trainAugcomp.py

import argparse

import math
import numpy as np
import torch
import torch.nn as nn
from torch.optim import SGD, lr_scheduler
from torch.utils.data import DataLoader
from tqdm import tqdm

from PIL import Image

from data.augmentations import get_transform
from FER import RafDataset,get_RAFDB_datasets


from util.general_utils import AverageMeter, init_experiment
from util.cluster_and_log_utils import log_accs_from_preds
from config import exp_root
from model import DINOHead, info_nce_logits, SupConLoss, DistillLoss, ContrastiveLearningViewGenerator, get_params_groups
from model import DynamicLSR
from model import DINOHeadExtended,WorstCaseEstimationLoss,TotalClusteringLoss
def train(student, train_loader, test_loader, unlabelled_train_loader, args):

    params_groups = get_params_groups(student)
    optimizer = SGD(params_groups, lr=args.lr, momentum=args.momentum, weight_decay=args.weight_decay)

    fp16_scaler = None
    if args.fp16:
        fp16_scaler = torch.cuda.amp.GradScaler()
    total_norm = 0
    exp_lr_scheduler = lr_scheduler.CosineAnnealingLR(
            optimizer,
            T_max=args.epochs,
            eta_min=args.lr * 1e-3,
        )


    cluster_criterion = DistillLoss(
                        args.warmup_teacher_temp_epochs,
                        args.epochs,
                        args.n_views,
                        args.warmup_teacher_temp,
                        args.teacher_temp,
                    )
    
    
    total_loss_func = TotalClusteringLoss(
        num_classes=args.num_labeled_classes
        feat_dim=args.feat_dim,
        device=device,
        wb_weight=1.0,  
        max_min_weight=0.1  
       
    ).to(device)
    
    worst_case_criterion = WorstCaseEstimationLoss(eta_prime=args.eta_prime)
    # worst_case_criterion = WorstCaseEstimationLoss_v2(eta_prime=args.eta_prime)
    # inductive
    best_test_acc_lab = 0
    # transductive
    best_train_acc_lab = 0
    best_train_acc_ubl = 0 
    best_train_acc_all = 0
    
    for epoch in range(args.epochs):
        loss_record = AverageMeter()

        student.train()
        for batch_idx, batch in enumerate(train_loader):
            images, class_labels, uq_idxs, mask_lab = batch
            mask_lab = mask_lab[:, 0]

            class_labels, mask_lab = class_labels.cuda(non_blocking=True), mask_lab.cuda(non_blocking=True).bool()
            images = torch.cat(images, dim=0).cuda(non_blocking=True)

            with torch.cuda.amp.autocast(fp16_scaler is not None):
                student_proj, student_out, aux_logits = student(images)
                teacher_out = student_out.detach()

                # clustering, sup
                sup_logits = torch.cat([f[mask_lab] for f in (student_out / 0.1).chunk(2)], dim=0)
                sup_labels = torch.cat([class_labels[mask_lab] for _ in range(2)], dim=0)
                cls_loss = nn.CrossEntropyLoss()(sup_logits, sup_labels)#cls_loss
                
                
                total_steps = len(train_loader) *  (epoch+1)
                initial_e = 0.1
                dynamic_lsr = DynamicLSR(initial_e=0.1, total_steps=total_steps)
                LSR_loss = dynamic_lsr(sup_logits, sup_labels)
               


                # clustering, unsup
                cluster_loss = cluster_criterion(student_out, teacher_out, epoch)
                avg_probs = (student_out / 0.1).softmax(dim=1).mean(dim=0)
                me_max_loss = - torch.sum(torch.log(avg_probs**(-avg_probs))) + math.log(float(len(avg_probs)))
                cluster_loss += args.memax_weight * me_max_loss

                # represent learning, unsup
                contrastive_logits, contrastive_labels = info_nce_logits(features=student_proj)
#                 print(contrastive_labels)
                contrastive_loss = torch.nn.CrossEntropyLoss()(contrastive_logits, contrastive_labels)
                

                # representation learning, sup
                student_proj = torch.cat([f[mask_lab].unsqueeze(1) for f in student_proj.chunk(2)], dim=1)
                student_proj = torch.nn.functional.normalize(student_proj, dim=-1)
                sup_con_labels = class_labels[mask_lab]
                sup_con_loss = SupConLoss()(student_proj, labels=sup_con_labels)
                
                if epoch >= args.cluster_epochs:
                    total_loss = total_loss_func(student_proj , sup_con_labels)
                else:
                    total_loss = torch.tensor(0.0).cuda()  
                # if epoch >= args.cluster_epochs:
                #     total_loss = total_loss_func(sup_con_logits ,sup_con_labels)
                # else:
                #     total_loss = torch.tensor(0.0).cuda()  
                    
                '''
                aughead
                '''
                # if epoch >= args.worst_epochs:
                #     student_out_views = student_out.chunk(2)
                #     aux_logits_views = aux_logits.chunk(2)
                    
                #     y_l = student_out_views[0][mask_lab]
                #     y_l_adv = aux_logits_views[0][mask_lab]
                #     y_u = student_out_views[0][~mask_lab]
                #     y_u_adv = aux_logits_views[0][~mask_lab]
                    
                #     worst_case_loss = args.trade_off_worst * worst_case_criterion(y_l, y_l_adv, y_u, y_u_adv)
                # else:
                #     worst_case_loss = torch.tensor(0.0).cuda()  
                
                student_out_views = student_out.chunk(2)
                aux_logits_views = aux_logits.chunk(2)
                
                y_l = student_out_views[0][mask_lab]
                y_l_adv = aux_logits_views[0][mask_lab]
                y_u = student_out_views[0][~mask_lab]
                y_u_adv = aux_logits_views[0][~mask_lab]

                worst_case_loss = args.worst_weight * worst_case_criterion(y_l, y_l_adv, y_u, y_u_adv)
                
                
                
                

                pstr = ''
                pstr += f'cls_loss: {cls_loss.item():.4f} '
                pstr += f'cluster_loss: {cluster_loss.item():.4f} '
                pstr += f'sup_con_loss: {sup_con_loss.item():.4f} '
                pstr += f'contrastive_loss: {contrastive_loss.item():.4f} '
                pstr += f'LSR_loss: {LSR_loss.item():.4f} '
                pstr += f'worst_case_loss: {worst_case_loss.item():.4f} '
                pstr += f'clustering_loss: {total_loss.item():.4f} '
#               

                loss = 0
                loss += (1 - args.sup_weight) * cluster_loss + args.sup_weight * cls_loss
                loss += (1 - args.sup_weight) * contrastive_loss + args.sup_weight * sup_con_loss
                loss += args.LSR_weight * LSR_loss
                loss += worst_case_loss
                loss += args.cluster *total_loss
                
                
                
            # Train acc
            loss_record.update(loss.item(), class_labels.size(0))
            optimizer.zero_grad()
            if fp16_scaler is None:
                loss.backward()
                torch.nn.utils.clip_grad_norm_(student.parameters(), max_norm=100.0)
                optimizer.step()
            else:
                fp16_scaler.scale(loss).backward()
                fp16_scaler.step(optimizer)
                fp16_scaler.update()

            if batch_idx % args.print_freq == 0:
                args.logger.info('Epoch: [{}][{}/{}]\t loss {:.5f}\t {}'
                            .format(epoch, batch_idx, len(train_loader), loss.item(), pstr))
                
        args.logger.info('Train Epoch: {} Avg Loss: {:.4f} '.format(epoch, loss_record.avg))

        args.logger.info('Testing on unlabelled examples in the training data...')
        all_acc, old_acc, new_acc = test(student, unlabelled_train_loader, epoch=epoch, save_name='Train ACC Unlabelled', args=args)

        original_data = np.loadtxt(f"./Results/clusterfinal.txt", delimiter=',')
        new_data = np.append(original_data, [all_acc, old_acc, new_acc])
        np.savetxt(f"./Results/clusterfinal.txt", new_data, fmt='%f', delimiter=',')


        args.logger.info('Train Accuracies: All {:.4f} | Old {:.4f} | New {:.4f}'.format(all_acc, old_acc, new_acc))


        exp_lr_scheduler.step()

        save_dict = {
            'model': student.state_dict(),
            'optimizer': optimizer.state_dict(),
            'epoch': epoch + 1,
        }
        
        

        if old_acc_test > best_test_acc_lab:

            args.logger.info(f'Best ACC on old Classes on disjoint test set: {old_acc_test:.4f}...')
            args.logger.info('Best Train Accuracies: All {:.4f} | Old {:.4f} | New {:.4f}'.format(all_acc, old_acc, new_acc))
            
            torch.save(save_dict, args.model_path[:-3] + f'_best.pt')
            args.logger.info("model saved to {}.".format(args.model_path[:-3] + f'_best.pt'))
            
            # inductive
            best_test_acc_lab = old_acc_test
            # transductive
            best_train_acc_lab = old_acc
            best_train_acc_ubl = new_acc
            best_train_acc_all = all_acc
        
        args.logger.info(f'Exp Name: {args.exp_name}')
        args.logger.info(f'Metrics with best model on test set: All: {best_train_acc_all:.4f} Old: {best_train_acc_lab:.4f} New: {best_train_acc_ubl:.4f}')



def test(model, test_loader, epoch, save_name, args):
    model.eval()

    preds, targets = [], []
    mask = np.array([])
    
    
    for batch_idx, (images, label, _) in enumerate(tqdm(test_loader)):
        images = images.cuda(non_blocking=True)
        with torch.no_grad():
            embedding, logits,_ = model(images)
            preds.append(logits.argmax(1).cpu().numpy())
            targets.append(label.cpu().numpy())
            mask = np.append(mask,
                             np.array([True if x.item() in args.train_classes else False for x in label]))
            
            
    data_embed_npy=torch.cat(data_emb_collect,axis=0).cpu().numpy()
    label_npy=torch.cat(label_collect,axis=0).cpu().numpy()

    
    preds = np.concatenate(preds)
    targets = np.concatenate(targets)


    all_acc, old_acc, new_acc = log_accs_from_preds(y_true=targets, y_pred=preds, mask=mask,
                                                    T=epoch, eval_funcs=args.eval_funcs, save_name=save_name,
                                                    args=args)

    return all_acc, old_acc, new_acc

        

if __name__ == "__main__":
    parser = argparse.ArgumentParser(description='cluster', formatter_class=argparse.ArgumentDefaultsHelpFormatter)
    parser.add_argument('--batch_size', default=128, type=int)
    parser.add_argument('--num_workers', default=8, type=int)
    parser.add_argument('--eval_funcs', nargs='+', help='Which eval functions to use', default=['v2', 'v2p'])

    parser.add_argument('--warmup_model_dir', type=str, default=None)
    parser.add_argument('--dataset_name', type=str, default='rafdb', help='options: cifar10, cifar100, imagenet_100, cub, scars, fgvc_aricraft, herbarium_19')
    parser.add_argument('--prop_train_labels', type=float, default=0.5)
    parser.add_argument('--use_ssb_splits', action='store_true', default=True)

    parser.add_argument('--grad_from_block', type=int, default=11)
    parser.add_argument('--lr', type=float, default=0.1)
    parser.add_argument('--gamma', type=float, default=0.1)
    parser.add_argument('--momentum', type=float, default=0.9)
    parser.add_argument('--weight_decay', type=float, default=5e-5)
    parser.add_argument('--epochs', default=200, type=int)
    parser.add_argument('--exp_root', type=str, default=exp_root)
    parser.add_argument('--transform', type=str, default='imagenet')
    parser.add_argument('--sup_weight', type=float, default=0.35)
    parser.add_argument('--n_views', default=2, type=int)

    parser.add_argument('--memax_weight', type=float, default=2)
    parser.add_argument('--warmup_teacher_temp', default=0.07, type=float, help='Initial value for the teacher temperature.')
    parser.add_argument('--teacher_temp', default=0.04, type=float, help='Final value (after linear warmup)of the teacher temperature.')
    parser.add_argument('--warmup_teacher_temp_epochs', default=30, type=int, help='Number of warmup epochs for the teacher temperature.')

    parser.add_argument('--fp16', action='store_true', default=False)
    parser.add_argument('--print_freq', default=10, type=int)

    parser.add_argument('--exp_name', default=None, type=str)    
    
    #------------------------
    #Aughead
    parser.add_argument('--eta-prime', default=2, type=float,
                        help='weight of adversarial loss on labeled data (default: 2)')
    parser.add_argument('--worst_weight', default=0.2, type=float,
                        help='weight of worst-case estimation loss on unlabeled data (default: 0.2)')
    parser.add_argument('--cluster_epochs', default=50, type=int, help='Number of warmup epochs for the clustering.')
    parser.add_argument('--cluster', default=0.2, type=int, help='weight for the clustering.')
    parser.add_argument('--LSR_weight', default=0.3, type=int, help='weight for the Dynamic LSR.')
    # ----------------------
    # INIT
    # ----------------------
    args = parser.parse_args([])
    device = torch.device('cuda:0')


    args.dataset_name = "ferplus"
    args.train_classes = (0,1,2,3)
    args.unlabeled_classes = (4,5,6,7)
    args.exp_name = "ferplus0123"

#     args.memax_weight = 1
    args.eval_funcs = 'v2'

    args.num_labeled_classes = len(args.train_classes)
    args.num_unlabeled_classes = len(args.unlabeled_classes)

    init_experiment(args, runner_name=['simgcd'])
    args.logger.info(f'Using evaluation function {args.eval_funcs[0]} to print results')

    torch.backends.cudnn.benchmark = True

    # ----------------------
    # BASE MODEL
    # ----------------------
    args.interpolation = 3
    args.crop_pct = 0.875
    
    
    backbone = torch.hub.load('facebookresearch/dino:main', 'dino_vitb16')

    if args.warmup_model_dir is not None:
        args.logger.info(f'Loading weights from {args.warmup_model_dir}')
        backbone.load_state_dict(torch.load(args.warmup_model_dir, map_location='cpu'))

    # NOTE: Hardcoded image size as we do not finetune the entire ViT model
    args.image_size = 224
    args.feat_dim = 768
    args.num_mlp_layers = 3
    args.mlp_out_dim = args.num_labeled_classes + args.num_unlabeled_classes


    # ----------------------
    # HOW MUCH OF BASE MODEL TO FINETUNE
    # ----------------------
    for m in backbone.parameters():
        m.requires_grad = False

    # Only finetune layers from block 'args.grad_from_block' onwards
    for name, m in backbone.named_parameters():
        if 'block' in name:
            block_num = int(name.split('.')[1])
            if block_num >= args.grad_from_block:
                m.requires_grad = True


    args.logger.info('model build')

    # --------------------
    # CONTRASTIVE TRANSFORM
    # --------------------
    train_transform, test_transform = get_transform(args.transform, image_size=args.image_size, args=args)
    train_transform = ContrastiveLearningViewGenerator(base_transform=train_transform, n_views=args.n_views)
    # --------------------
    # DATASETS
    # --------------------


    train_dataset, test_dataset, unlabelled_train_examples_test, datasets = get_RAFDB_datasets(args.dataset_name,
                                                                                               train_transform,
                                                                                               test_transform,
                                                                                               args)

    # --------------------
    # SAMPLER
    # Sampler which balances labelled and unlabelled examples in each batch
    # --------------------
    label_len = len(train_dataset.labelled_dataset)
    unlabelled_len = len(train_dataset.unlabelled_dataset)
    sample_weights = [1 if i < label_len else label_len / unlabelled_len for i in range(len(train_dataset))]
    sample_weights = torch.DoubleTensor(sample_weights)
    sampler = torch.utils.data.WeightedRandomSampler(sample_weights, num_samples=len(train_dataset))

    # --------------------
    # DATALOADERS
    # --------------------
    train_loader = DataLoader(train_dataset, num_workers=args.num_workers, batch_size=args.batch_size, shuffle=False,
                              sampler=sampler, drop_last=True, pin_memory=True)
    
    test_loader_unlabelled = DataLoader(unlabelled_train_examples_test, num_workers=args.num_workers,
                                        batch_size=256, shuffle=False, pin_memory=False)
    
    test_loader_labelled = DataLoader(test_dataset, num_workers=args.num_workers,
                                      batch_size=256, shuffle=False, pin_memory=False)

    # ----------------------
    # PROJECTION HEAD
    # ----------------------
    projector = DINOHeadExtended(in_dim=args.feat_dim, out_dim=args.mlp_out_dim, nlayers=args.num_mlp_layers)
    model = nn.Sequential(backbone, projector).to(device)

    # ----------------------
    # TRAIN
    # ----------------------
    train(model, train_loader, test_loader_labelled, test_loader_unlabelled, args)
#     train(model, train_loader, None, test_loader_unlabelled, args)