afe
The API for AFE as a library.
Functions in this file are re-exported from other modules. See those modules for documentation.
Submodules
Attributes
Functions
|
Loads an AwesomeNet from yaml and npz files |
|
Saves an AwesomeNet to a yaml and npz file. Overwrites existing files. |
|
Loads a TensorFlow model into an _AwesomeNet format |
|
Loads a PyTorch model into an _AwesomeNet format. |
|
Loads a TensorFlow Lite model into an _AwesomeNet format. |
|
Loads an ONNX model into an _AwesomeNet format. |
|
Loads a Keras model into an _AwesomeNet format. |
|
Loads a caffe model into an _AwesomeNet format. |
|
Loads a caffe2 model into an _AwesomeNet format. |
|
Quantizes a network. |
|
Executes a network. |
|
Compile a network using Product Compiler. |
|
Wrapper to take elf file and generate summary KPI file, and trace file. |
Package Contents
- afe.load_awesomenet(model_name: str, network_directory: str = '') afe.ir.net.AwesomeNet[source]
Loads an AwesomeNet from yaml and npz files
- Parameters:
model_name – The name of the model, used in file names.
network_directory – The directory to load the files from.
- afe.save_awesomenet(net: afe.ir.net.AwesomeNet, model_name: str, output_directory: str = '') None[source]
Saves an AwesomeNet to a yaml and npz file. Overwrites existing files.
- Parameters:
net – AwesomeNet to save.
model_name – The name of the model, used in file names.
output_directory – The directory to save the files to.
- afe.load_tensorflow_model(model_path: str, shape_dict: Dict[str, Tuple[int, Ellipsis]], output_names: List[str], layout: str = 'NHWC', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads a TensorFlow model into an _AwesomeNet format
- Parameters:
model_path – Path to a .pb TensorFlow model
shape_dict – Dictionary of input names to input shapes eg. {‘input’, (1,224,224,3)}
output_names – List of output names of the network eg. [‘output1’, ‘output2’]
layout – any variation of the characters NHWC representing Batch Size, Height, Width, and Channels
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.load_pytorch_model(model_path: str, input_names: List[str], input_shapes: List[Tuple[int, Ellipsis]], input_dtypes: List[afe.ir.tensor_type.ScalarType] | None = None, layout: str = 'NCHW', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads a PyTorch model into an _AwesomeNet format.
- Parameters:
model_path – Path to a PyTorch file (.pt) that contains the entire model
input_names – List of input names. eg [‘input’]
input_shapes – List of input shapes corresponding to input names. eg. [(1, 224, 224, 3)]
input_dtypes – List of input datatypes corresponding to input names. eg [‘float32’]
layout – Input data layout. Default is ‘NCHW’
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.load_tflite_model(model_path: str, shape_dict: Dict[str, Tuple[int, Ellipsis]], dtype_dict: Dict[str, afe.ir.tensor_type.ScalarType], layout: str = 'NHWC', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads a TensorFlow Lite model into an _AwesomeNet format.
- Parameters:
model_path – Path to a .tflite file containing the TensorFlow Lite model
shape_dict – Dictionary of input names to input shapes eg. {‘input’, (1,224,224,3)}
dtype_dict – Dictionary of input names to input types eg. {‘input’, ‘float32’}
layout – Input data layout. Default is ‘NHWC’
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.load_onnx_model(model_path: str, shape_dict: Dict[str, Tuple[int, Ellipsis]], dtype_dict: Dict[str, afe.ir.tensor_type.ScalarType], layout: str = 'NCHW', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads an ONNX model into an _AwesomeNet format.
- Parameters:
model_path – Path to a .onnx file containing the onnx model
shape_dict – Dictionary of input names to input shapes eg. {‘input’, (1,224,224,3)}
dtype_dict – Dictionary of input names to input types eg. {‘input’, ‘float32’}
layout – Input data layout. Default is ‘NCHW’
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.load_keras_model(model_path: str, shape_dict: Dict[str, Tuple[int, Ellipsis]], layout: str = 'NCHW', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads a Keras model into an _AwesomeNet format.
- Parameters:
model_path – Path to a .h5 file containing the keras model
shape_dict – Dictionary of input names to input shapes eg. {‘input’, (1,224,224,3)}
layout – Input data layout. Default is ‘NCHW’
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.load_caffe_model(prototxt_file_path: str, caffemodel_file_path: str, shape_dict: Dict[str, Tuple[int, Ellipsis]], dtype_dict: Dict[str, afe.ir.tensor_type.ScalarType], layout: str = 'NCHW', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads a caffe model into an _AwesomeNet format.
- Parameters:
prototxt_file_path – filepath to the caffe .prototxt file
caffemodel_file_path – filepath to the caffe .caffemodel file
shape_dict – Dictionary of input names to input shapes eg. {‘input’, (1,224,224,3)}
dtype_dict – Dictionary of input names to input types eg. {‘input’, ‘float32’}
layout – Input data layout. Default is ‘NCHW’
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.load_caffe2_model(init_net_file_path: str, predict_net_file_path: str, shape_dict: Dict[str, Tuple[int, Ellipsis]], dtype_dict: Dict[str, afe.ir.tensor_type.ScalarType], layout: str = 'NCHW', is_quantized: bool = False) afe.ir.net.AwesomeNet[source]
Loads a caffe2 model into an _AwesomeNet format.
- Parameters:
init_net_file_path – filepath to the caffe2 .pb init_net file
predict_net_file_path – filepath to the caffe2 .pb predict_net file
shape_dict – Dictionary of input names to input shapes eg. {‘input’, (1,224,224,3)}
dtype_dict – Dictionary of input names to input types eg. {‘input’, ‘float32’}
layout – Input data layout. Default is ‘NCHW’
is_quantized – Whether the model is pre-quantized. Default is False
- Returns:
An _AwesomeNet
- afe.quantize_net(net: afe.ir.net.AwesomeNet, input_generator: sima_utils.data.data_generator.DataGenerator, asymmetry: bool, per_channel: bool, max_calibration_samples: int)[source]
Quantizes a network.
- Parameters:
net – an _AwesomeNet
input_generator – a DataGenerator we use to feed inputs into the _AwesomeNet during calibration
asymmetry – If True this function performs asymmetric quantization. Otherwise, it performs symmetric quantization
per_channel – If True this function performs per_channel quantization.
max_calibration_samples – Maximum number of samples we use for calibration
- afe.execute_net(net: afe.ir.net.AwesomeNet, inputs: Dict[str, numpy.ndarray], dequantize_outputs: bool = True) List[numpy.ndarray][source]
Executes a network.
- Parameters:
net – an _AwesomeNet
inputs – Dictionary of placeholder node names (str) to the input data
dequantize_outputs – Whether to dequantize the output data to floating point
:return The result of executing the output node
- afe.compile_net(net: afe.ir.net.AwesomeNet, output_elf_path: str, compress: bool = True)[source]
Compile a network using Product Compiler.
- Parameters:
net – an _AwesomeNet.
output_elf_path – Path in which elf files should be created.
compress – If True mlc file is compressed before generating .elf file.
- afe.simulate_isim_module(elf_file: str, output_path: str, output_kpi_file_name: str | None = None, output_trace_file_name: str | None = None, generate_csv_kpi: bool | None = True, override_existing: bool | None = False, target: sima_utils.common.Platform | None = None)[source]
Wrapper to take elf file and generate summary KPI file, and trace file.
- Parameters:
elf_file – Path to the location of the elf file to simulate in ISIM.
output_path – Location of where to store all output files. If location does not exist, it will be created.
output_kpi_file_name – Name for output kpi file. Default: elf_file name without ‘.elf’ extension and finishing with ‘_kpis.txt’ postfix.
output_trace_file_name – Name for output trace file. Default: elf_file name without ‘.elf’ extension and finishing with ‘_trace.mpack’ postfix.
generate_csv_kpi – Whether to generate a CSV file with the KPIs in addition to the txt.
override_existing – Whether to override existing kpis and trace files from a previous run.
target – [Optional] platform target.