Python FileStorageObserver.create方法代碼示例

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def run_single_experiment(dataset: str,
                          savedir: str,
                          named_configs: List,
                          config_updates: Dict[str, Any]):
    from tape.__main__ import proteins

    config_updates.update({
        'training': {'learning_rate': 1e-4, 'use_memory_saving_gradients': True},
        'num_epochs': 1000,
        'steps_per_epoch': 200,
        'tasks': dataset})

    if not os.path.exists(savedir):
        os.mkdir(savedir)
    shutil.rmtree(proteins.observers[0].basedir)
    proteins.observers[0] = FileStorageObserver.create(
        os.path.join(savedir, dataset))

    proteins.run(
        named_configs=named_configs,
        config_updates=config_updates) 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def config():
    locals().update({k: v.default for k, v in inspect.signature(get_enhancer).parameters.items()})

    session_id = 'dev'
    storage_dir: str = None
    database_rttm: str = None
    activity_rttm: str = database_rttm

    job_id = 1
    number_of_jobs = 1

    assert storage_dir is not None, (storage_dir, 'overwrite the storage_dir from the command line')
    assert database_rttm is not None, (database_rttm, 'overwrite the database_rttm from the command line')
    assert activity_rttm is not None, (database_rttm, 'overwrite the activity_rttm from the command line')

    if dlp_mpi.IS_MASTER:
        experiment.observers.append(FileStorageObserver.create(str(
            Path(storage_dir).expanduser().resolve() / 'sacred'
        ))) 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def config():
    chime6 = False
    if chime6:
        locals().update({k: v.default for k, v in inspect.signature(get_enhancer_chime6).parameters.items()})
    else:
        locals().update({k: v.default for k, v in inspect.signature(get_enhancer).parameters.items()})

    session_id = 'dev'
    storage_dir: str = None

    job_id = 1
    number_of_jobs = 1

    assert storage_dir is not None, (storage_dir, 'overwrite the storage_dir from the command line')

    if dlp_mpi.IS_MASTER:
        experiment.observers.append(FileStorageObserver.create(str(
            Path(storage_dir).expanduser().resolve() / 'sacred'
        ))) 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def setup_file_observer():
    file_obs_path = os.path.join(results_path, "sacred")
    logger.info("FileStorageObserver path: {}".format(file_obs_path))
    logger.info("Using the FileStorageObserver in results/sacred")
    ex.observers.append(FileStorageObserver.create(file_obs_path))
    pass 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def single_run(config_updates, rundir, _id):
    for i in range(3):
        try:
            run = single_exp._create_run(config_updates=config_updates)
            observer = FileStorageObserver.create(basedir=rundir)
            run._id = _id
            run.observers = [observer]
            run()
            break
        except TypeError:
            if i < 2:
                print("Run %i failed at start, retrying..." % _id)
            else:
                print("Giving up %i" % _id)
            continue 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def single_run(config_updates, rundir, _id):
    run = single_exp._create_run(config_updates=config_updates)
    observer = FileStorageObserver.create(basedir=rundir)
    run._id = _id
    run.observers = [observer]
    try:
        run()
    except:
        print('Run %i failed' % _id) 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def polish(trial):
    """Load model from checkpoint, then fix the order of the factor
    matrices (using the largest logits), and re-optimize using L-BFGS to find
    the nearest local optima.
    """
    # Hack: create new instance without call __init__, since trainable.__init__
    # creates result_dir and log_dir in the wrong place (~/ray_results)
    trainable_cls = TrainableBP
    trainable = trainable_cls.__new__(trainable_cls)
    trainable._setup(trial.config)
    trainable.restore(str(Path(trial.logdir) / trial._checkpoint.value))
    loss = trainable.polish(N_LBFGS_STEPS, save_to_self_model=True)
    torch.save(trainable.model.state_dict(), str((Path(trial.logdir) / trial._checkpoint.value).parent / 'polished_model.pth'))
    return loss 

# 需要導入模塊: from sacred.observers import FileStorageObserver [as 別名]
# 或者: from sacred.observers.FileStorageObserver import create [as 別名]
def config():
    logdir = 'runs/transcriber-' + datetime.now().strftime('%y%m%d-%H%M%S')
    device = 'cuda' if torch.cuda.is_available() else 'cpu'
    iterations = 500000
    resume_iteration = None
    checkpoint_interval = 1000
    train_on = 'MAESTRO'

    batch_size = 8
    sequence_length = 327680
    model_complexity = 48

    if torch.cuda.is_available() and torch.cuda.get_device_properties(torch.cuda.current_device()).total_memory < 10e9:
        batch_size //= 2
        sequence_length //= 2
        print(f'Reducing batch size to {batch_size} and sequence_length to {sequence_length} to save memory')

    learning_rate = 0.0006
    learning_rate_decay_steps = 10000
    learning_rate_decay_rate = 0.98

    leave_one_out = None

    clip_gradient_norm = 3

    validation_length = sequence_length
    validation_interval = 500

    ex.observers.append(FileStorageObserver.create(logdir))