def optimize_graph_step(graph: Graph, config: Config) -> None:
"""Optimizing graph that imported from tensorflow pb.
Args:
graph (Graph): Graph that optimization passes are applying to
config (Config): Collection of configurations
Returns:
"""
pass_remove_identities(graph)
pass_transpose(graph)
if config.activate_hard_quantization:
pass_lookup(graph)
pass_propagate_quantization_details_into_conv(graph)
if config.threshold_skipping:
pass_compute_thresholds(graph)
pass_pack_weights(graph)
pass_quantize_convolutions(graph)
if config.threshold_skipping:
pass_propagate_output_type_backward(graph)
pass_propagate_datatypes(graph)
pass_propagate_format(graph)
pass_constant_folding(graph)
pass_simplify_batchnorm(graph)
def optimize_graph_step(model: Model, config: Config) -> None:
"""Optimize graph in the model.
Parameters
----------
model : Model
Model that contains the graph
config : Config
Collection of configurations
"""
graph: Graph = model.graph
pass_remove_identities(graph)
pass_transpose(graph)
if config.activate_hard_quantization:
pass_lookup(graph)
pass_propagate_quantization_details_into_conv(graph)
if config.threshold_skipping:
pass_compute_thresholds(graph)
pass_pack_weights(graph)
pass_quantize_convolutions(graph)
pass_fix_qtz_types_and_format(graph)
pass_propagate_output_type_backward(graph)
pass_propagate_datatypes(graph)
pass_propagate_format(graph)
pass_constant_folding(graph)
def test_pass_pack_weights(self) -> None:
"""Test pass."""
data1 = np.float32(np.random.rand(1, 2, 2, 3))
data2 = np.float32(np.random.rand(1, 2, 2, 3))
graph1 = self.create_sample_graph(data1, data2)
pass_pack_weights(graph1)
self.assertEqual(graph1.get_op('conv2').input_ops['W'].op_type, 'Constant',
'[Failed] Found input kernel weights not a constant')
graph_2_1 = self.create_sample_graph_2(data1)
graph_2_2 = self.create_sample_graph_2(data1)
pass_pack_weights(graph_2_2)
self.assertEqual(graph_2_1, graph_2_2,
'[Failed] Found optimized graph not the same')
print("Test pass #4 pack_weights passed!")