fbnet.py

# Copyright (c) Qualcomm Innovation Center, Inc.
# All rights reserved
#
# This source code is licensed under the BSD-style license found in the
# LICENSE file in the root directory of this source tree.

import json
import os
import re
from multiprocessing.connection import Client

import numpy as np
import timm
from executorch.backends.qualcomm.quantizer.quantizer import QuantDtype
from executorch.examples.qualcomm.utils import (
    build_executorch_binary,
    get_imagenet_dataset,
    make_output_dir,
    setup_common_args_and_variables,
    SimpleADB,
    topk_accuracy,
)


def main(args):
    if not args.compile_only and args.device is None:
        raise RuntimeError(
            "device serial is required if not compile only. "
            "Please specify a device serial by -s/--device argument."
        )

    # ensure the working directory exist.
    os.makedirs(args.artifact, exist_ok=True)

    instance = timm.create_model("fbnetc_100", pretrained=True).eval()

    data_num = 100
    inputs, targets, input_list = get_imagenet_dataset(
        dataset_path=f"{args.dataset}",
        data_size=data_num,
        image_shape=(299, 299),
    )

    pte_filename = "fbnet"

    build_executorch_binary(
        instance,
        inputs[0],
        args.model,
        f"{args.artifact}/{pte_filename}",
        inputs,
        quant_dtype=QuantDtype.use_8a8w,
        shared_buffer=args.shared_buffer,
    )

    if args.compile_only:
        return

    adb = SimpleADB(
        qnn_sdk=os.getenv("QNN_SDK_ROOT"),
        build_path=f"{args.build_folder}",
        pte_path=f"{args.artifact}/{pte_filename}.pte",
        workspace=f"/data/local/tmp/executorch/{pte_filename}",
        device_id=args.device,
        host_id=args.host,
        soc_model=args.model,
    )
    adb.push(inputs=inputs, input_list=input_list)
    adb.execute()

    # collect output data
    output_data_folder = f"{args.artifact}/outputs"
    make_output_dir(output_data_folder)

    output_raws = []

    def post_process():
        for f in sorted(
            os.listdir(output_data_folder), key=lambda f: int(f.split("_")[1])
        ):
            filename = os.path.join(output_data_folder, f)
            if re.match(r"^output_[0-9]+_[1-9].raw$", f):
                os.remove(filename)
            else:
                output = np.fromfile(filename, dtype=np.float32)
                output_raws.append(output)

    adb.pull(output_path=args.artifact, callback=post_process)

    # top-k analysis
    predictions = []
    for i in range(data_num):
        predictions.append(
            np.fromfile(
                os.path.join(output_data_folder, f"output_{i}_0.raw"), dtype=np.float32
            )
        )

    k_val = [1, 5]
    topk = [topk_accuracy(predictions, targets, k).item() for k in k_val]
    if args.ip and args.port != -1:
        with Client((args.ip, args.port)) as conn:
            conn.send(json.dumps({f"top_{k}": topk[i] for i, k in enumerate(k_val)}))
    else:
        for i, k in enumerate(k_val):
            print(f"top_{k}->{topk[i]}%")


if __name__ == "__main__":
    parser = setup_common_args_and_variables()
    parser.add_argument(
        "-a",
        "--artifact",
        help="path for storing generated artifacts by this example. Default ./fbnet",
        default="./fbnet",
        type=str,
    )

    parser.add_argument(
        "-d",
        "--dataset",
        help=(
            "path to the validation folder of ImageNet dataset. "
            "e.g. --dataset imagenet-mini/val "
            "for https://www.kaggle.com/datasets/ifigotin/imagenetmini-1000)"
        ),
        type=str,
        required=True,
    )

    args = parser.parse_args()
    try:
        main(args)
    except Exception as e:
        if args.ip and args.port != -1:
            with Client((args.ip, args.port)) as conn:
                conn.send(json.dumps({"Error": str(e)}))
        else:
            raise Exception(e)