archai.algos.gumbelsoftmax package

Submodules

archai.algos.gumbelsoftmax.gs_arch_trainer module

class archai.algos.gumbelsoftmax.gs_arch_trainer.GsArchTrainer(conf_train: archai.common.config.Config, model: torch.nn.modules.module.Module, checkpoint: Optional[archai.common.checkpoint.CheckPoint])[source]

Bases: archai.nas.arch_trainer.ArchTrainer

create_optimizer(conf_optim: archai.common.config.Config, params) → torch.optim.optimizer.Optimizer[source]
pre_step(x: torch.Tensor, y: torch.Tensor) → None[source]

archai.algos.gumbelsoftmax.gs_exp_runner module

class archai.algos.gumbelsoftmax.gs_exp_runner.GsExperimentRunner(config_filename: str, base_name: str, clean_expdir=False)[source]

Bases: archai.nas.exp_runner.ExperimentRunner

finalizers()archai.nas.finalizers.Finalizers[source]
model_desc_builder()archai.algos.gumbelsoftmax.gs_model_desc_builder.GsModelDescBuilder[source]
trainer_class() → Optional[Type[archai.nas.arch_trainer.ArchTrainer]][source]

archai.algos.gumbelsoftmax.gs_finalizers module

class archai.algos.gumbelsoftmax.gs_finalizers.GsFinalizers[source]

Bases: archai.nas.finalizers.Finalizers

finalize_node(node: torch.nn.modules.container.ModuleList, node_index: int, node_desc: archai.nas.model_desc.NodeDesc, max_final_edges: int, *args, **kwargs)archai.nas.model_desc.NodeDesc[source]

archai.algos.gumbelsoftmax.gs_model_desc_builder module

class archai.algos.gumbelsoftmax.gs_model_desc_builder.GsModelDescBuilder[source]

Bases: archai.nas.model_desc_builder.ModelDescBuilder

build_nodes(stem_shapes: List[List[int]], conf_cell: archai.common.config.Config, cell_index: int, cell_type: archai.nas.model_desc.CellType, node_count: int, in_shape: List[int], out_shape: List[int]) → Tuple[List[List[int]], List[archai.nas.model_desc.NodeDesc]][source]
pre_build(conf_model_desc: archai.common.config.Config) → None[source]

archai.algos.gumbelsoftmax.gs_op module

class archai.algos.gumbelsoftmax.gs_op.GsOp(op_desc: archai.nas.model_desc.OpDesc, arch_params: Optional[archai.nas.arch_params.ArchParams], affine: bool)[source]

Bases: archai.nas.operations.Op

The output of GsOp is weighted output of all allowed primitives.

PRIMITIVES = ['max_pool_3x3', 'avg_pool_3x3', 'skip_connect', 'sep_conv_3x3', 'sep_conv_5x5', 'dil_conv_3x3', 'dil_conv_5x5', 'none']
can_drop_path() → bool[source]
finalize(sampled_weights) → Tuple[archai.nas.model_desc.OpDesc, Optional[float]][source]

for trainable op, return final op and its rank

forward(x)[source]

Defines the computation performed at every call.

Should be overridden by all subclasses.

Note

Although the recipe for forward pass needs to be defined within this function, one should call the Module instance afterwards instead of this since the former takes care of running the registered hooks while the latter silently ignores them.

ops() → Iterator[Tuple[archai.nas.operations.Op, float]][source]

Return contituent ops, if this op is primitive just return self

set_op_sampled_weights(sampled_weights: torch.Tensor)[source]

Sets the weight for each op

Module contents