Update

README.md

@@ -181,7 +181,7 @@ To run inference on Renesas R-Car X5H, you need:
 ## Download
 
 ```bash
-huggingface-cli download Renesas/RetinaNet-ONNX fp32/retinanet-9.onnx
+hf download Renesas/RetinaNet-R101-ONNX --repo-type=model --include "fp32/*"
 ```
 
 ## Inference

compile_config/RetinaNet_l2seg.json

@@ -0,0 +1,312 @@
+{
+  "segments": [
+    {
+      "name": "Segment_MP",
+      "inputs": [
+        "input"
+      ],
+      "outputs": [
+        "740"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_1",
+      "inputs": [
+        "740"
+      ],
+      "outputs": [
+        "772"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_merge_2_3_4",
+      "inputs": [
+        "772"
+      ],
+      "outputs": [
+        "932"
+      ],
+      "n_buffer": 1,
+      "tiling_info": {
+        "932": {
+          "l2_tile_shape": [
+            1,
+            30,
+            40,
+            1024
+          ]
+        }
+      }
+    },
+    {
+      "name": "Segment_6",
+      "inputs": [
+        "932"
+      ],
+      "outputs": [
+        "972"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_7_8_9",
+      "inputs": [
+        "972"
+      ],
+      "outputs": [
+        "1004"
+      ],
+      "n_buffer": 1,
+      "tiling_info": {
+        "1004": {
+          "l2_tile_shape": [
+            1,
+            15,
+            20,
+            2048
+          ]
+        }
+      }
+    },
+    {
+      "name": "Segment_10",
+      "inputs": [
+        "1004"
+      ],
+      "outputs": [
+        "1005"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_11",
+      "inputs": [
+        "1004"
+      ],
+      "outputs": [
+        "1028"
+      ],
+      "n_buffer": 1,
+      "tiling_info": {
+        "1028": {
+          "l2_tile_shape": [
+            1,
+            8,
+            10,
+            256
+          ]
+        }
+      }
+    },
+    {
+      "name": "Segment_12_merge_resize_1",
+      "inputs": [
+        "740",
+        "1016"
+      ],
+      "outputs": [
+        "1031"
+      ],
+      "n_buffer": 1,
+      "tiling_info": {
+        "1031": {
+          "l2_tile_shape": [
+            1,
+            60,
+            80,
+            256
+          ]
+        }
+      }
+    },
+    {
+      "name": "Segment_output1.0",
+      "inputs": [
+        "1031"
+      ],
+      "outputs": [
+        "output1"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output2.0",
+      "inputs": [
+        "1032"
+      ],
+      "outputs": [
+        "output2"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output3.0",
+      "inputs": [
+        "1033"
+      ],
+      "outputs": [
+        "output3"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output4.0",
+      "inputs": [
+        "1028"
+      ],
+      "outputs": [
+        "output4"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output5.0",
+      "inputs": [
+        "1030"
+      ],
+      "outputs": [
+        "output5"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output6",
+      "inputs": [
+        "1031"
+      ],
+      "outputs": [
+        "output6"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output7",
+      "inputs": [
+        "1032"
+      ],
+      "outputs": [
+        "output7"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output8",
+      "inputs": [
+        "1033"
+      ],
+      "outputs": [
+        "output8"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output9",
+      "inputs": [
+        "1028"
+      ],
+      "outputs": [
+        "output9"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "Segment_output10",
+      "inputs": [
+        "1030"
+      ],
+      "outputs": [
+        "output10"
+      ],
+      "n_buffer": 1
+    },
+    {
+      "name": "seg_33_merge_resize_0",
+      "inputs": [
+        "972",
+        "1005"
+      ],
+      "outputs": [
+        "1016"
+      ],
+      "n_buffer": 1,
+      "tiling_info": {
+        "1016": {
+          "l2_tile_shape": [
+            1,
+            30,
+            40,
+            256
+          ]
+        }
+      }
+    }
+  ],
+
+  "layer_groups": [
+    {
+      "name": "LayerGroup_0",
+      "inputs": [
+        "666"
+      ],
+      "outputs": [
+        "673"
+      ]
+    },
+    {
+      "name": "LayerGroup_1",
+      "inputs": [
+        "678"
+      ],
+      "outputs": [
+        "688"
+      ]
+    },
+    {
+      "name": "LayerGroup_2",
+      "inputs": [
+        "688"
+      ],
+      "outputs": [
+        "698"
+      ]
+    },
+    {
+      "name": "LayerGroup_3",
+      "inputs": [
+        "698"
+      ],
+      "outputs": [
+        "705"
+      ]
+    },
+    {
+      "name": "LayerGroup_4",
+      "inputs": [
+        "710"
+      ],
+      "outputs": [
+        "720"
+      ]
+    },
+    {
+      "name": "LayerGroup_5",
+      "inputs": [
+        "720"
+      ],
+      "outputs": [
+        "730"
+      ]
+    },
+    {
+      "name": "LayerGroup_6",
+      "inputs": [
+        "730"
+      ],
+      "outputs": [
+        "740"
+      ]
+    }
+  ]
+}

int8/benchmarks/x5h_mwmx_npu_apm80_12core.yaml

@@ -40,3 +40,42 @@ memory:
 
 power:
   avg_w: null
+
+# Exact commands verified against the NNAC "Getting Started" chapter, plus the
+# network_configuration block verified against the internal model_zoo compile
+# config for this model's 12-core slice (see
+# compile_config/RetinaNet_l2seg.json). Rendered by the AI-Dashboard in place
+# of the generic placeholder flow — see downloadRunHTML() / parse_reproduce().
+reproduce:
+  steps:
+    - title: Download the ONNX model and compile config
+      command: hf download Renesas/RetinaNet-R101-ONNX --repo-type=model --include "fp32/*" --include "compile_config/*"
+    - title: Configure the network for the 12-core build
+      command: |
+        cat >> ${WORKDIR}/nnac_config/config.yaml << 'EOF'
+        network_configuration:
+          retinanet_9:
+            optimization_options:
+              - -segmentation_json ./compile_config/RetinaNet_l2seg.json
+              - -enable_dma_fusion
+              - -enable_stu_multi_channel
+              - -stu_multi_channel_config 0x15212420
+              - -stu_weight
+              - -distributed_coef_stu
+        EOF
+        # If config.yaml already has a network_configuration: key (e.g. from a
+        # previous model), merge this block into it instead of appending a duplicate key.
+    - title: Compile with the NNAC toolchain (INT8 auto-cast from the FP32 graph)
+      command: |
+        python3 nnac_frontend/legalize.py -d binary/nnx ./fp32/retinanet-9.onnx
+    - title: Set up the R-Car X5H board
+      command: Configure the board per the AI Compiler (NNAC) "Getting Started" guide, section 3.4 (host TFTP/NFS setup, bootloader flashing, U-Boot, Linux boot, login) -- exact steps depend on your board/network setup.
+    - title: Copy the compiled artifact to the board
+      command: Copy ${WORKDIR}/binary/nnx to the board -- method may vary (NFS mount, scp, USB, etc.).
+    - title: Run on R-Car X5H (single NPU cluster, 12 AI cores)
+      command: |
+        cd binary
+        ./host_app ./arc_prog_npus ./nnx/retinanet-9
+  notes: >-
+    hash[n] = 0x...(OK) means the run's output matches the reference hash in
+    hash.txt; latency is the NPX execution time reported in cycles and ms.