
GRAPH_TYPE graph)
Initialize the operation and return its output variables
|
boolean |
OpImportFilter.skipOp(NODE_TYPE nodeDef,
SameDiff initWith,
Map<String,ATTR_TYPE> attributesForNode,
GRAPH_TYPE graph)
If true: the op should be skipped for import, and its output variables should not be created.
|
Modifier and Type | Method and Description |
---|---|
static SameDiff |
TFGraphMapper.importGraph(File f)
Import a frozen TensorFlow protobuf (.pb) file from the specified file
|
static SameDiff |
TFGraphMapper.importGraph(File f,
Map<String,TFImportOverride> importOverride,
TFOpImportFilter opFilter)
Import a frozen TensorFlow protobuf (.pb) file from the specified file, with optional overrides
|
static SameDiff |
TFGraphMapper.importGraph(GraphDef tfGraph)
Import a TensorFlow model from a GraphDef
|
static SameDiff |
TFGraphMapper.importGraph(GraphDef tfGraph,
Map<String,TFImportOverride> importOverride,
TFOpImportFilter opFilter)
Import a TensorFlow model from a GraphDef, with optional import overrides
|
static SameDiff |
TFGraphMapper.importGraph(InputStream is)
Import a frozen TensorFlow protobuf (.pb) file, via an input stream
|
static SameDiff |
TFGraphMapper.importGraph(InputStream is,
Map<String,TFImportOverride> importOverride,
TFOpImportFilter opFilter)
Import a frozen TensorFlow protobuf (.pb) file via an input stream, with optional overrides
|
static SameDiff |
TFGraphMapper.importGraphTxt(InputStream is,
Map<String,TFImportOverride> importOverride,
TFOpImportFilter opFilter)
Import a frozen TensorFlow protobuf file in text format (.pb.txt) file via an input stream, with optional overrides
|
Modifier and Type | Method and Description |
---|---|
SDVariable |
Activation.asSameDiff(SameDiff sd,
SDVariable input)
Get the Activation as a SameDiff variable
|
SDVariable |
Activation.asSameDiff(String variableName,
SameDiff sd,
SDVariable input)
Get the Activation as a SameDiff variable
|
Modifier and Type | Method and Description |
---|---|
void |
BaseOp.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
BaseBroadcastBoolOp.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
DynamicCustomOp.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
BaseReduceOp.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
BaseBroadcastOp.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
NoOp.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
BaseOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BaseBroadcastBoolOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DynamicCustomOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BaseReduceOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BaseBroadcastOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
NoOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
BaseBroadcastBoolOp(SameDiff sameDiff) |
BaseBroadcastBoolOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BaseBroadcastBoolOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BaseBroadcastBoolOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BaseBroadcastBoolOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BaseBroadcastBoolOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BaseBroadcastBoolOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BaseBroadcastOp(SameDiff sameDiff) |
BaseBroadcastOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BaseBroadcastOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BaseBroadcastOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BaseBroadcastOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BaseBroadcastOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BaseBroadcastOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BaseIndexAccumulation(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
BaseIndexAccumulation(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
boolean keepDims,
int[] dimensions) |
BaseOp(SameDiff sameDiff,
boolean inPlace,
Object[] extraArgs) |
BaseOp(SameDiff sameDiff,
Object[] extraArgs) |
BaseReduceBoolOp(SameDiff sameDiff,
SDVariable input,
int... dimensions) |
BaseReduceBoolOp(SameDiff sameDiff,
SDVariable input,
int[] dimensions,
boolean keepDims) |
BaseReduceBoolOp(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
BaseReduceFloatOp(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
BaseReduceFloatOp(SameDiff sameDiff,
SDVariable input,
int... dimensions) |
BaseReduceFloatOp(SameDiff sameDiff,
SDVariable input,
int[] dimensions,
boolean keepDims) |
BaseReduceFloatOp(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
BaseReduceLongOp(SameDiff sameDiff,
SDVariable input,
int... dimensions) |
BaseReduceLongOp(SameDiff sameDiff,
SDVariable input,
int[] dimensions,
boolean keepDims) |
BaseReduceLongOp(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
BaseReduceOp(SameDiff sameDiff) |
BaseReduceOp(SameDiff sameDiff,
SDVariable i_v) |
BaseReduceOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
BaseReduceOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions,
boolean keepDims) |
BaseReduceOp(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
BaseReduceOp(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions,
boolean keepDims) |
BaseReduceSameOp(SameDiff sameDiff,
SDVariable input,
int... dimensions) |
BaseReduceSameOp(SameDiff sameDiff,
SDVariable input,
int[] dimensions,
boolean keepDims) |
BaseReduceSameOp(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
BaseScalarBoolOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
BaseScalarBoolOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
BaseScalarBoolOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace,
Object[] extraArgs) |
BaseScalarBoolOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
Object[] extraArgs) |
BaseScalarOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
BaseScalarOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
BaseScalarOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace,
Object[] extraArgs) |
BaseScalarOp(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
Object[] extraArgs) |
BaseTransformAnyOp(SameDiff sameDiff) |
BaseTransformAnyOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BaseTransformAnyOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
Object[] extraArgs) |
BaseTransformAnyOp(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs) |
BaseTransformAnyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BaseTransformAnyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
BaseTransformAnyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
Object[] extraArgs) |
BaseTransformBoolOp(SameDiff sameDiff) |
BaseTransformBoolOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BaseTransformBoolOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
Object[] extraArgs) |
BaseTransformBoolOp(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs) |
BaseTransformBoolOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BaseTransformBoolOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
BaseTransformBoolOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
Object[] extraArgs) |
BaseTransformFloatOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BaseTransformFloatOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
Object[] extraArgs) |
BaseTransformFloatOp(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs) |
BaseTransformOp(SameDiff sameDiff) |
BaseTransformOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BaseTransformOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
Object[] extraArgs) |
BaseTransformOp(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs) |
BaseTransformOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BaseTransformOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
BaseTransformOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
Object[] extraArgs) |
BaseTransformSameOp(SameDiff sameDiff) |
BaseTransformSameOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BaseTransformSameOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
Object[] extraArgs) |
BaseTransformSameOp(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs) |
BaseTransformSameOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BaseTransformSameOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
BaseTransformSameOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
Object[] extraArgs) |
BaseTransformStrictOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BaseTransformStrictOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
Object[] extraArgs) |
BaseTransformStrictOp(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs) |
BaseTransformStrictOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BaseTransformStrictOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
DynamicCustomOp(SameDiff sameDiff,
SDVariable arg) |
DynamicCustomOp(SameDiff sameDiff,
SDVariable[] args) |
DynamicCustomOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
DynamicCustomOp(String opName,
SameDiff sameDiff,
SDVariable[] args) |
DynamicCustomOp(String opName,
SameDiff sameDiff,
SDVariable[] args,
boolean inPlace)
Initialize this for
SameDiff execution
Any extra int or float arguments for operations
must be added to the respective TArguments
or IArguments lists upon construction |
NoOp(SameDiff sd,
SDVariable in) |
Modifier and Type | Method and Description |
---|---|
void |
FakeQuantWithMinMaxVarsPerChannel.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BitCast.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Modifier and Type | Method and Description |
---|---|
void |
BroadcastTo.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BiasAdd.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
BiasAdd(SameDiff sameDiff,
SDVariable input,
SDVariable bias,
boolean nchw) |
BiasAddGrad(SameDiff sameDiff,
SDVariable input,
SDVariable bias,
SDVariable gradient,
boolean nchw) |
BroadcastAddOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastAddOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastAddOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastAMax(SameDiff sameDiff) |
BroadcastAMax(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastAMax(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastAMax(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastAMax(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastAMax(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastAMax(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastAMin(SameDiff sameDiff) |
BroadcastAMin(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastAMin(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastAMin(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastAMin(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastAMin(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastAMin(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastCopyOp(SameDiff sameDiff) |
BroadcastCopyOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastCopyOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastCopyOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastCopyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastCopyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastCopyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastDivOp(SameDiff sameDiff) |
BroadcastDivOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastDivOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastDivOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastGradientArgs(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastGradientArgs(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastGradientArgs(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastMax(SameDiff sameDiff) |
BroadcastMax(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastMax(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastMax(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastMax(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastMax(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastMax(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastMin(SameDiff sameDiff) |
BroadcastMin(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastMin(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastMin(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastMin(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastMin(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastMin(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastMulOp(SameDiff sameDiff) |
BroadcastMulOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastMulOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastMulOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastMulOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastMulOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastMulOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastRDivOp(SameDiff sameDiff) |
BroadcastRDivOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastRDivOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastRDivOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastRDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastRDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastRDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastRSubOp(SameDiff sameDiff) |
BroadcastRSubOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastRSubOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastRSubOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastRSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastRSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastRSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastSubOp(SameDiff sameDiff) |
BroadcastSubOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastSubOp(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastSubOp(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastTo(SameDiff sameDiff,
SDVariable input,
SDVariable shape) |
Constructor and Description |
---|
BroadcastEqualTo(SameDiff sameDiff) |
BroadcastEqualTo(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastEqualTo(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastEqualTo(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastGreaterThan(SameDiff sameDiff) |
BroadcastGreaterThan(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastGreaterThan(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastGreaterThan(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastGreaterThan(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastGreaterThan(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastGreaterThan(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastGreaterThanOrEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastLessThan(SameDiff sameDiff) |
BroadcastLessThan(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastLessThan(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastLessThan(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastLessThan(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastLessThan(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastLessThan(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastLessThanOrEqual(SameDiff sameDiff) |
BroadcastLessThanOrEqual(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastLessThanOrEqual(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastLessThanOrEqual(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastLessThanOrEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastLessThanOrEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastLessThanOrEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
BroadcastNotEqual(SameDiff sameDiff) |
BroadcastNotEqual(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
boolean inPlace) |
BroadcastNotEqual(SameDiff sameDiff,
SDVariable i_v,
int[] dimension,
Object[] extraArgs) |
BroadcastNotEqual(SameDiff sameDiff,
SDVariable i_v,
long[] shape,
boolean inPlace,
int[] dimension,
Object[] extraArgs) |
BroadcastNotEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
int[] dimension) |
BroadcastNotEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension) |
BroadcastNotEqual(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
int[] dimension,
Object[] extraArgs) |
Modifier and Type | Method and Description |
---|---|
void |
Select.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
Select(SameDiff sameDiff,
SDVariable[] args) |
Select(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
Where(SameDiff sameDiff,
SDVariable[] args) |
Where(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
WhereNumpy(SameDiff sameDiff,
SDVariable[] args) |
WhereNumpy(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
Modifier and Type | Method and Description |
---|---|
void |
Exit.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Switch.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
LoopCond.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
NextIteration.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BaseCompatOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Enter.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Merge.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
BaseCompatOp(SameDiff sameDiff,
SDVariable[] inputs) |
Enter(SameDiff sameDiff,
SDVariable[] inputs) |
Enter(SameDiff sameDiff,
String frameName,
SDVariable input) |
Enter(SameDiff sameDiff,
String frameName,
SDVariable input,
boolean isConstant) |
Exit(SameDiff sameDiff,
SDVariable x) |
Merge(SameDiff sd,
SDVariable[] inputs) |
Merge(SameDiff sd,
SDVariable a,
SDVariable b) |
NextIteration(SameDiff sameDiff,
SDVariable x) |
StopGradient(SameDiff sd,
SDVariable in) |
Switch(SameDiff sameDiff,
SDVariable input,
SDVariable predicate) |
Modifier and Type | Method and Description |
---|---|
void |
CropAndResize.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ResizeNearestNeighbor.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ResizeBilinear.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ResizeBicubic.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ExtractImagePatches.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
CropAndResize(SameDiff sameDiff,
SDVariable image,
SDVariable cropBoxes,
SDVariable boxIndices,
SDVariable cropOutSize,
CropAndResize.Method method,
double extrapolationValue) |
ExtractImagePatches(SameDiff samediff,
SDVariable input,
int[] kSizes,
int[] strides,
int[] rates,
boolean sameMode) |
NonMaxSuppression(SameDiff sameDiff,
SDVariable boxes,
SDVariable scores,
SDVariable maxOutSize,
SDVariable iouThreshold,
SDVariable scoreThreshold) |
NonMaxSuppressionV3(SameDiff sameDiff,
SDVariable boxes,
SDVariable scores,
SDVariable maxOutSize,
SDVariable iouThreshold,
SDVariable scoreThreshold) |
ResizeBicubic(SameDiff sameDiff,
SDVariable image,
SDVariable size,
boolean alignCorners,
boolean alignPixelCenters) |
ResizeBilinear(SameDiff sd,
SDVariable input,
int height,
int width,
boolean alignCorners,
boolean halfPixelCenters) |
Constructor and Description |
---|
FirstIndex(SameDiff sameDiff,
SDVariable i_v,
Condition condition,
boolean keepDims,
int... dimensions) |
IAMax(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
IAMin(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
IMax(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
IMin(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
LastIndex(SameDiff sameDiff,
SDVariable i_v,
Condition condition,
boolean keepDims,
int... dimensions) |
Modifier and Type | Method and Description |
---|---|
void |
ArgMin.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ArgMax.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Modifier and Type | Method and Description |
---|---|
void |
ExternalErrorsFunction.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
ExternalErrorsFunction.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
ExternalErrorsFunction(SameDiff sd,
List<SDVariable> inputs,
Map<String,INDArray> gradients) |
Modifier and Type | Method and Description |
---|---|
void |
Pooling2D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
MaxPooling3D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
DeConv2D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
LocalResponseNormalization.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
BatchNorm.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
DepthwiseConv2D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
AvgPooling2D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
AvgPooling3D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Conv2D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
MaxPooling2D.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Pooling2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DeConv2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DeConv2DTF.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
LocalResponseNormalization.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Conv3D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
SpaceToDepth.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Pooling3D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BatchNorm.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DepthwiseConv2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
AvgPooling2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DeConv3DTF.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
MaxPoolWithArgmax.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DepthToSpace.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Conv2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
MaxPooling2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
AvgPooling2D(SameDiff sameDiff,
SDVariable input,
Pooling2DConfig config) |
AvgPooling3D(SameDiff sameDiff,
INDArray arrayInput,
INDArray arrayOutput,
Pooling3DConfig config) |
AvgPooling3D(SameDiff sameDiff,
SDVariable input,
Pooling3DConfig config) |
BatchNorm(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputArrays,
boolean inPlace,
boolean applyGamma,
boolean applyBeta,
double epsilon,
int[] axis) |
BatchNormDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputArrays,
boolean inPlace,
boolean applyGamma,
boolean applyBeta,
double epsilon,
int[] axis) |
Col2Im(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputs,
Conv2DConfig conv2DConfig) |
Col2Im(SameDiff sd,
SDVariable input,
Conv2DConfig config) |
Conv1D(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv1DConfig config) |
Conv1DDerivative(SameDiff sameDiff,
SDVariable[] inputs,
Conv1DConfig config) |
Conv1DDerivative(SameDiff sd,
SDVariable input,
SDVariable weights,
SDVariable bias,
SDVariable gradOut,
Conv1DConfig config) |
Conv2D(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv2DConfig config) |
Conv2DDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv2DConfig config) |
Conv3D(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv3DConfig config) |
Conv3DDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv3DConfig conv3DConfig) |
DeConv2D(SameDiff sameDiff,
SDVariable[] inputs,
DeConv2DConfig config) |
DeConv2DDerivative(SameDiff sameDiff,
SDVariable[] inputs,
DeConv2DConfig config) |
DeConv2DTF(SameDiff sameDiff,
SDVariable[] inputs,
DeConv2DConfig config) |
DeConv3D(SameDiff sameDiff,
SDVariable input,
SDVariable weights,
SDVariable bias,
DeConv3DConfig config) |
DeConv3DDerivative(SameDiff sameDiff,
SDVariable input,
SDVariable weights,
SDVariable bias,
SDVariable grad,
DeConv3DConfig config) |
DeConv3DTF(SameDiff sameDiff,
SDVariable shape,
SDVariable weights,
SDVariable input,
DeConv3DConfig config) |
DepthToSpace(SameDiff sameDiff,
SDVariable[] args,
int blockSize,
String dataFormat) |
DepthwiseConv2D(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv2DConfig config) |
Im2col(SameDiff sameDiff,
SDVariable[] inputFunctions,
INDArray[] inputArrays,
INDArray[] outputs,
Conv2DConfig conv2DConfig) |
Im2col(SameDiff sd,
SDVariable input,
Conv2DConfig config) |
Im2colBp(SameDiff sd,
SDVariable input,
Conv2DConfig config) |
Im2colBp(SameDiff sameDiff,
SDVariable i2cInput,
SDVariable gradAtOutput,
Conv2DConfig conv2DConfig) |
LocalResponseNormalization(SameDiff sameDiff,
SDVariable[] inputFunctions,
boolean inPlace,
LocalResponseNormalizationConfig config) |
LocalResponseNormalizationDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
boolean inPlace,
LocalResponseNormalizationConfig config) |
MaxPooling2D(SameDiff sameDiff,
SDVariable input,
Pooling2DConfig config) |
MaxPooling3D(SameDiff sameDiff,
INDArray arrayInput,
INDArray arrayOutput,
Pooling3DConfig config) |
MaxPooling3D(SameDiff sameDiff,
SDVariable input,
Pooling3DConfig config) |
MaxPoolWithArgmax(SameDiff sameDiff,
SDVariable input,
Pooling2DConfig config) |
Pooling2D(SameDiff sameDiff,
SDVariable[] inputs,
Pooling2DConfig config) |
Pooling2DDerivative(SameDiff sameDiff,
SDVariable[] inputs,
Pooling2DConfig config) |
Pooling3D(SameDiff sameDiff,
SDVariable[] inputs,
INDArray[] inputArrays,
INDArray[] outputs,
boolean inPlace,
Pooling3DConfig pooling3DConfig,
Pooling3D.Pooling3DType type) |
Pooling3DDerivative(SameDiff sameDiff,
SDVariable[] inputs,
INDArray[] inputArrays,
INDArray[] outputs,
boolean inPlace,
Pooling3DConfig pooling3DConfig,
Pooling3D.Pooling3DType type) |
SConv2D(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv2DConfig conv2DConfig) |
SConv2DDerivative(SameDiff sameDiff,
SDVariable[] inputFunctions,
Conv2DConfig conv2DConfig) |
SpaceToDepth(SameDiff sameDiff,
SDVariable[] args,
int blockSize,
String dataFormat) |
Upsampling2d(SameDiff sameDiff,
SDVariable input,
boolean nchw,
int scaleH,
int scaleW) |
Upsampling2dDerivative(SameDiff sameDiff,
SDVariable input,
SDVariable gradient,
boolean nchw,
int scaleH,
int scaleW) |
Modifier and Type | Method and Description |
---|---|
void |
SRU.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
SRUCell.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
LSTMCell.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
SRU.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
LSTMBlockCell.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
LSTMLayer.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
SRUCell.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
LSTMCell.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
GRUCell(SameDiff sameDiff,
SDVariable x,
SDVariable hLast,
GRUWeights weights) |
LSTMBlockCell(SameDiff sameDiff,
SDVariable x,
SDVariable cLast,
SDVariable yLast,
LSTMWeights weights,
LSTMConfiguration configuration) |
LSTMCell(SameDiff sameDiff,
LSTMCellConfiguration configuration) |
LSTMLayer(SameDiff sameDiff,
SDVariable maxTSLength,
SDVariable x,
SDVariable cLast,
SDVariable yLast,
LSTMWeights weights,
LSTMConfiguration configuration) |
SRU(SameDiff sameDiff,
SDVariable x,
SDVariable initialC,
SDVariable mask,
SRUWeights weights) |
SRUCell(SameDiff sameDiff,
SDVariable x,
SDVariable cLast,
SRUWeights weights) |
Modifier and Type | Method and Description |
---|---|
void |
SoftmaxCrossEntropyLoss.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
SparseSoftmaxCrossEntropyLossWithLogits.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Modifier and Type | Method and Description |
---|---|
void |
TensorMmul.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Mmul.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorMmul.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Mmul.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
NormalizeMoments.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Moments.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
All(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
Any(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
IsInf(SameDiff sameDiff,
SDVariable i_v,
int[] dims) |
IsInf(SameDiff sameDiff,
SDVariable i_v,
int[] dims,
boolean keepDims) |
IsNaN(SameDiff sameDiff,
SDVariable i_v,
int[] dims) |
IsNaN(SameDiff sameDiff,
SDVariable i_v,
int[] dims,
boolean keepDims) |
Constructor and Description |
---|
BaseReductionBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
BaseReductionBp(SameDiff sameDiff,
SDVariable origInput1,
SDVariable origInput2,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
CumProdBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean exclusive,
boolean reverse,
int... axis) |
CumSumBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean exclusive,
boolean reverse,
int... axis) |
DotBp(SameDiff sameDiff,
SDVariable origInput1,
SDVariable origInput2,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
MaxBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
MeanBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
MinBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
Norm1Bp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
Norm2Bp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
NormMaxBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
ProdBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
SquaredNormBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
StandardDeviationBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean biasCorrected,
boolean keepDims,
int... dimensions) |
SumBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean keepDims,
int... dimensions) |
VarianceBp(SameDiff sameDiff,
SDVariable origInput,
SDVariable gradAtOutput,
boolean biasCorrected,
boolean keepDims,
int... dimensions) |
Constructor and Description |
---|
BatchMmul(SameDiff sameDiff,
SDVariable[] matrices,
boolean transposeA,
boolean transposeB) |
LogSumExp(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Constructor and Description |
---|
AMean(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
AMean(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
Bias(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions,
double mean) |
Bias(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions,
double mean) |
Entropy(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
LogEntropy(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
Mean(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Norm1(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Norm2(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
NormMax(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
NormMax(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
ShannonEntropy(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
ShannonEntropy(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
SquaredNorm(SameDiff sameDiff,
SDVariable input,
boolean keepDims,
int... dimensions) |
Constructor and Description |
---|
CountNonZero(SameDiff sameDiff,
SDVariable input,
int... dimensions) |
CountZero(SameDiff sameDiff,
SDVariable input,
int... dimensions) |
MatchCondition(SameDiff sameDiff,
SDVariable in,
Condition condition,
boolean keepDims,
int... dimensions) |
Constructor and Description |
---|
AMax(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
AMax(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
AMin(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
AMin(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
ASum(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
ASum(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
Max(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Max(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
Min(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Prod(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Prod(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
Sum(SameDiff sameDiff,
SDVariable i_v,
boolean keepDims,
int[] dimensions) |
Sum(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
Modifier and Type | Method and Description |
---|---|
static List<SDVariable> |
CosineSimilarity.doDiff(SameDiff sameDiff,
DifferentialFunctionFactory f,
SDVariable x,
SDVariable y,
SDVariable gradOut,
boolean keepDims,
int... dimensions) |
Constructor and Description |
---|
BaseReduce3Op(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
BaseReduce3Op(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int... dimensions) |
CosineDistance(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int... dimensions) |
CosineSimilarity(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
CosineSimilarity(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
Dot(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int... dimensions) |
EqualsWithEps(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions,
double eps) |
EqualsWithEps(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions,
double eps) |
EuclideanDistance(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
EuclideanDistance(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int[] dimensions) |
HammingDistance(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int... dimensions) |
JaccardDistance(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int... dimensions) |
ManhattanDistance(SameDiff sameDiff,
SDVariable i_v,
int[] dimensions) |
ManhattanDistance(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2,
int... dimensions) |
Modifier and Type | Method and Description |
---|---|
void |
Relu6.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
LeakyReLU.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
LeakyReLU(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double alpha) |
LeakyReLU(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs,
double alpha) |
LogX(SameDiff sameDiff,
SDVariable i_v,
double base) |
Pow(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double pow) |
Pow(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs,
double pow) |
PowDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double pow) |
PRelu(SameDiff sameDiff,
SDVariable x,
SDVariable alpha,
int... sharedAxes) |
RectifiedLinear(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double cutoff) |
RectifiedLinearDerivative(SameDiff sd,
SDVariable input,
SDVariable gradient) |
Relu6(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double cutoff) |
ReplaceNans(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double set) |
ReplaceNans(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs,
double set) |
ScalarAdd(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarAdd(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace,
Object[] extraArgs) |
ScalarAdd(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
Object[] extraArgs) |
ScalarDivision(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarDivision(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarFMod(SameDiff sd,
SDVariable in,
Number number) |
ScalarMax(SameDiff sd,
SDVariable in,
Number number) |
ScalarMin(SameDiff sd,
SDVariable in,
Number number) |
ScalarMultiplication(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarMultiplication(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarRemainder(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarRemainder(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarReverseDivision(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarReverseDivision(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarReverseSubtraction(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarReverseSubtraction(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarSet(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarSet(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
ScalarSubtraction(SameDiff sameDiff,
SDVariable i_v,
Number scalar) |
ScalarSubtraction(SameDiff sameDiff,
SDVariable i_v,
Number scalar,
boolean inPlace) |
Step(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double cutoff) |
Modifier and Type | Method and Description |
---|---|
void |
ScatterNd.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterAdd.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterNdAdd.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterSub.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterDiv.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterMin.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterUpdate.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterMul.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterMax.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterNdSub.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ScatterNdUpdate.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Modifier and Type | Method and Description |
---|---|
void |
DiagPart.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
ParallelStack.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
MergeMax.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Gather.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Repeat.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Diag.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Unstack.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Shape.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Rank.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
MergeAvg.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Reshape.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Transpose.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
MergeSum.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Stack.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
DiagPart.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Concat.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ConfusionMatrix.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Create.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Split.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
OneHot.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
StridedSlice.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ZerosLike.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ParallelStack.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Squeeze.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
MergeMax.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
OnesLike.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Gather.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ApplyGradientDescent.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Repeat.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Diag.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Unstack.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
SplitV.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Shape.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Linspace.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
SequenceMask.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Size.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
MergeAvg.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Reshape.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Tile.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Transpose.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ShapeN.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ExpandDims.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
MergeSum.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Stack.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
BroadcastDynamicShape(SameDiff sameDiff,
SDVariable in,
SDVariable shape) |
Concat(SameDiff sameDiff,
int concatDimension,
SDVariable... inputs) |
ConfusionMatrix(SameDiff sameDiff,
SDVariable labels,
SDVariable pred,
DataType dataType) |
ConfusionMatrix(SameDiff sameDiff,
SDVariable labels,
SDVariable pred,
Integer numClasses) |
ConfusionMatrix(SameDiff sameDiff,
SDVariable labels,
SDVariable pred,
Integer numClasses,
SDVariable weights) |
ConfusionMatrix(SameDiff sameDiff,
SDVariable labels,
SDVariable pred,
SDVariable weights) |
Create(String name,
SameDiff sameDiff,
SDVariable input,
boolean initialize) |
Create(String name,
SameDiff sameDiff,
SDVariable input,
char order,
boolean initialize,
DataType dataType) |
Cross(SameDiff sameDiff,
SDVariable[] args) |
Diag(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
DiagPart(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
ExpandDims(SameDiff sameDiff,
SDVariable[] args) |
ExpandDims(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
ExpandDims(SameDiff sameDiff,
SDVariable[] args,
int axis) |
Eye(SameDiff sameDiff,
int numRows) |
Eye(SameDiff sameDiff,
int numRows,
int numCols) |
Eye(SameDiff sameDiff,
int numRows,
int numCols,
DataType dataType) |
Eye(SameDiff sameDiff,
int numRows,
int numCols,
DataType dataType,
int[] batchDimension) |
Eye(SameDiff sameDiff,
SDVariable numRows) |
Eye(SameDiff sameDiff,
SDVariable numRows,
SDVariable numCols) |
Eye(SameDiff sameDiff,
SDVariable numRows,
SDVariable numCols,
SDVariable batch_shape) |
Gather(SameDiff sameDiff,
SDVariable input,
int[] indices,
int axis,
boolean inPlace) |
Gather(SameDiff sameDiff,
SDVariable input,
SDVariable indices,
int axis,
boolean inPlace) |
GatherNd(SameDiff sameDiff,
SDVariable input,
SDVariable indices,
boolean inPlace) |
Linspace(SameDiff sameDiff,
SDVariable from,
SDVariable to,
SDVariable length,
DataType dataType) |
MergeAvg(SameDiff sameDiff,
SDVariable... inputs) |
MergeMax(SameDiff sameDiff,
SDVariable... inputs) |
MergeSum(SameDiff sameDiff,
SDVariable... inputs) |
MeshGrid(SameDiff sd,
boolean cartesian,
SDVariable... inputs) |
OneHot(SameDiff sameDiff,
SDVariable indices,
int depth) |
OneHot(SameDiff sameDiff,
SDVariable indices,
int depth,
int axis,
double on,
double off,
DataType dataType) |
OnesLike(String name,
SameDiff sameDiff,
SDVariable input) |
OnesLike(String name,
SameDiff sameDiff,
SDVariable input,
DataType dataType) |
ParallelStack(SameDiff sameDiff,
SDVariable[] values) |
Permute(SameDiff sameDiff,
SDVariable i_v,
int... permuteDims) |
Permute(SameDiff sd,
SDVariable input,
SDVariable permuteDims) |
Rank(SameDiff sameDiff,
SDVariable input,
boolean inPlace) |
ReductionShape(SameDiff sameDiff,
SDVariable shape,
SDVariable axis,
boolean keepDims) |
Repeat(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace,
int axis) |
Repeat(SameDiff sameDiff,
SDVariable[] args,
int axis) |
Reshape(SameDiff sameDiff,
SDVariable i_v,
long[] shape) |
Reshape(SameDiff sameDiff,
SDVariable i_v,
SDVariable shape) |
SequenceMask(SameDiff sameDiff,
SDVariable input,
DataType dataType) |
SequenceMask(SameDiff sameDiff,
SDVariable input,
int maxLen,
DataType dataType) |
SequenceMask(SameDiff sameDiff,
SDVariable input,
SDVariable maxLen,
DataType dataType) |
Shape(SameDiff sameDiff,
SDVariable input,
boolean inPlace) |
ShapeN(SameDiff sameDiff,
SDVariable[] inputs,
boolean inPlace) |
Size(SameDiff sameDiff,
SDVariable input) |
SizeAt(SameDiff sameDiff,
SDVariable input,
int dimension) |
Slice(SameDiff sameDiff,
SDVariable input,
int[] begin,
int[] size) |
Slice(SameDiff sameDiff,
SDVariable input,
SDVariable begin,
SDVariable end) |
Squeeze(SameDiff sameDiff,
SDVariable arg,
int[] squeezeDims) |
Stack(SameDiff sameDiff,
SDVariable[] values,
int axis) |
StridedSlice(SameDiff sameDiff,
SDVariable in,
int[] begin,
int[] end,
int[] strides) |
StridedSlice(SameDiff sameDiff,
SDVariable in,
int[] begin,
int[] end,
int[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
StridedSlice(SameDiff sameDiff,
SDVariable in,
long[] begin,
long[] end,
long[] strides) |
StridedSlice(SameDiff sameDiff,
SDVariable in,
long[] begin,
long[] end,
long[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
Tile(SameDiff sameDiff,
SDVariable i_v,
int[] axis) |
Tile(SameDiff sameDiff,
SDVariable i_v,
SDVariable axis) |
Transpose(SameDiff sameDiff,
SDVariable i_v) |
Transpose(SameDiff sameDiff,
SDVariable in,
int[] permuteDims) |
Transpose(SameDiff sameDiff,
SDVariable in,
SDVariable permuteDims) |
Unstack(SameDiff sameDiff,
SDVariable value,
int axis) |
Unstack(SameDiff sameDiff,
SDVariable value,
int axis,
int num) |
ZerosLike(String name,
SameDiff sameDiff,
SDVariable input) |
ZerosLike(String name,
SameDiff sameDiff,
SDVariable input,
boolean inPlace) |
Constructor and Description |
---|
ConcatBp(SameDiff sameDiff,
int concatDimension,
SDVariable... inputsAndGrad) |
ConcatBp(SameDiff sameDiff,
SDVariable... inputsGradAxis) |
SliceBp(SameDiff sameDiff,
SDVariable input,
SDVariable gradient,
int[] begin,
int[] size) |
SliceBp(SameDiff sameDiff,
SDVariable input,
SDVariable gradient,
SDVariable begin,
SDVariable size) |
StridedSliceBp(SameDiff sameDiff,
SDVariable in,
SDVariable grad,
long[] begin,
long[] end,
long[] strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
StridedSliceBp(SameDiff sameDiff,
SDVariable in,
SDVariable grad,
SDVariable begin,
SDVariable end,
SDVariable strides,
int beginMask,
int endMask,
int ellipsisMask,
int newAxisMask,
int shrinkAxisMask) |
TileBp(SameDiff sameDiff,
SDVariable in,
SDVariable grad,
int[] repeat) |
TileBp(SameDiff sameDiff,
SDVariable in,
SDVariable repeat,
SDVariable grad) |
Modifier and Type | Method and Description |
---|---|
SameDiff |
TensorArray.getSameDiff() |
Modifier and Type | Method and Description |
---|---|
void |
TensorArraySize.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorArrayRead.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorArraySplit.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorArrayConcat.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorArrayScatter.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorArrayGather.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
TensorArray.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
TensorArraySize.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
TensorArrayRead.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BaseTensorOp.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
StandardDeviation(SameDiff sameDiff,
SDVariable i_v,
boolean biasCorrected,
boolean keepDims,
int[] dimensions) |
Variance(SameDiff sameDiff,
SDVariable i_v,
boolean biasCorrected,
boolean keepDims,
int[] dimensions) |
Modifier and Type | Method and Description |
---|---|
void |
Cholesky.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Pad.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
BinCount.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
HistogramFixedWidth.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
CheckNumerics.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
NthElement.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
Angle(SameDiff sameDiff,
SDVariable input) |
BaseDynamicTransformOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
BinCount(SameDiff sd,
SDVariable in,
SDVariable weights,
Integer minLength,
Integer maxLength,
DataType outputType) |
CheckNumerics(SameDiff sd,
SDVariable input,
SDVariable message) |
HistogramFixedWidth(SameDiff sameDiff,
SDVariable values,
SDVariable valuesRange,
SDVariable numBins) |
IdentityN(SameDiff sameDiff,
SDVariable input) |
IdentityN(SameDiff sameDiff,
SDVariable[] inputs) |
MaxOut(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
Number max) |
MaxOut(SameDiff sameDiff,
SDVariable i_v,
Object[] extraArgs,
Number max) |
Pad(SameDiff sd,
SDVariable in,
SDVariable padding,
Pad.Mode mode,
double padValue) |
ReluLayer(SameDiff sameDiff,
SDVariable input,
SDVariable weights,
SDVariable bias) |
Constructor and Description |
---|
Assign(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
IsMax(SameDiff sameDiff,
SDVariable i_v) |
Constructor and Description |
---|
BooleanNot(SameDiff sameDiff,
SDVariable i_v) |
IsFinite(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
IsInf(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
IsNaN(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
MatchConditionTransform(SameDiff sameDiff,
SDVariable in,
Condition condition) |
Modifier and Type | Method and Description |
---|---|
void |
ClipByNorm.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
ClipByValue.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
ClipByNorm.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ClipByValue.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
ClipByNorm(SameDiff sameDiff,
SDVariable x,
double clipValue,
int... dimensions) |
ClipByNormBp(SameDiff sameDiff,
SDVariable x,
SDVariable eps,
double clipValue,
int... dimensions) |
ClipByValue(SameDiff sameDiff,
SDVariable x,
double clipValueMin,
double clipValueMax) |
ClipByValue(SameDiff sameDiff,
SDVariable x,
double clipValueMin,
double clipValueMax,
boolean inPlace) |
Constructor and Description |
---|
CompareAndReplace(SameDiff sameDiff,
SDVariable to,
SDVariable from,
Condition condition) |
CompareAndSet(SameDiff sameDiff,
SDVariable to,
Number set,
Condition condition) |
Eps(SameDiff sameDiff) |
Eps(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
Eps(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
Eps(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
Object[] extraArgs) |
Modifier and Type | Method and Description |
---|---|
void |
Fill.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
CumProd.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Assign.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
CumSum.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Dilation2D.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
MirrorPad.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DynamicPartition.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
UniqueWithCounts.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Fill.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ReverseV2.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
FakeQuantWithMinMaxArgs.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ParallelConcat.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Unique.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
ReverseSequence.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
InTopK.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
CumProd.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
TopK.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Assign.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
FakeQuantWithMinMaxVars.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
Svd.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
CumSum.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DynamicStitch.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
Assign(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
ATan2(SameDiff sameDiff,
SDVariable y,
SDVariable x) |
BatchToSpace(SameDiff sameDiff,
SDVariable[] args,
int[] blocks,
int[][] crops,
boolean inPlace) |
BatchToSpaceND(SameDiff sameDiff,
SDVariable[] args,
int[] blocks,
int[][] crops,
boolean inPlace) |
BitsHammingDistance(SameDiff sd,
SDVariable x,
SDVariable y) |
BitwiseAnd(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
BitwiseOr(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
BitwiseXor(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
Choose(SameDiff sameDiff,
SDVariable[] args,
Condition condition) |
Choose(String opName,
SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
CumProd(SameDiff sameDiff,
SDVariable x,
boolean exclusive,
boolean reverse,
int... axis) |
CumProd(SameDiff sameDiff,
SDVariable x,
int... axis) |
CumSum(SameDiff sameDiff,
SDVariable x,
boolean exclusive,
boolean reverse,
int... axis) |
CumSum(SameDiff sameDiff,
SDVariable x,
int... axis) |
CyclicRShiftBits(SameDiff sameDiff,
SDVariable x,
SDVariable shift) |
CyclicShiftBits(SameDiff sameDiff,
SDVariable x,
SDVariable shift) |
Dilation2D(SameDiff sameDiff,
SDVariable[] inputAndWeights,
int[] strides,
int[] rates,
boolean isSameMode,
boolean inPlace) |
DotProductAttention(SameDiff sameDiff,
SDVariable queries,
SDVariable keys,
SDVariable values,
SDVariable mask,
boolean scaled,
boolean withWeights) |
DotProductAttentionBp(SameDiff sameDiff,
SDVariable queries,
SDVariable keys,
SDVariable values,
SDVariable eps,
SDVariable mask,
boolean scaled) |
DynamicPartition(SameDiff sameDiff,
SDVariable input,
SDVariable partitions,
int numPartitions) |
DynamicStitch(SameDiff sameDiff,
SDVariable[] indices,
SDVariable[] inputs) |
EqualTo(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
FakeQuantWithMinMaxArgs(SameDiff sd,
SDVariable input,
float min,
float max,
boolean narrowRange,
int numBits) |
FakeQuantWithMinMaxVars(SameDiff sd,
SDVariable input,
SDVariable min,
SDVariable max,
boolean narrowRange,
int numBits) |
Fill(SameDiff sameDiff,
SDVariable shape,
DataType outputDataType,
double value) |
GreaterThan(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
GreaterThanOrEqual(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
InTopK(SameDiff sd,
SDVariable predictions,
SDVariable targets,
int k) |
InvertPermutation(SameDiff sameDiff,
SDVariable input,
boolean inPlace) |
IsNonDecreasing(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
IsNumericTensor(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
IsStrictlyIncreasing(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
LayerNorm(SameDiff sameDiff,
SDVariable input,
SDVariable gain,
boolean channelsFirst,
int... dimensions) |
LayerNorm(SameDiff sameDiff,
SDVariable input,
SDVariable gain,
SDVariable bias,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(SameDiff sameDiff,
SDVariable input,
SDVariable gain,
SDVariable gradient,
boolean channelsFirst,
int... dimensions) |
LayerNormBp(SameDiff sameDiff,
SDVariable input,
SDVariable gain,
SDVariable bias,
SDVariable gradient,
boolean channelsFirst,
int... dimensions) |
LessThan(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
LessThanOrEqual(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
ListDiff(SameDiff sd,
SDVariable x,
SDVariable y) |
LogicalAnd(SameDiff sd,
SDVariable in1,
SDVariable in2) |
LogicalNot(SameDiff sd,
SDVariable in1,
SDVariable in2) |
LogicalOr(SameDiff sd,
SDVariable in1,
SDVariable in2) |
LogicalXor(SameDiff sd,
SDVariable in1,
SDVariable in2) |
LogMatrixDeterminant(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
LogSoftMax(SameDiff sameDiff,
SDVariable i_v) |
LogSoftMax(SameDiff sameDiff,
SDVariable i_v,
int dimension) |
MatrixDeterminant(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
MatrixDiag(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
MatrixDiagPart(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
MatrixInverse(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
MatrixSetDiag(SameDiff sameDiff,
SDVariable in,
SDVariable diag,
boolean inPlace) |
Max(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
Max(SameDiff sameDiff,
SDVariable first,
SDVariable second) |
Min(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
Min(SameDiff sameDiff,
SDVariable first,
SDVariable second) |
MultiHeadDotProductAttention(SameDiff sameDiff,
SDVariable queries,
SDVariable keys,
SDVariable values,
SDVariable Wq,
SDVariable Wk,
SDVariable Wv,
SDVariable Wo,
SDVariable mask,
boolean scaled,
boolean withWeights) |
MultiHeadDotProductAttentionBp(SameDiff sameDiff,
SDVariable queries,
SDVariable keys,
SDVariable values,
SDVariable Wq,
SDVariable Wk,
SDVariable Wv,
SDVariable Wo,
SDVariable eps,
SDVariable mask,
boolean scaled) |
NotEqualTo(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
Pow(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
Reverse(SameDiff sameDiff,
SDVariable i_v,
int... dimensions) |
ReverseSequence(SameDiff sameDiff,
SDVariable i_v,
SDVariable seqLengths) |
ReverseSequence(SameDiff sameDiff,
SDVariable i_v,
SDVariable seqLengths,
int seqDim,
int batchDim) |
RShiftBits(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
ShiftBits(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
SoftMax(SameDiff sameDiff,
SDVariable[] args) |
SoftMax(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
SoftMax(SameDiff sameDiff,
SDVariable[] args,
int dimension) |
SoftMax(SameDiff sameDiff,
SDVariable[] args,
int dimension,
boolean inPlace) |
SpaceToBatch(SameDiff sameDiff,
SDVariable[] args,
int[] blocks,
int[][] padding,
boolean inPlace) |
SpaceToBatchND(SameDiff sameDiff,
SDVariable[] args,
int[] blocks,
int[][] padding,
boolean inPlace) |
Standardize(SameDiff sameDiff,
SDVariable i_v,
int... dimensions) |
StandardizeBp(SameDiff sameDiff,
SDVariable i_v,
SDVariable grad,
int... dimensions) |
Svd(SameDiff sd,
SDVariable input,
boolean fullUV,
boolean computeUv) |
Svd(SameDiff sd,
SDVariable input,
boolean fullUV,
boolean computeUv,
int switchNum) |
ThresholdRelu(SameDiff sd,
SDVariable input,
boolean inPlace,
double cutoff) |
ThresholdRelu(SameDiff sd,
SDVariable input,
double cutoff) |
TopK(SameDiff sd,
SDVariable in,
int k,
boolean sorted) |
Trace(SameDiff sd,
SDVariable in) |
Unique(SameDiff sd,
SDVariable in) |
UniqueWithCounts(SameDiff sd,
SDVariable in) |
XwPlusB(SameDiff sameDiff,
SDVariable input,
SDVariable weights,
SDVariable bias) |
Zeta(SameDiff sameDiff,
SDVariable x,
SDVariable q) |
Constructor and Description |
---|
SegmentMax(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds) |
SegmentMean(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds) |
SegmentMin(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds) |
SegmentProd(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds) |
SegmentSum(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds) |
Modifier and Type | Method and Description |
---|---|
void |
Cast.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
Cast(SameDiff sameDiff,
SDVariable arg,
DataType dst) |
Constructor and Description |
---|
RSqrt(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Sqrt(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Constructor and Description |
---|
BinaryMinimalRelativeError(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BinaryMinimalRelativeError(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BinaryMinimalRelativeError(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
BinaryRelativeError(SameDiff sameDiff) |
BinaryRelativeError(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
BinaryRelativeError(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
BinaryRelativeError(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
RelativeError(SameDiff sameDiff) |
RelativeError(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
RelativeError(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
RelativeError(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
Set(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Constructor and Description |
---|
AddOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
Axpy(SameDiff sameDiff,
double p) |
Axpy(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double p) |
Axpy(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace,
double p) |
Axpy(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
double p) |
CopyOp(SameDiff sameDiff) |
CopyOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
CopyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
CopyOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
DivOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
FloorDivOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
FloorDivOp(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
FloorModOp(SameDiff sameDiff,
SDVariable x,
SDVariable y) |
FModOp(SameDiff sameDiff) |
FModOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
FModOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
FModOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
MergeAddOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
ModOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
MulOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
PowPairwise(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
PowPairwise(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
RDivOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
RealDivOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
RemainderOp(SameDiff sameDiff) |
RemainderOp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
RemainderOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
RemainderOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
RSubOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
RSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
RSubOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
SquaredDifferenceOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
SubOp(SameDiff sameDiff,
SDVariable[] args,
boolean inPlace) |
TruncateDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
TruncateDivOp(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
Constructor and Description |
---|
AddBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
BaseArithmeticBackpropOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
DivBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
FloorDivBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
FloorModBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
ModBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
MulBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
RDivBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
RSubBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
SquaredDifferenceBpOp(SameDiff sameDiff,
SDVariable[] args) |
SubBpOp(SameDiff sameDiff,
SDVariable x,
SDVariable y,
SDVariable eps) |
Constructor and Description |
---|
And(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
And(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double comparable) |
And(SameDiff sameDiff,
SDVariable ix,
SDVariable iy) |
Not(SameDiff sameDiff,
SDVariable i_v) |
Or(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double comparable) |
Or(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
Or(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
Xor(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double comparable) |
Xor(SameDiff sameDiff,
SDVariable ix,
SDVariable iy) |
Constructor and Description |
---|
Abs(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
AMax(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2) |
AMin(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2) |
Ceil(SameDiff sameDiff,
SDVariable i_v) |
Ceil(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Cube(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Floor(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Identity(SameDiff sd,
SDVariable input) |
Max(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2) |
Min(SameDiff sameDiff,
SDVariable i_v,
SDVariable i_v2) |
Negative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
OneMinus(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Reciprocal(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
Round(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Sign(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Square(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
TimesOneMinus(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Constructor and Description |
---|
UnsortedSegmentMax(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds,
int numSegments) |
UnsortedSegmentMean(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds,
int numSegments) |
UnsortedSegmentMin(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds,
int numSegments) |
UnsortedSegmentProd(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds,
int numSegments) |
UnsortedSegmentSqrtN(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds,
int numSegments) |
UnsortedSegmentSum(SameDiff sameDiff,
SDVariable data,
SDVariable segmentIds,
int numSegments) |
Modifier and Type | Method and Description |
---|---|
void |
RectifiedTanh.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
RectifiedTanh.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
ACos(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
ACosh(SameDiff sameDiff,
SDVariable i_v) |
ACosh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
ASin(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
ASinh(SameDiff sameDiff,
SDVariable i_v) |
ASinh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
ATan(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
ATanh(SameDiff sameDiff,
SDVariable i_v) |
ATanh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Cos(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Cosh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
ELU(SameDiff sameDiff,
SDVariable i_v) |
Erf(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Erfc(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Exp(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Expm1(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
GELU(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
boolean precise) |
GELUDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
HardSigmoid(SameDiff sameDiff,
SDVariable in,
boolean inPlace) |
HardTanh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Log(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Log1p(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
LogSigmoid(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Mish(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
MishDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
MishDerivative(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
MishDerivative(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
PreciseGELU(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
boolean precise) |
PreciseGELUDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
boolean precise) |
RationalTanh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
RectifiedTanh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Rint(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SELU(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SetRange(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double min,
double max) |
Sigmoid(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SigmoidDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace)
Deprecated.
|
SigmoidDerivative(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2)
Deprecated.
|
SigmoidDerivative(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace)
Deprecated.
|
Sin(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Sinh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SoftPlus(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SoftSign(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Stabilize(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace,
double realMin,
double cutOff,
double k) |
Swish(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SwishDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
SwishDerivative(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2) |
SwishDerivative(SameDiff sameDiff,
SDVariable i_v1,
SDVariable i_v2,
boolean inPlace) |
Tan(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
Tanh(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace) |
TanhDerivative(SameDiff sameDiff,
SDVariable i_v,
boolean inPlace)
Deprecated.
|
Modifier and Type | Method and Description |
---|---|
void |
SaveV2.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
RestoreV2.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
BaseRandomOp(SameDiff sd,
long[] shape) |
BaseRandomOp(SameDiff sameDiff,
SDVariable i_v) |
Constructor and Description |
---|
RandomStandardNormal(SameDiff sameDiff,
SDVariable[] args) |
Modifier and Type | Method and Description |
---|---|
void |
DistributionUniform.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
DistributionUniform(SameDiff sd,
SDVariable shape,
double min,
double max) |
DistributionUniform(SameDiff sd,
SDVariable shape,
double min,
double max,
DataType dataType) |
RandomBernoulli(SameDiff sd,
SDVariable shape,
double p) |
RandomExponential(SameDiff sd,
SDVariable shape,
double lambda) |
RandomNormal(SameDiff sameDiff,
SDVariable shape,
double mean,
double stdev) |
Modifier and Type | Method and Description |
---|---|
void |
DropOutInverted.initFromOnnx(Onnx.NodeProto node,
SameDiff initWith,
Map<String,Onnx.AttributeProto> attributesForNode,
Onnx.GraphProto graph) |
void |
Range.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
void |
DropOutInverted.initFromTensorFlow(NodeDef nodeDef,
SameDiff initWith,
Map<String,AttrValue> attributesForNode,
GraphDef graph) |
Constructor and Description |
---|
BernoulliDistribution(SameDiff sd,
double prob,
long[] shape) |
BinomialDistribution(SameDiff sd,
int trials,
double probability,
long[] shape) |
DropOut(SameDiff sameDiff,
SDVariable input,
double p) |
DropOutInverted(SameDiff sameDiff,
SDVariable input,
double p) |
GaussianDistribution(SameDiff sd,
double mean,
double stddev,
long[] shape) |
Linspace(SameDiff sd,
double from,
double to,
long length) |
LogNormalDistribution(SameDiff sd,
double mean,
double stdev,
long... shape) |
Range(SameDiff sd,
double from,
double to,
double step,
DataType dataType) |
Range(SameDiff sd,
SDVariable from,
SDVariable to,
SDVariable step,
DataType dataType) |
TruncatedNormalDistribution(SameDiff sd,
double mean,
double stddev,
long[] shape) |
UniformDistribution(SameDiff sd,
double from,
double to,
long[] shape) |
Copyright © 2019. All rights reserved.