train.py

# Copyright 2020 InterDigital Communications, Inc.
#
# Licensed under the Apache License, Version 2.0 (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     http://www.apache.org/licenses/LICENSE-2.0
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# This file may have been modified by Bytedance Inc. (“Bytedance Modifications”). All Bytedance Modifications are Copyright 2022 Bytedance Inc.


# Copyright 2023 Bytedance Inc.
# All rights reserved.
# Licensed under the BSD 3-Clause Clear License (the "License");
# you may not use this file except in compliance with the License.
# You may obtain a copy of the License at
#
#     https://choosealicense.com/licenses/bsd-3-clause-clear/
#
# Unless required by applicable law or agreed to in writing, software
# distributed under the License is distributed on an "AS IS" BASIS,
# WITHOUT WARRANTIES OR CONDITIONS OF ANY KIND, either express or implied.
# See the License for the specific language governing permissions and
# limitations under the License.

# Redistribution and use in source and binary forms, with or without
# modification, are permitted (subject to the limitations in the disclaimer
# below) provided that the following conditions are met:

# * Redistributions of source code must retain the above copyright notice,
#   this list of conditions and the following disclaimer.
# * Redistributions in binary form must reproduce the above copyright notice,
#   this list of conditions and the following disclaimer in the documentation
#   and/or other materials provided with the distribution.
# * Neither the name of InterDigital Communications, Inc nor the names of its
#   contributors may be used to endorse or promote products derived from this
#   software without specific prior written permission.

# NO EXPRESS OR IMPLIED LICENSES TO ANY PARTY'S PATENT RIGHTS ARE GRANTED BY
# THIS LICENSE. THIS SOFTWARE IS PROVIDED BY THE COPYRIGHT HOLDERS AND
# CONTRIBUTORS "AS IS" AND ANY EXPRESS OR IMPLIED WARRANTIES, INCLUDING, BUT
# NOT LIMITED TO, THE IMPLIED WARRANTIES OF MERCHANTABILITY AND FITNESS FOR A
# PARTICULAR PURPOSE ARE DISCLAIMED. IN NO EVENT SHALL THE COPYRIGHT HOLDER OR
# CONTRIBUTORS BE LIABLE FOR ANY DIRECT, INDIRECT, INCIDENTAL, SPECIAL,
# EXEMPLARY, OR CONSEQUENTIAL DAMAGES (INCLUDING, BUT NOT LIMITED TO,
# PROCUREMENT OF SUBSTITUTE GOODS OR SERVICES; LOSS OF USE, DATA, OR PROFITS;
# OR BUSINESS INTERRUPTION) HOWEVER CAUSED AND ON ANY THEORY OF LIABILITY,
# WHETHER IN CONTRACT, STRICT LIABILITY, OR TORT (INCLUDING NEGLIGENCE OR
# OTHERWISE) ARISING IN ANY WAY OUT OF THE USE OF THIS SOFTWARE, EVEN IF
# ADVISED OF THE POSSIBILITY OF SUCH DAMAGE.

import argparse
import math
import random
import shutil
import sys

import torch
import torch.nn as nn
import torch.optim as optim

from torch.utils.data import DataLoader
from torchvision import transforms

from compressai.datasets import ImageFolder
from Cheng2020Attention import Cheng2020Attention


class RateDistortionLoss(nn.Module):
    """Custom rate distortion loss with a Lagrangian parameter."""

    def __init__(self):
        super().__init__()
        self.mse = nn.MSELoss()

    def forward(self, output, target, lmbda):
        N, _, H, W = target.size()
        out = {}
        num_pixels = N * H * W

        out["bpp_loss"] = sum(
            (torch.log(likelihoods).sum() / (-math.log(2) * num_pixels))
            for likelihoods in output["likelihoods"].values()
        )
        out["mse_loss"] = self.mse(output["x_hat"], target) * 255 ** 2
        out["loss"] = lmbda  * out["mse_loss"] + out["bpp_loss"]

        return out


class AverageMeter:
    """Compute running average."""

    def __init__(self):
        self.val = 0
        self.avg = 0
        self.sum = 0
        self.count = 0

    def update(self, val, n=1):
        self.val = val
        self.sum += val * n
        self.count += n
        self.avg = self.sum / self.count


class CustomDataParallel(nn.DataParallel):
    """Custom DataParallel to access the module methods."""

    def __getattr__(self, key):
        try:
            return super().__getattr__(key)
        except AttributeError:
            return getattr(self.module, key)


def configure_optimizers(net, args):
    """Separate parameters for the main optimizer and the auxiliary optimizer.
    Return two optimizers"""

    parameters = {
        n
        for n, p in net.named_parameters()
        if not n.endswith(".quantiles") and p.requires_grad
    }
    aux_parameters = {
        n
        for n, p in net.named_parameters()
        if n.endswith(".quantiles") and p.requires_grad
    }

    # Make sure we don't have an intersection of parameters
    params_dict = dict(net.named_parameters())
    inter_params = parameters & aux_parameters
    union_params = parameters | aux_parameters

    assert len(inter_params) == 0
    assert len(union_params) - len(params_dict.keys()) == 0

    optimizer = optim.Adam(
        (params_dict[n] for n in sorted(parameters)),
        lr=args.learning_rate,
    )
    aux_optimizer = optim.Adam(
        (params_dict[n] for n in sorted(aux_parameters)),
        lr=args.aux_learning_rate,
    )
    return optimizer, aux_optimizer


def train_one_epoch(
    model, criterion, train_dataloader, optimizer, aux_optimizer, epoch, clip_max_norm, noise, stage
):
    model.train()
    device = next(model.parameters()).device

    for i, d in enumerate(train_dataloader):
        d = d.to(device)
        if stage > 1:
            s = random.randint(0, model.levels - 1)  # choose random level from [0, levels-1]
        else:
            s = model.levels - 1

        optimizer.zero_grad()
        aux_optimizer.zero_grad()

        out_net = model(d, noise, stage, s)

        out_criterion = criterion(out_net, d, model.lmbda[s])
        out_criterion["loss"].backward()
        if clip_max_norm > 0:
            torch.nn.utils.clip_grad_norm_(model.parameters(), clip_max_norm)
        optimizer.step()

        aux_loss = model.aux_loss()
        aux_loss.backward()
        aux_optimizer.step()

        if i % 200 == 0:
            print(
                f"Train epoch {epoch} stage{stage}: ["
                f"{i*len(d)}/{len(train_dataloader.dataset)}"
                f" ({100. * i / len(train_dataloader):.0f}%)]"
				f" \tlambda: {model.lmbda[s]} s: {s:.3f}, scale: {model.Gain.data[s].detach().cpu().numpy():0.4f}, |"
                f'\tLoss: {out_criterion["loss"].item():.3f} |'
                f'\tMSE loss: {out_criterion["mse_loss"].item():.3f} |'
                f'\tBpp loss: {out_criterion["bpp_loss"].item():.2f} |'
                f"\tAux loss: {aux_loss.item():.2f}"
            )


def test_epoch(epoch, test_dataloader, model, criterion, noise, stage):
    model.eval()
    device = next(model.parameters()).device
    print("stage:{}, noise quantization:{}".format(stage, noise))


    loss_total = 0
    bpp_loss_total = 0
    mse_loss_total = 0

    with torch.no_grad():
        for s in range(model.levels):
            loss = AverageMeter()
            bpp_loss = AverageMeter()
            mse_loss = AverageMeter()
            aux_loss = AverageMeter()
            for d in test_dataloader:
                d = d.to(device)
                out_net = model(x=d, noise=noise, stage=stage, s=s)
                out_criterion = criterion(out_net, d, model.lmbda[s])

                aux_loss.update(model.aux_loss().item())
                bpp_loss.update(out_criterion["bpp_loss"].item())
                loss.update(out_criterion["loss"].item())
                mse_loss.update(out_criterion["mse_loss"].item())

            loss_total += loss.avg
            bpp_loss_total += bpp_loss.avg
            mse_loss_total += mse_loss.avg
			
            print(
                f"Test epoch {epoch}, lambda: {model.lmbda[s]}, s: {s}, scale: {model.Gain.data[s].cpu().numpy():0.4f}, stage {stage}:"
                f"\tLoss: {loss.avg:.3f} |"
                f"\tMSE loss: {mse_loss.avg:.3f} |"
                f"\tBpp loss: {bpp_loss.avg:.4f} |"
                f"\tAux loss: {aux_loss.avg:.4f}"
            )
    print(
        f"Test epoch {epoch} : Total Average losses:"
        f"\tLoss: {loss_total:.3f} |"
        f"\tMSE loss: {mse_loss_total:.3f} |"
        f"\tBpp loss: {bpp_loss_total:.4f} \n"
    )

    return loss_total


def save_checkpoint(state, is_best, filename="checkpoint.pth.tar"):
    torch.save(state, filename)
    if is_best:
        shutil.copyfile(filename, "checkpoint_best_loss.pth.tar")


def parse_args(argv):
    parser = argparse.ArgumentParser(description="Example training script.")
    parser.add_argument(
        "-d", "--dataset", type=str, required=True, help="Training dataset"
    )
    parser.add_argument(
        "-e",
        "--epochs",
        default=100,
        type=int,
        help="Number of epochs (default: %(default)s)",
    )
    parser.add_argument(
        "-lr",
        "--learning-rate",
        default=1e-4,
        type=float,
        help="Learning rate (default: %(default)s)",
    )
    parser.add_argument(
        "-n",
        "--num-workers",
        type=int,
        default=4,
        help="Dataloaders threads (default: %(default)s)",
    )
    parser.add_argument(
        "--batch-size", type=int, default=16, help="Batch size (default: %(default)s)"
    )
    parser.add_argument(
        "--test-batch-size",
        type=int,
        default=64,
        help="Test batch size (default: %(default)s)",
    )
    parser.add_argument(
        "--aux-learning-rate",
        default=1e-3,
        type=float,
        help="Auxiliary loss learning rate (default: %(default)s)",
    )
    parser.add_argument(
        "--patch-size",
        type=int,
        nargs=2,
        default=(256, 256),
        help="Size of the patches to be cropped (default: %(default)s)",
    )
    parser.add_argument("--cuda", action="store_true", help="Use cuda")
    parser.add_argument(
        "--save", action="store_true", default=True, help="Save model to disk"
    )
    parser.add_argument(
        "--seed", type=float, help="Set random seed for reproducibility"
    )
    parser.add_argument(
        "--clip_max_norm",
        default=1.0,
        type=float,
        help="gradient clipping max norm (default: %(default)s",
    )
    parser.add_argument("--checkpoint", type=str, help="Path to a checkpoint")
    parser.add_argument('--stage', default=0, type=int, help='trainning stage')
    parser.add_argument("--ste", default=0, type=int, help="Using ste round in the finetune stage")
    parser.add_argument('--loadFromPretrainedSinglemodel', default=0, type=int, help='load models from single rate')
    parser.add_argument("--refresh", default=0, type=int, help="refresh the setting of optimizer and epoch",
    )
    args = parser.parse_args(argv)
    return args


def main(argv):
    args = parse_args(argv)
    import os
    os.environ['CUDA_VISIBLE_DEVICES'] = '1'
    if args.seed is not None:
        torch.manual_seed(args.seed)
        random.seed(args.seed)

    train_transforms = transforms.Compose(
        [transforms.RandomCrop(args.patch_size), transforms.ToTensor()]
    )

    test_transforms = transforms.Compose(
        [transforms.CenterCrop(args.patch_size), transforms.ToTensor()]
    )

    train_dataset = ImageFolder(args.dataset, split="train", transform=train_transforms)
    test_dataset = ImageFolder(args.dataset, split="test", transform=test_transforms)

    device = "cuda" if args.cuda and torch.cuda.is_available() else "cpu"

    train_dataloader = DataLoader(
        train_dataset,
        batch_size=args.batch_size,
        num_workers=args.num_workers,
        shuffle=True,
        pin_memory=(device == "cuda"),
    )

    test_dataloader = DataLoader(
        test_dataset,
        batch_size=args.test_batch_size,
        num_workers=args.num_workers,
        shuffle=False,
        pin_memory=(device == "cuda"),
    )

    net = Cheng2020Attention()
    net = net.to(device)



    optimizer, aux_optimizer = configure_optimizers(net, args)
    lr_scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, "min", factor=0.5, patience=20)
    criterion = RateDistortionLoss()

    last_epoch = 0
    best_loss = float("inf")
    if args.checkpoint:  # load from previous checkpoint
        if args.loadFromPretrainedSinglemodel:
            print("Loading single lambda pretrained checkpoint: ", args.checkpoint)
            checkpoint = torch.load(args.checkpoint, map_location=device)
            if "state_dict" in checkpoint:
                ckpt = checkpoint["state_dict"]
            else:
                ckpt = checkpoint
            model_dict = net.state_dict()
            pretrained_dict = {k: v for k, v in ckpt.items() if
                               k in model_dict.keys() and v.shape == model_dict[k].shape}

            model_dict.update(pretrained_dict)
            net.load_state_dict(model_dict)
        else:
            print("Loading: ", args.checkpoint)
            checkpoint = torch.load(args.checkpoint, map_location=device)
            last_epoch = checkpoint["epoch"] + 1
            best_loss = checkpoint["best_loss"]
            net.load_state_dict(checkpoint["state_dict"])
            optimizer.load_state_dict(checkpoint["optimizer"])
            aux_optimizer.load_state_dict(checkpoint["aux_optimizer"])
            lr_scheduler.load_state_dict(checkpoint["lr_scheduler"])
    if args.refresh:
        optimizer.param_groups[0]['lr'] = args.learning_rate
        aux_optimizer.param_groups[0]['lr'] = args.aux_learning_rate
        lr_scheduler = optim.lr_scheduler.ReduceLROnPlateau(optimizer, "min", factor=0.5, patience=20)
        last_epoch = 0

    if args.cuda and torch.cuda.device_count() > 1:
        net = CustomDataParallel(net)

    stage = args.stage
    noise = True
    ste = False  
    if args.ste or stage>2:
        ste = True
        noise = False
    
    for epoch in range(last_epoch, args.epochs):
        print("noise quant: {}, ste quant:{}, stage:{}".format(noise, ste, stage))
        print(f"Learning rate: {optimizer.param_groups[0]['lr']}")
        train_one_epoch(
            net,
            criterion,
            train_dataloader,
            optimizer,
            aux_optimizer,
            epoch,
            args.clip_max_norm,
			noise,
            stage,
        )
        loss = test_epoch(epoch, test_dataloader, net, criterion, noise, stage, )
        lr_scheduler.step(loss)

        is_best = loss < best_loss
        best_loss = min(loss, best_loss)

        if args.save:
            save_checkpoint(
                {
                    "epoch": epoch,
                    "state_dict": net.state_dict(),
                    "loss": loss,
					"best_loss": best_loss,
                    "optimizer": optimizer.state_dict(),
                    "aux_optimizer": aux_optimizer.state_dict(),
                    "lr_scheduler": lr_scheduler.state_dict(),
                },
                is_best,
            )


if __name__ == "__main__":
    main(sys.argv[1:])