afe.ir.node

Classes

AwesomeNode

An awesome node is responsible for storing attributes and operations used by the AwesomeNet when

Functions

`node_is_sima_ir`(→ bool)	Return True is the AwesomeNode contains a SiMa IR
`node_is_backend_ir`(→ bool)	Return True is the AwesomeNode contains a SiMa IR
`node_is_constant`(→ bool)	Return True is the AwesomeNode contains a Constant SiMaIR
`node_is_placeholder`(→ bool)	Return True is the AwesomeNode contains a Placeholder SiMaIR
`node_is_awesomenet`(→ bool)	Return True is the AwesomeNode contains a AwesomeNet
`node_is_external`(→ bool)	Return True is the AwesomeNode contains a External Graph
`node_is_ev`(→ bool)	Return True if the AwesomeNode is an EV node.
`node_is_relu`(→ bool)
`node_is_mean`(→ bool)
`node_is_upsampling`(→ bool)
`node_is_subgraph`(→ bool)	Return True is the AwesomeNode contains a sub-graph.
`node_is_tuple`(→ bool)	Return True is the AwesomeNode contains a Tuple.
`node_is_tuple_get_item`(→ bool)	Return True is the AwesomeNode contains a TupleGetItem.
`node_is_unpack`(→ bool)	Return True is the AwesomeNode contains an Unpack transform.
`node_is_fp32_node`(→ bool)	Return True if AwesomeNode's output type is floating point.
`node_is_pool_node`(→ bool)	Return True if the AwesomeNode contains Pool operator.
`node_is_requantization_node`(→ bool)	Return True if the AwesomeNode contains Requantize operator.
`node_uses_observer`(→ bool)	Return True if node uses calibration information during quantization.
`get_node_inputs`(→ Dict[afe.ir.defines.NodeName, Any])	Generates a dictionary using the keys from a node's input_names and the values corresponding

Module Contents

class afe.ir.node.AwesomeNode[source]

An awesome node is responsible for storing attributes and operations used by the AwesomeNet when executing the network

Parameters:

name – Name of the node
input_names – Parameter names of this AwesomeNode. These represent the inputs expected by the code inside this node.
input_node_names – Argument names of this AwesomeNode. These represent the inputs supplied by the net that contains this node. Data for each argument gets forwarded to the corresponding parameter.
ir – Union[SiMaIR, “afe.ir.net.AwesomeNet”, “afe.backends.BackendIR”]. The IR can be: * SiMaIR - A frontend IR that contains single operation or ExternalOp * AwesomeNet - A frontend IR that contains a sub-graph that can be supported by SiMa’s backends * BackendIR - A backend IR that can be compiled, or already compiled by certain backend
_status – Status. Node status. Default is Status.RELAY
_layer_stats – Layer statistics. For each MLA node, quantization error is calculated, that information is than forwarded to .sima.json file, and it can be viewed in Netron.

name: afe.ir.defines.NodeName[source]

input_names: List[afe.ir.defines.InputName][source]

input_node_names: List[afe.ir.defines.NodeName][source]

ir: afe.ir.sima_ir.SiMaIR | afe.ir.net.AwesomeNet | afe.backends.BackendIR[source]

property status: afe.ir.defines.Status[source]

get_type() → afe.ir.tensor_type.NodeType[source]

set_batch_size(batch_size: int)[source]

Modifies AwesomeNode’s internal parameters to accommodate for a given batch size.

Parameters:: batch_size – Integer value representing the batch size of the inputs to the AwesomeNet.

update_layer_stats(layer_stats: afe.ir.defines.LayerStats)[source]

property layer_stats: afe.ir.defines.LayerStats[source]

afe.ir.node.node_is_sima_ir(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a SiMa IR

Parameters:: node – AwesomeNode

afe.ir.node.node_is_backend_ir(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a SiMa IR

Parameters:: node – AwesomeNode

afe.ir.node.node_is_constant(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a Constant SiMaIR

Parameters:: node – AwesomeNode

afe.ir.node.node_is_placeholder(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a Placeholder SiMaIR

Parameters:: node – AwesomeNode

afe.ir.node.node_is_awesomenet(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a AwesomeNet

Parameters:: node – AwesomeNode

afe.ir.node.node_is_external(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a External Graph

Parameters:: node – AwesomeNode

afe.ir.node.node_is_ev(node: AwesomeNode) → bool[source]

Return True if the AwesomeNode is an EV node.

Parameters:: node – AwesomeNode

afe.ir.node.node_is_relu(node: AwesomeNode) → bool[source]

afe.ir.node.node_is_mean(node: AwesomeNode) → bool[source]

afe.ir.node.node_is_upsampling(node: AwesomeNode) → bool[source]

afe.ir.node.node_is_subgraph(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a sub-graph.

Parameters

param node:: AwesomeNode

Return

return:: bool

afe.ir.node.node_is_tuple(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a Tuple.

Parameters

param node:: AwesomeNode

Return

return:: bool

afe.ir.node.node_is_tuple_get_item(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains a TupleGetItem.

Parameters

param node:: AwesomeNode

Return

return:: bool

afe.ir.node.node_is_unpack(node: AwesomeNode) → bool[source]

Return True is the AwesomeNode contains an Unpack transform.

Parameters

param node:: AwesomeNode

Return

return:: bool

afe.ir.node.node_is_fp32_node(node: AwesomeNode) → bool[source]

Return True if AwesomeNode’s output type is floating point.

Parameters:: node – AwesomeNode to be analyzed.
Returns:: bool. True if AwesomeNode’s output type is fp32, otherwise False.

afe.ir.node.node_is_pool_node(node: AwesomeNode) → bool[source]

Return True if the AwesomeNode contains Pool operator.

Parameters:: node – AwesomeNode to be analyzed.
Returns:: bool. True if the AwesomeNode’s operator is Pooling operator (MaxPool, AvgPool).

afe.ir.node.node_is_requantization_node(node: AwesomeNode) → bool[source]

Return True if the AwesomeNode contains Requantize operator.

Parameters:: node – AwesomeNode to be analyzed.
Returns:: bool. True if the AwesomeNode’s operator is Requantize operator.

afe.ir.node.node_uses_observer(node: AwesomeNode) → bool[source]

Return True if node uses calibration information during quantization.

Parameters:: node – AwesomeNode to be analyzed.
Returns:: bool. True if AwesomeNode uses calibration information during quantization. Otherwise, quantization information is obtained using quantization from its input node(s).

afe.ir.node.get_node_inputs(node: AwesomeNode, node_outputs: Dict[afe.ir.defines.NodeName, Any]) → Dict[afe.ir.defines.NodeName, Any][source]

Generates a dictionary using the keys from a node’s input_names and the values corresponding to the data of nodes whose names are in the node’s input_node_names.

Parameters:

node – The AwesomeNode we are gathering inputs for.
node_outputs – Dictionary containing outputs from other nodes from across the network.

Returns:

A dictionary containing node inputs.