autoencoder_pruned_finetune_sensing.py

# Copyright (c) Meta Platforms, Inc. and affiliates.
# All rights reserved.

# This source code is licensed under the license found in the
# LICENSE file in the root directory of this source tree.
# --------------------------------------------------------
# References:
# DeiT: https://github.com/facebookresearch/deit
# BEiT: https://github.com/microsoft/unilm/tree/master/beit
# --------------------------------------------------------

import argparse
import datetime
import json
import numpy as np
import os
import time
from pathlib import Path

import torch
import torch.backends.cudnn as cudnn
from torch.utils.tensorboard import SummaryWriter

import util.lr_decay as lrd
import util.misc as misc
from util.misc import NativeScalerWithGradNormCount as NativeScaler
from timm.layers import trunc_normal_
from timm.loss import LabelSmoothingCrossEntropy, SoftTargetCrossEntropy
from timm.data.mixup import Mixup

import models_vit
from advanced_finetuning.lora import create_lora_model

from pruned_engine_finetune import train_one_epoch, evaluate
from dataset_classes.csi_sensing import CSISensingDataset

from advanced_finetuning.lora import LoRALayer, LinearWithLoRA

def get_args_parser():
    parser = argparse.ArgumentParser('MAE fine-tuning for CSI sensing', add_help=False)
    parser.add_argument('--batch_size', default=64, type=int,
                        help='Batch size per GPU (effective batch size is batch_size * accum_iter * # gpus')
    parser.add_argument('--epochs', default=300, type=int)
    parser.add_argument('--accum_iter', default=1, type=int,
                        help='Accumulate gradient iterations (for increasing the effective batch size under memory constraints)')

    # Model parameters
    parser.add_argument('--model', default='seg_vit_large_patch16', type=str, metavar='MODEL',
                        help='Name of model to train')

    parser.add_argument('--model_path', default='', type=str, metavar='MODEL',
                        help='Path of the pruned model to fine-tune')
    
    parser.add_argument('--input_size', default=224, type=int,
                        help='images input size')

    parser.add_argument('--drop_path', type=float, default=0.1, metavar='PCT',
                        help='Drop path rate (default: 0.1)')
    
    parser.add_argument('--frozen_blocks', type=int, default=12, help='number of encoder blocks to freeze. '
                                                          'Freezes all by default')

    parser.add_argument('--lora', action='store_true', help='Whether to use LoRa (default: False)')

    parser.add_argument('--lora_rank', type=int, default=8, help='Rank of LoRa (default: 8)')

    parser.add_argument('--lora_alpha', type=float, default=1, help='Alpha for LoRa (default: 0.5)')

    # Optimizer parameters
    parser.add_argument('--clip_grad', type=float, default=None, metavar='NORM',
                        help='Clip gradient norm (default: None, no clipping)')
    parser.add_argument('--weight_decay', type=float, default=0.05,
                        help='weight decay (default: 0.05)')

    parser.add_argument('--lr', type=float, default=None, metavar='LR',
                        help='learning rate (absolute lr)')
    parser.add_argument('--blr', type=float, default=1e-3, metavar='LR',
                        help='base learning rate: absolute_lr = base_lr * total_batch_size / 256')
    parser.add_argument('--layer_decay', type=float, default=0.75,
                        help='layer-wise lr decay from ELECTRA/BEiT')

    parser.add_argument('--min_lr', type=float, default=1e-6, metavar='LR',
                        help='lower lr bound for cyclic schedulers that hit 0')

    parser.add_argument('--warmup_epochs', type=int, default=25, metavar='N',
                        help='epochs to warmup LR')

    # Augmentation parameters
    parser.add_argument('--color_jitter', type=float, default=None, metavar='PCT',
                        help='Color jitter factor (enabled only when not using Auto/RandAug)')
    parser.add_argument('--aa', type=str, default='rand-m9-mstd0.5-inc1', metavar='NAME',
                        help='Use AutoAugment policy. "v0" or "original". " + "(default: rand-m9-mstd0.5-inc1)'),
    parser.add_argument('--smoothing', type=float, default=0.1,
                        help='Label smoothing (default: 0)')

    # * Random Erase params
    parser.add_argument('--reprob', type=float, default=0.25, metavar='PCT',
                        help='Random erase prob (default: 0.25)')
    parser.add_argument('--remode', type=str, default='pixel',
                        help='Random erase mode (default: "pixel")')
    parser.add_argument('--recount', type=int, default=1,
                        help='Random erase count (default: 1)')
    parser.add_argument('--resplit', action='store_true', default=False,
                        help='Do not random erase first (clean) augmentation split')

    # * Mixup params
    parser.add_argument('--mixup', type=float, default=0,
                        help='mixup alpha, mixup enabled if > 0.')
    parser.add_argument('--cutmix', type=float, default=0,
                        help='cutmix alpha, cutmix enabled if > 0.')
    parser.add_argument('--cutmix_minmax', type=float, nargs='+', default=None,
                        help='cutmix min/max ratio, overrides alpha and enables cutmix if set (default: None)')
    parser.add_argument('--mixup_prob', type=float, default=1.0,
                        help='Probability of performing mixup or cutmix when either/both is enabled')
    parser.add_argument('--mixup_switch_prob', type=float, default=0.5,
                        help='Probability of switching to cutmix when both mixup and cutmix enabled')
    parser.add_argument('--mixup_mode', type=str, default='batch',
                        help='How to apply mixup/cutmix params. Per "batch", "pair", or "elem"')

    # * Finetuning params
    parser.add_argument('--finetune', default='',
                        help='finetune from checkpoint')
    parser.add_argument('--global_pool', default='token')

    # Dataset parameters
    parser.add_argument('--data_path', default='', type=str,
                        help='dataset path')
    parser.add_argument('--nb_classes', default=6, type=int,
                        help='number of the classification types')

    parser.add_argument('--output_dir', default='./has_output_dir',
                        help='path where to save, empty for no saving')
    parser.add_argument('--log_dir', default='./has_output_dir',
                        help='path where to tensorboard log')
    parser.add_argument('--device', default='cuda',
                        help='device to use for training / testing')
    parser.add_argument('--seed', default=0, type=int)
    parser.add_argument('--resume', default='',
                        help='resume from checkpoint')

    parser.add_argument('--start_epoch', default=0, type=int, metavar='N',
                        help='start epoch')
    parser.add_argument('--eval', action='store_true',
                        help='Perform evaluation only')
    parser.add_argument('--dist_eval', action='store_true', default=False,
                        help='Enabling distributed evaluation (recommended during training for faster monitor')
    parser.add_argument('--num_workers', default=10, type=int)
    parser.add_argument('--pin_mem', action='store_true',
                        help='Pin CPU memory in DataLoader for more efficient (sometimes) transfer to GPU.')
    parser.add_argument('--no_pin_mem', action='store_false', dest='pin_mem')
    parser.set_defaults(pin_mem=True)

    # distributed training parameters
    parser.add_argument('--world_size', default=1, type=int,
                        help='number of distributed processes')
    parser.add_argument('--local_rank', default=-1, type=int)
    parser.add_argument('--dist_on_itp', action='store_true')
    parser.add_argument('--dist_url', default='env://',
                        help='url used to set up distributed training')
    parser.add_argument('--downsampled', action='store_true', default=False)

    return parser


def main(args):
    misc.init_distributed_mode(args)

    print('job dir: {}'.format(os.path.dirname(os.path.realpath(__file__))))
    print("{}".format(args).replace(', ', ',\n'))

    device = torch.device(args.device)

    # fix the seed for reproducibility
    seed = args.seed + misc.get_rank()
    torch.manual_seed(seed)
    np.random.seed(seed)

    cudnn.benchmark = True

    dataset_train = CSISensingDataset(Path('/home/ict317-3/Mohammad/mae/fine-tuning_datasets/NTU-Fi_HAR/train'), downsampled=args.downsampled)
    dataset_val = CSISensingDataset(Path('/home/ict317-3/Mohammad/mae/fine-tuning_datasets/NTU-Fi_HAR/test'), downsampled=args.downsampled)

    num_tasks = misc.get_world_size()
    global_rank = misc.get_rank()
    sampler_train = torch.utils.data.DistributedSampler(
        dataset_train, num_replicas=num_tasks, rank=global_rank, shuffle=True
    )
    print("Sampler_train = %s" % str(sampler_train))
    if args.dist_eval:
        if len(dataset_val) % num_tasks != 0:
            print('Warning: Enabling distributed evaluation with an eval dataset not divisible by process number. '
                  'This will slightly alter validation results as extra duplicate entries are added to achieve '
                  'equal num of samples per-process.')
        sampler_val = torch.utils.data.DistributedSampler(
            dataset_val, num_replicas=num_tasks, rank=global_rank, shuffle=True)  # shuffle=True to reduce monitor bias
    else:
        sampler_val = torch.utils.data.SequentialSampler(dataset_val)

    if global_rank == 0 and args.log_dir is not None and not args.eval:
        os.makedirs(args.log_dir, exist_ok=True)
        log_writer = SummaryWriter(log_dir=args.log_dir)
    else:
        log_writer = None

    data_loader_train = torch.utils.data.DataLoader(
        dataset_train, sampler=sampler_train,
        batch_size=args.batch_size,
        num_workers=args.num_workers,
        pin_memory=args.pin_mem,
        drop_last=True,
    )

    data_loader_val = torch.utils.data.DataLoader(
        dataset_val, sampler=sampler_val,
        batch_size=args.batch_size,
        num_workers=args.num_workers,
        pin_memory=args.pin_mem,
        drop_last=False
    )

    mixup_active = args.mixup > 0 or args.cutmix > 0. or args.cutmix_minmax is not None
    if mixup_active:
        print("Mixup is activated!")
        mixup_fn = Mixup(
            mixup_alpha=args.mixup, cutmix_alpha=args.cutmix, cutmix_minmax=args.cutmix_minmax,
            prob=args.mixup_prob, switch_prob=args.mixup_switch_prob, mode=args.mixup_mode,
            label_smoothing=args.smoothing, num_classes=args.nb_classes)
    else:
        mixup_fn = None

    # model = models_vit.__dict__[args.model](global_pool=args.global_pool, num_classes=args.nb_classes,
    #                                         drop_path_rate=args.drop_path)
    # if args.lora:
    #     model = create_lora_model(model, args.lora_rank, args.lora_alpha)

    # if args.finetune and not args.eval:
    #     checkpoint = torch.load(args.finetune, map_location='cpu')

    #     print("Load pre-trained checkpoint from: %s" % args.finetune)
    #     checkpoint_model = checkpoint['model']
    #     state_dict = model.state_dict()
    #     for k in ['head.weight', 'head.bias', 'pos_embed']:
    #         if k in checkpoint_model and checkpoint_model[k].shape != state_dict[k].shape:
    #             print(f"Removing key {k} from pretrained checkpoint")
    #             del checkpoint_model[k]
    #     checkpoint_model['patch_embed.proj.weight'] = checkpoint_model['patch_embed.proj.weight'].expand(-1, 3, -1, -1)

    #     # load pre-trained model
    #     msg = model.load_state_dict(checkpoint_model, strict=False)
    #     print(msg)
    #     # manually initialize fc layer
    #     trunc_normal_(model.head.weight, std=2e-5)

    # if not args.lora:
    #     model.freeze_encoder()

    model_path = args.model_path 
    print("The path of the pruned model is: ", model_path)
    model = torch.load(model_path, weights_only=False)

    print(model)
    
    if args.frozen_blocks is not None:
        # Freeze the patch embedding layer
        for param in model.patch_embed.proj.parameters():
            param.requires_grad = False

        # Freeze all transformer blocks
        for param in model.blocks.parameters():
            param.requires_grad = False

        
        for param in model.decoder_blocks.parameters():
            param.requires_grad = False

        # Freeze the final normalization layer
        # for param in model.norm.parameters():
        #     param.requires_grad = False

        # Freeze positional embeddings and tokens
        if hasattr(model, "cls_token"):
            model.cls_token.requires_grad = False
        if hasattr(model, "pos_embed"):
            model.pos_embed.requires_grad = False
        if hasattr(model, "mask_token"):
            model.mask_token.requires_grad = False
        if hasattr(model, "decoder_pos_embed"):
            model.decoder_pos_embed.requires_grad = False
            


    # Confirm that only the classification head is trainable
    for name, param in model.named_parameters():
        print(f"{name}: {'Trainable' if param.requires_grad else 'Frozen'}")

    print("The number of frozen blocks is: ", args.frozen_blocks)

    model.to(device)

    # model_without_ddp = model
    n_parameters = sum(p.numel() for p in model.parameters() if p.requires_grad)

    # print("Model = %s" % str(model_without_ddp))
    print('number of params (M): %.2f' % (n_parameters / 1.e6))

    eff_batch_size = args.batch_size * args.accum_iter * misc.get_world_size()
    
    if args.lr is None:  # only base_lr is specified
        args.lr = args.blr * eff_batch_size / 256

    print("base lr: %.2e" % (args.lr * 256 / eff_batch_size))
    print("actual lr: %.2e" % args.lr)

    print("accumulate grad iterations: %d" % args.accum_iter)
    print("effective batch size: %d" % eff_batch_size)

    # if args.distributed:
    #     model = torch.nn.parallel.DistributedDataParallel(model, device_ids=[args.gpu])
    #     model_without_ddp = model.module

    # build optimizer with layer-wise lr decay (lrd)
    param_groups = lrd.param_groups_lrd(model, args.weight_decay, layer_decay=args.layer_decay)
    optimizer = torch.optim.AdamW(param_groups, lr=args.lr)
    loss_scaler = NativeScaler()

    if mixup_fn is not None:
        # smoothing is handled with mixup label transform
        criterion = SoftTargetCrossEntropy()
    elif args.smoothing > 0.:
        criterion = LabelSmoothingCrossEntropy(smoothing=args.smoothing)
    else:
        criterion = torch.nn.CrossEntropyLoss()

    print("criterion = %s" % str(criterion))

    # misc.load_model(args=args, model_without_ddp=model_without_ddp, optimizer=optimizer, loss_scaler=loss_scaler)

    if args.eval:
        test_stats = evaluate(data_loader_val, model, criterion, device)
        print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")
        exit(0)

    print(f"Start training for {args.epochs} epochs")
    start_time = time.time()
    max_accuracy = 0.0
    for epoch in range(args.start_epoch, args.epochs):
        if args.distributed:
            data_loader_train.sampler.set_epoch(epoch)
        train_stats = train_one_epoch(
            model, criterion, data_loader_train,
            optimizer, device, epoch, loss_scaler,
            args.clip_grad, mixup_fn,
            log_writer=log_writer,
            args=args
        )
        # if args.output_dir and (epoch + 1) % 10 == 0:
        #     misc.save_model(
        #         args=args, model=model, model_without_ddp=model_without_ddp, optimizer=optimizer,
        #         loss_scaler=loss_scaler, epoch=epoch)

        test_stats = evaluate(data_loader_val, model, criterion, device)
        print(f"Accuracy of the network on the {len(dataset_val)} test images: {test_stats['acc1']:.1f}%")
        max_accuracy = max(max_accuracy, test_stats["acc1"])
        print(f'Max accuracy: {max_accuracy:.2f}%')

        if test_stats["acc1"] == max_accuracy:
            print("A new better model has been saved ... ")
            torch.save(model, 'has_output_dir/best_model.pth')

        if log_writer is not None:
            log_writer.add_scalar('perf/test_acc1', test_stats['acc1'], epoch)
            log_writer.add_scalar('perf/test_acc3', test_stats['acc3'], epoch)
            log_writer.add_scalar('perf/test_loss', test_stats['loss'], epoch)

        log_stats = {**{f'train_{k}': v for k, v in train_stats.items()},
                     **{f'test_{k}': v for k, v in test_stats.items()},
                     'epoch': epoch,
                     'n_parameters': n_parameters}

        # if args.output_dir and misc.is_main_process():
        #     if log_writer is not None:
        #         log_writer.flush()
        #     with open(os.path.join(args.output_dir, "log.txt"), mode="a", encoding="utf-8") as f:
        #         f.write(json.dumps(log_stats) + "\n")

    total_time = time.time() - start_time
    total_time_str = str(datetime.timedelta(seconds=int(total_time)))
    print('Training time {}'.format(total_time_str))


if __name__ == '__main__':
    args = get_args_parser()
    args = args.parse_args()
    if args.output_dir:
        Path(args.output_dir).mkdir(parents=True, exist_ok=True)
    main(args)