Python models.build_model方法代码示例

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def main():
    init_env('1')
    loaders = make_data_loaders(cfg)
    model = build_model(cfg)
    model = model.cuda()
    task_name = 'base_unet'
    log_dir = os.path.join(cfg.LOG_DIR, task_name)
    cfg.TASK_NAME = task_name
    mkdir(log_dir)
    logger = setup_logger('train', log_dir, filename='train.log')
    logger.info(cfg)
    logger = setup_logger('eval', log_dir, filename='eval.log')
    optimizer, scheduler = make_optimizer(cfg, model)
    metrics = get_metrics(cfg)
    losses = get_losses(cfg)
    train_val(model, loaders, optimizer, scheduler, losses, metrics) 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def main():
    config = configure()
    task = tasks.load_task(config)
    model = models.build_model(config.model, config.opt)

    for i_epoch in range(config.opt.iters):

        train_loss, train_acc, _ = \
                do_iter(task.train, model, config, train=True)
        val_loss, val_acc, val_predictions = \
                do_iter(task.val, model, config, vis=True)
        test_loss, test_acc, test_predictions = \
                do_iter(task.test, model, config)

        logging.info(
                "%5d  |  %8.3f  %8.3f  %8.3f  |  %8.3f  %8.3f  %8.3f",
                i_epoch,
                train_loss, val_loss, test_loss,
                train_acc, val_acc, test_acc)

        with open("logs/val_predictions_%d.json" % i_epoch, "w") as pred_f:
            print >>pred_f, json.dumps(val_predictions)

        #with open("logs/test_predictions_%d.json" % i_epoch, "w") as pred_f:
        #    print >>pred_f, json.dumps(test_predictions) 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def main():
    N = 50
    D = 20

    settings = ExperimentSettings()
    settings.max_rank=2
    settings.gaussian_auto_ard = False
    settings.constant_gaussian_std = 1.0
    settings.constant_noise_std = 0.1
    
    #X = np.float32(np.random.randn(N, D))

    m = build_model(('lowrank', ('chain', 'g'), 'g'), (N, D), settings)
    #m = build_model(('chain', 'g'), (N, D), settings)
    X = m.sample()
    #X /= np.std(X)

    best_structure = do_structure_search(X, settings) 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def __init__(self, model_path, gpu_id=0):
        from models import build_model
        from data_loader import get_dataloader
        from post_processing import get_post_processing
        from utils import get_metric
        self.gpu_id = gpu_id
        if self.gpu_id is not None and isinstance(self.gpu_id, int) and torch.cuda.is_available():
            self.device = torch.device("cuda:%s" % self.gpu_id)
            torch.backends.cudnn.benchmark = True
        else:
            self.device = torch.device("cpu")
        checkpoint = torch.load(model_path, map_location=torch.device('cpu'))
        config = checkpoint['config']
        config['arch']['backbone']['pretrained'] = False

        self.validate_loader = get_dataloader(config['dataset']['validate'], config['distributed'])

        self.model = build_model(config['arch'].pop('type'), **config['arch'])
        self.model.load_state_dict(checkpoint['state_dict'])
        self.model.to(self.device)

        self.post_process = get_post_processing(config['post_processing'])
        self.metric_cls = get_metric(config['metric']) 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def score_model(structure, X, settings):
    N, D = X.shape
    m = build_model(structure, (N, D), settings)
    m.observe(X)

    jm = Model(m)

    jm.train(silent=True,
             stopping_rule=settings.stopping_rule,
             adam_rate=settings.adam_rate)
    score = jm.monte_carlo_elbo(n_samples=settings.n_elbo_samples)

    return score 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def train(config):
    # load train data
    print("start load data")
    train_data_df = load_data_from_csv(os.path.join(config.data_dir, config.file_names[0]))
    validate_data_df = load_data_from_csv(os.path.join(config.data_dir, config.file_names[1]))
    # explore data
    print("explore train data!")
    explore_data_analysis(train_data_df)
    print("explore dev data!")
    explore_data_analysis(validate_data_df)

    content_train = train_data_df.iloc[:, 0]

    content_val = validate_data_df.iloc[:, 0]

    if config.write_vocab:
        write_vocab(content_train, os.path.join(config.data_dir, config.file_prefix + 'vocab.data'), min_count=5)

    print("start convert str2id!")
    word2id = load_vocab(os.path.join(config.data_dir, config.file_prefix + 'vocab.data'))
    train_data = list(map(lambda x: string2id(x, word2id), content_train))
    print("train_data的长度", len(train_data))
    val_data = list(map(lambda x: string2id(x, word2id), content_val))

    print("create experiment dir")

    config = prepare_experiment(config, len(word2id), len(train_data_df))

    set_logger(config)
    train_label = train_data_df.iloc[:, 1]
    val_label = validate_data_df.iloc[:, 1]
    train_set = DataSet(config.batch_size, train_data, train_label, config.sequence_length)
    dev_set = DataSet(config.batch_size, val_data, val_label, config.sequence_length)
    print("-----start train  model------")
    model = build_model(config)
    train_module(model, config, train_set, dev_set)
    print("finish train %s model") 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def __init__(self, model_path, post_p_thre=0.7, gpu_id=None):
        '''
        初始化pytorch模型
        :param model_path: 模型地址(可以是模型的参数或者参数和计算图一起保存的文件)
        :param gpu_id: 在哪一块gpu上运行
        '''
        self.gpu_id = gpu_id

        if self.gpu_id is not None and isinstance(self.gpu_id, int) and torch.cuda.is_available():
            self.device = torch.device("cuda:%s" % self.gpu_id)
        else:
            self.device = torch.device("cpu")
        print('device:', self.device)
        checkpoint = torch.load(model_path, map_location=self.device)

        config = checkpoint['config']
        config['arch']['backbone']['pretrained'] = False
        self.model = build_model(config['arch'].pop('type'), **config['arch'])
        self.post_process = get_post_processing(config['post_processing'])
        self.post_process.box_thresh = post_p_thre
        self.img_mode = config['dataset']['train']['dataset']['args']['img_mode']
        self.model.load_state_dict(checkpoint['state_dict'])
        self.model.to(self.device)
        self.model.eval()

        self.transform = []
        for t in config['dataset']['train']['dataset']['args']['transforms']:
            if t['type'] in ['ToTensor', 'Normalize']:
                self.transform.append(t)
        self.transform = get_transforms(self.transform) 

# 需要导入模块: import models [as 别名]
# 或者: from models import build_model [as 别名]
def main(config):
    import torch
    from models import build_model, build_loss
    from data_loader import get_dataloader
    from trainer import Trainer
    from post_processing import get_post_processing
    from utils import get_metric
    if torch.cuda.device_count() > 1:
        torch.cuda.set_device(args.local_rank)
        torch.distributed.init_process_group(backend="nccl", init_method="env://", world_size=torch.cuda.device_count(), rank=args.local_rank)
        config['distributed'] = True
    else:
        config['distributed'] = False
    config['local_rank'] = args.local_rank

    train_loader = get_dataloader(config['dataset']['train'], config['distributed'])
    assert train_loader is not None
    if 'validate' in config['dataset']:
        validate_loader = get_dataloader(config['dataset']['validate'], False)
    else:
        validate_loader = None

    criterion = build_loss(config['loss']).cuda()

    config['arch']['backbone']['in_channels'] = 3 if config['dataset']['train']['dataset']['args']['img_mode'] != 'GRAY' else 1
    model = build_model(config['arch'])

    post_p = get_post_processing(config['post_processing'])
    metric = get_metric(config['metric'])

    trainer = Trainer(config=config,
                      model=model,
                      criterion=criterion,
                      train_loader=train_loader,
                      post_process=post_p,
                      metric_cls=metric,
                      validate_loader=validate_loader)
    trainer.train()