TorchLean API

NN.API.Neural.Summary

Model Summaries #

Structured model-summary rendering for checked sequential models.

Instances For

    Dimensions of a tensor shape, outermost first. Scalars have no dimensions.

    Instances For

      User-facing tensor shape display for one model-summary shape.

      Instances For

        User-facing display for a list of parameter tensor shapes.

        Instances For

          Structured per-layer summary derived from a checked sequential model.

          • index :

            Zero-based position in the sequential layer list.

          • kind : String

            User-facing layer kind string.

          • inputShape : Shape

            Checked input shape for this layer.

          • outputShape : Shape

            Checked output shape for this layer.

          • paramShapes : List Shape

            Parameter tensor shapes owned by this layer.

          • paramCount :

            Total scalar parameter count for this layer.

          Instances For

            One-line rendering of one layer summary.

            Instances For

              Structured whole-model summary derived from a checked sequential model.

              • inputShape : Shape

                Checked input shape for the full model.

              • outputShape : Shape

                Checked output shape for the full model.

              • Per-layer summaries in order.

              • layerCount :

                Total number of layers in the sequential model.

              • totalParams :

                Total scalar parameter count across all layers.

              Instances For

                Header line for the model summary.

                Instances For

                  Multi-line rendering of the structured model summary.

                  Instances For

                    Recursive worker that records one summary row for each layer in a sequential model.

                    Instances For
                      def TorchLean.nn.summary {σ τ : Shape} (model : Sequential σ τ) :

                      Structured checked model summary for a sequential model.

                      Instances For
                        def TorchLean.nn.info {σ τ : Shape} (model : Sequential σ τ) :

                        Model description derived from a sequential model value.

                        This walks the checked Seq itself, so the printed layer list stays attached to the model that will actually run.

                        Instances For
                          def TorchLean.nn.printInfo {σ τ : Shape} (model : Sequential σ τ) :

                          Print the checked model description for a sequential model.

                          Instances For