(imagenet.Imagenet, dataset.pre_process_vgg, dataset.PostProcessCommon(offset=-1), {"image_size": [224, 224, 3]}), "imagenet_mobilenet": (imagenet.Imagenet, dataset.pre_process_mobilenet, dataset.PostProcessArgMax(offset=-1), {"image_size": [224, 224, 3]}), "imagenet_mobilenet_ncore": (imagenet.Imagenet, dataset.pre_process_mobilenet_uint8, dataset.PostProcessArgMax(offset=-1), {"image_size": [224, 224, 3]}), "coco-300": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), {"image_size": [300, 300, 3]}), "coco-300-ncore": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), {"image_size": [300, 300, 3]}), "coco-300-pt": (coco.Coco, dataset.pre_process_coco_pt_mobilenet, coco.PostProcessCocoPt(False,0.3), {"image_size": [300, 300, 3]}), "coco-1200": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCoco(), {"image_size": [1200, 1200, 3]}), "coco-1200-onnx": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoOnnx(), {"image_size": [1200, 1200, 3]}), "coco-1200-pt": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoPt(True,0.05), {"image_size": [1200, 1200, 3],"use_label_map": True}), "coco-1200-tf": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoTf(), {"image_size": [1200, 1200, 3],"use_label_map": False}), }
(imagenet.Imagenet, dataset.pre_process_mobilenet, dataset.PostProcessArgMax(offset=-1), {"image_size": [224, 224, 3]}), "coco": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), {"image_size": [-1, -1, 3]}), "coco-300": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), {"image_size": [300, 300, 3]}), "coco-1200": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCoco(), {"image_size": [1200, 1200, 3]}), "coco-1200-onnx": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoOnnx(), {"image_size": [1200, 1200, 3]}), "coco-1200-pt": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoPt(), {"image_size": [1200, 1200, 3]}), } # pre-defined command line options so simplify things. They are used as defaults and can be # overwritten from command line DEFAULT_LATENCY_BUCKETS = "0.010,0.050,0.100" SUPPORTED_PROFILES = { "defaults": { "dataset": "imagenet", "backend": "tensorflow", "cache": 0, "time": 60, "queries-single": 1024, "queries-multi": 24576,
"image_size": [224, 224, 3] }), "imagenet_mobilenet": (imagenet.Imagenet, dataset.pre_process_mobilenet, dataset.PostProcessArgMax(offset=-1), { "image_size": [224, 224, 3] }), "coco": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), { "image_size": [-1, -1, 3] }), "coco-300": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), { "image_size": [300, 300, 3] }), "coco-300-pt": (coco.Coco, dataset.pre_process_coco_pt_mobilenet, coco.PostProcessCocoPt(False, 0.3), { "image_size": [300, 300, 3] }), "coco-1200": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCoco(), { "image_size": [1200, 1200, 3] }), "coco-1200-onnx": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoOnnx(), { "image_size": [1200, 1200, 3] }), "coco-1200-pt": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoPt(True, 0.05), { "image_size": [1200, 1200, 3] }), "coco-1200-tf": (coco.Coco, dataset.pre_process_coco_resnet34_tf,
"image_size": [-1, -1, 3] }), "coco-300": (coco.Coco, dataset.pre_process_coco_mobilenet, coco.PostProcessCoco(), { "image_size": [300, 300, 3] }), "coco-1200": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCoco(), { "image_size": [1200, 1200, 3] }), "coco-1200-onnx": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoOnnx(), { "image_size": [1200, 1200, 3] }), "coco-1200-pt": (coco.Coco, dataset.pre_process_coco_resnet34, coco.PostProcessCocoPt(), { "image_size": [1200, 1200, 3] }), } # pre-defined command line options so simplify things. They are used as defaults and can be # overwritten from command line DEFAULT_LATENCY_BUCKETS = "0.010,0.050,0.100,0.200,0.400" SUPPORTED_PROFILES = { "defaults": { "dataset": "imagenet", "backend": "tensorflow", "cache": 0, "time": 128, "max-latency": DEFAULT_LATENCY_BUCKETS,