diff --git a/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormBackward/Default/sweep.json b/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormBackward/Default/sweep.json index e92ee7ff0c8e..c6a6cbc5ab34 100644 --- a/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormBackward/Default/sweep.json +++ b/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormBackward/Default/sweep.json @@ -2198,149 +2198,6 @@ "seed": 1, "reference_source": "c++ integration suite: Smoke/IntegrationGpuBatchnormBackwardCalcStats3dFp32.Correctness_2" } - }, - { - "id": "runtime_epsilon", - "values": { - "io_data_type": "bfloat16", - "tensors": [ - { - "uid": 0, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 1, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 2, - "dims": [ - 2, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 3, - "dims": [ - 2, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "bfloat16" - }, - { - "uid": 4, - "dims": [ - 2, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "bfloat16" - }, - { - "uid": 5, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 6, - "dims": [ - 1 - ], - "strides": [ - 1 - ], - "data_type": "double" - }, - { - "uid": 7, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - } - ] - }, - "tensor_patches": [ - { - "uid": 6, - "set": { - "is_runtime_pass_by_value": true - }, - "remove": [ - "value_type", - "value" - ] - } - ], - "metadata": { - "format_version": 1, - "operation": "BatchnormPointwise", - "generator": "manual", - "generator_version": "1.0.0", - "generated_at": "2026-07-27T00:00:00Z", - "seed": 42, - "notes": "Runtime PBV scalars: epsilon (uid=6) promoted to is_runtime_pass_by_value=true, detection value 1.0. The harness supplies the detection value(s) via inputs; both the CPU reference and the provider read from the variantPack. A provider that ignores the runtime tensor and uses any other value produces diverging output.", - "inputs": { - "6": { - "kind": "fixed", - "value": 1.0 - } - } - } } ] } diff --git a/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Default/sweep.json b/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Default/sweep.json index e78d025150cc..b55c9370016a 100644 --- a/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Default/sweep.json +++ b/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Default/sweep.json @@ -1190,7 +1190,7 @@ } }, { - "id": "runtime_epsilon_momentum", + "id": "2_3_1_1_bfp16_nhwc_rlo0_norslope_norhi", "values": { "io_data_type": "bfloat16", "tensors": [ @@ -1199,12 +1199,14 @@ "dims": [ 1, 3, + 1, 1 ], "strides": [ 3, 1, - 1 + 3, + 3 ], "data_type": "float" }, @@ -1213,12 +1215,14 @@ "dims": [ 1, 3, + 1, 1 ], "strides": [ 3, 1, - 1 + 3, + 3 ], "data_type": "float" }, @@ -1227,12 +1231,14 @@ "dims": [ 2, 3, + 1, 1 ], "strides": [ 3, 1, - 1 + 3, + 3 ], "data_type": "float" }, @@ -1241,12 +1247,14 @@ "dims": [ 2, 3, + 1, 1 ], "strides": [ 3, 1, - 1 + 3, + 3 ], "data_type": "bfloat16" }, @@ -1255,12 +1263,14 @@ "dims": [ 2, 3, + 1, 1 ], "strides": [ 3, 1, - 1 + 3, + 3 ], "data_type": "bfloat16" }, @@ -1269,11 +1279,13 @@ "dims": [ 1, 3, + 1, 1 ], "strides": [ 3, 1, + 1, 1 ], "data_type": "float" @@ -1290,79 +1302,15 @@ }, { "uid": 7, - "dims": [ - 1 - ], - "strides": [ - 1 - ], - "data_type": "double" - }, - { - "uid": 8, "dims": [ 1, - 3, - 1 - ], - "strides": [ 3, 1, 1 ], - "data_type": "float" - }, - { - "uid": 9, - "dims": [ - 1, - 3, - 1 - ], "strides": [ 3, 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 10, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 11, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, - 1, - 1 - ], - "data_type": "float" - }, - { - "uid": 12, - "dims": [ - 1, - 3, - 1 - ], - "strides": [ - 3, 1, 1 ], @@ -1370,55 +1318,55 @@ } ], "attributes": { - "relu_lower_clip": null, + "relu_lower_clip": 0.0, "relu_lower_clip_slope": null, "relu_upper_clip": null } }, - "tensor_patches": [ - { - "uid": 6, - "set": { - "is_runtime_pass_by_value": true - }, - "remove": [ - "value_type", - "value" - ] - }, - { - "uid": 7, - "set": { - "is_runtime_pass_by_value": true - }, - "remove": [ - "value_type", - "value" - ] - } - ], "metadata": { "format_version": 1, - "operation": "BatchnormPointwise", - "generator": "manual", + "generator": "capture-bundles", "generator_version": "1.0.0", - "generated_at": "2026-07-27T00:00:00Z", - "seed": 42, - "notes": "Runtime PBV scalars: epsilon (uid=6) promoted to is_runtime_pass_by_value=true, detection value 1.0; momentum (uid=7) promoted to is_runtime_pass_by_value=true, detection value 0.9. The harness supplies the detection value(s) via inputs; both the CPU reference and the provider read from the variantPack. A provider that ignores the runtime tensor and uses any other value produces diverging output.", "inputs": { - "6": { - "kind": "fixed", - "value": 1.0 + "4": { + "hi": 1.0, + "kind": "free", + "lo": -1.0, + "seed": 1 }, "7": { - "kind": "fixed", - "value": 0.9 + "hi": 2.0, + "kind": "free", + "lo": -2.0, + "seed": 2 + }, + "5": { + "hi": 2.0, + "kind": "free", + "lo": -2.0, + "seed": 3 + }, + "1": { + "hi": 2.0, + "kind": "free", + "lo": -2.0, + "seed": 1001 + }, + "0": { + "hi": 2.0, + "kind": "free", + "lo": 0.10000000149011612, + "seed": 2001 } - } + }, + "notes": "Captured from C++ graph test Full/IntegrationGpuBatchnormFwdTrainingActiv2dBfp16.Correctness/52", + "operation": "Full/IntegrationGpuBatchnormFwdTrainingActiv2dBfp16", + "seed": 42, + "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv2dBfp16.Correctness_52" } }, { - "id": "2_3_1_1_bfp16_nhwc_rlo0_norslope_norhi", + "id": "2_3_3_1_1_bfp16_ncdhw_norlo_norslope_rhi6", "values": { "io_data_type": "bfloat16", "tensors": [ @@ -1428,13 +1376,15 @@ 1, 3, 1, + 1, 1 ], "strides": [ 3, 1, - 3, - 3 + 1, + 1, + 1 ], "data_type": "float" }, @@ -1444,13 +1394,15 @@ 1, 3, 1, + 1, 1 ], "strides": [ 3, 1, - 3, - 3 + 1, + 1, + 1 ], "data_type": "float" }, @@ -1459,14 +1411,16 @@ "dims": [ 2, 3, + 3, 1, 1 ], "strides": [ + 9, 3, 1, - 3, - 3 + 1, + 1 ], "data_type": "float" }, @@ -1475,14 +1429,16 @@ "dims": [ 2, 3, + 3, 1, 1 ], "strides": [ + 9, 3, 1, - 3, - 3 + 1, + 1 ], "data_type": "bfloat16" }, @@ -1491,14 +1447,16 @@ "dims": [ 2, 3, + 3, 1, 1 ], "strides": [ + 9, 3, 1, - 3, - 3 + 1, + 1 ], "data_type": "bfloat16" }, @@ -1508,12 +1466,14 @@ 1, 3, 1, + 1, 1 ], "strides": [ 3, 1, 1, + 1, 1 ], "data_type": "float" @@ -1534,21 +1494,23 @@ 1, 3, 1, + 1, 1 ], "strides": [ 3, 1, 1, + 1, 1 ], "data_type": "float" } ], "attributes": { - "relu_lower_clip": 0.0, + "relu_lower_clip": null, "relu_lower_clip_slope": null, - "relu_upper_clip": null + "relu_upper_clip": 6.0 } }, "metadata": { @@ -1580,23 +1542,23 @@ "lo": -2.0, "seed": 1001 }, - "0": { + "2": { "hi": 2.0, "kind": "free", "lo": 0.10000000149011612, "seed": 2001 } }, - "notes": "Captured from C++ graph test Full/IntegrationGpuBatchnormFwdTrainingActiv2dBfp16.Correctness/52", - "operation": "Full/IntegrationGpuBatchnormFwdTrainingActiv2dBfp16", + "notes": "Captured from C++ graph test Full/IntegrationGpuBatchnormFwdTrainingActiv3dBfp16.Correctness/13", + "operation": "Full/IntegrationGpuBatchnormFwdTrainingActiv3dBfp16", "seed": 42, - "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv2dBfp16.Correctness_52" + "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv3dBfp16.Correctness_13" } }, { - "id": "2_3_3_1_1_bfp16_ncdhw_norlo_norslope_rhi6", + "id": "2_3_3_1_1_fp32_ndhwc_rlo0p1_norslope_rhi0p5", "values": { - "io_data_type": "bfloat16", + "io_data_type": "float", "tensors": [ { "uid": 0, @@ -1610,9 +1572,9 @@ "strides": [ 3, 1, - 1, - 1, - 1 + 3, + 3, + 3 ], "data_type": "float" }, @@ -1628,9 +1590,9 @@ "strides": [ 3, 1, - 1, - 1, - 1 + 3, + 3, + 3 ], "data_type": "float" }, @@ -1645,10 +1607,10 @@ ], "strides": [ 9, - 3, 1, - 1, - 1 + 3, + 3, + 3 ], "data_type": "float" }, @@ -1663,12 +1625,12 @@ ], "strides": [ 9, - 3, 1, - 1, - 1 + 3, + 3, + 3 ], - "data_type": "bfloat16" + "data_type": "float" }, { "uid": 4, @@ -1681,12 +1643,12 @@ ], "strides": [ 9, - 3, 1, - 1, - 1 + 3, + 3, + 3 ], - "data_type": "bfloat16" + "data_type": "float" }, { "uid": 5, @@ -1736,9 +1698,9 @@ } ], "attributes": { - "relu_lower_clip": null, + "relu_lower_clip": 0.10000000149011612, "relu_lower_clip_slope": null, - "relu_upper_clip": 6.0 + "relu_upper_clip": 0.5 } }, "metadata": { @@ -1770,39 +1732,35 @@ "lo": -2.0, "seed": 1001 }, - "2": { + "0": { "hi": 2.0, "kind": "free", "lo": 0.10000000149011612, "seed": 2001 } }, - "notes": "Captured from C++ graph test Full/IntegrationGpuBatchnormFwdTrainingActiv3dBfp16.Correctness/13", - "operation": "Full/IntegrationGpuBatchnormFwdTrainingActiv3dBfp16", + "notes": "Captured from C++ graph test Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32.Correctness/38", + "operation": "Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32", "seed": 42, - "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv3dBfp16.Correctness_13" + "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32.Correctness_38" } }, { - "id": "2_3_3_1_1_fp32_ndhwc_rlo0p1_norslope_rhi0p5", + "id": "runtime_epsilon", "values": { - "io_data_type": "float", + "io_data_type": "bfloat16", "tensors": [ { "uid": 0, "dims": [ 1, 3, - 1, - 1, 1 ], "strides": [ 3, 1, - 3, - 3, - 3 + 1 ], "data_type": "float" }, @@ -1811,16 +1769,12 @@ "dims": [ 1, 3, - 1, - 1, 1 ], "strides": [ 3, 1, - 3, - 3, - 3 + 1 ], "data_type": "float" }, @@ -1829,16 +1783,12 @@ "dims": [ 2, 3, - 3, - 1, 1 ], "strides": [ - 9, - 1, - 3, 3, - 3 + 1, + 1 ], "data_type": "float" }, @@ -1847,51 +1797,39 @@ "dims": [ 2, 3, - 3, - 1, 1 ], "strides": [ - 9, - 1, - 3, 3, - 3 + 1, + 1 ], - "data_type": "float" + "data_type": "bfloat16" }, { "uid": 4, "dims": [ 2, 3, - 3, - 1, 1 ], "strides": [ - 9, - 1, - 3, 3, - 3 + 1, + 1 ], - "data_type": "float" + "data_type": "bfloat16" }, { "uid": 5, "dims": [ 1, 3, - 1, - 1, 1 ], "strides": [ 3, 1, - 1, - 1, 1 ], "data_type": "float" @@ -1911,66 +1849,48 @@ "dims": [ 1, 3, - 1, - 1, 1 ], "strides": [ 3, 1, - 1, - 1, 1 ], "data_type": "float" } ], "attributes": { - "relu_lower_clip": 0.10000000149011612, + "relu_lower_clip": null, "relu_lower_clip_slope": null, - "relu_upper_clip": 0.5 + "relu_upper_clip": null } }, + "tensor_patches": [ + { + "uid": 6, + "set": { + "is_runtime_pass_by_value": true + }, + "remove": [ + "value_type", + "value" + ] + } + ], "metadata": { "format_version": 1, - "generator": "capture-bundles", + "operation": "BatchnormPointwise", + "generator": "manual", "generator_version": "1.0.0", + "generated_at": "2026-07-27T00:00:00Z", + "seed": 42, + "notes": "Runtime PBV scalars: epsilon (uid=6) promoted to is_runtime_pass_by_value=true, detection value 1.0. The harness supplies the detection value(s) via inputs; both the CPU reference and the provider read from the variantPack. A provider that ignores the runtime tensor and uses any other value produces diverging output.", "inputs": { - "4": { - "hi": 1.0, - "kind": "free", - "lo": -1.0, - "seed": 1 - }, - "7": { - "hi": 2.0, - "kind": "free", - "lo": -2.0, - "seed": 2 - }, - "5": { - "hi": 2.0, - "kind": "free", - "lo": -2.0, - "seed": 3 - }, - "1": { - "hi": 2.0, - "kind": "free", - "lo": -2.0, - "seed": 1001 - }, - "0": { - "hi": 2.0, - "kind": "free", - "lo": 0.10000000149011612, - "seed": 2001 + "6": { + "kind": "fixed", + "value": 1.0 } - }, - "notes": "Captured from C++ graph test Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32.Correctness/38", - "operation": "Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32", - "seed": 42, - "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32.Correctness_38" + } } } ] diff --git a/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Variant2/sweep.json b/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Variant2/sweep.json index d20c0a2d48bf..da37d495aeef 100644 --- a/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Variant2/sweep.json +++ b/dnn-providers/integration-tests/integration-test-bundles/quick/BatchnormPointwise/Variant2/sweep.json @@ -2500,6 +2500,234 @@ "seed": 42, "reference_source": "c++ integration suite: Full/IntegrationGpuBatchnormFwdTrainingActiv3dFp32.Correctness_2" } + }, + { + "id": "runtime_epsilon_momentum", + "values": { + "io_data_type": "bfloat16", + "tensors": [ + { + "uid": 0, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 1, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 2, + "dims": [ + 2, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 3, + "dims": [ + 2, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "bfloat16" + }, + { + "uid": 4, + "dims": [ + 2, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "bfloat16" + }, + { + "uid": 5, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 6, + "dims": [ + 1 + ], + "strides": [ + 1 + ], + "data_type": "double" + }, + { + "uid": 7, + "dims": [ + 1 + ], + "strides": [ + 1 + ], + "data_type": "double" + }, + { + "uid": 8, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 9, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 10, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 11, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + }, + { + "uid": 12, + "dims": [ + 1, + 3, + 1 + ], + "strides": [ + 3, + 1, + 1 + ], + "data_type": "float" + } + ], + "attributes": { + "relu_lower_clip": null, + "relu_lower_clip_slope": null, + "relu_upper_clip": null + } + }, + "tensor_patches": [ + { + "uid": 6, + "set": { + "is_runtime_pass_by_value": true + }, + "remove": [ + "value_type", + "value" + ] + }, + { + "uid": 7, + "set": { + "is_runtime_pass_by_value": true + }, + "remove": [ + "value_type", + "value" + ] + } + ], + "metadata": { + "format_version": 1, + "operation": "BatchnormPointwise", + "generator": "manual", + "generator_version": "1.0.0", + "generated_at": "2026-07-27T00:00:00Z", + "seed": 42, + "notes": "Runtime PBV scalars: epsilon (uid=6) promoted to is_runtime_pass_by_value=true, detection value 1.0; momentum (uid=7) promoted to is_runtime_pass_by_value=true, detection value 0.9. The harness supplies the detection value(s) via inputs; both the CPU reference and the provider read from the variantPack. A provider that ignores the runtime tensor and uses any other value produces diverging output.", + "inputs": { + "6": { + "kind": "fixed", + "value": 1.0 + }, + "7": { + "kind": "fixed", + "value": 0.9 + } + } + } } ] }