def build_evaluator(cls, cfg, dataset_name, output_folder=None): """ Create evaluator(s) for a given dataset. This uses the special metadata "evaluator_type" associated with each builtin dataset. For your own dataset, you can simply create an evaluator manually in your script and do not have to worry about the hacky if-else logic here. """ if output_folder is None: output_folder = os.path.join(cfg.OUTPUT_DIR, "inference") evaluator_list = [] evaluator_type = MetadataCatalog.get(dataset_name).evaluator_type if evaluator_type == "lvis": return LVISEvaluator(dataset_name, cfg, True, output_folder) if evaluator_type == "coco": return COCOEvaluator(dataset_name, cfg, True, output_folder) if evaluator_type == "cityscapes": assert ( torch.cuda.device_count() >= comm.get_rank() ), "CityscapesEvaluator currently do not work with multiple machines." return CityscapesEvaluator(dataset_name) if len(evaluator_list) == 0: raise NotImplementedError( "no Evaluator for the dataset {} with the type {}".format( dataset_name, evaluator_type)) if len(evaluator_list) == 1: return evaluator_list[0] return DatasetEvaluators(evaluator_list)
def setup(args): cfg = get_cfg() cfg.merge_from_file(args.config_file) cfg.SOLVER.BASE_LR = 0.001 # Avoid NaNs. Not useful in this script anyway. cfg.merge_from_list(args.opts) cfg.freeze() setup_logger(distributed_rank=comm.get_rank()) return cfg
def setup(args): cfg = get_cfg() add_densepose_config(cfg) cfg.merge_from_file(args.config_file) cfg.merge_from_list(args.opts) cfg.freeze() default_setup(cfg, args) # Setup logger for "densepose" module setup_logger(output=cfg.OUTPUT_DIR, distributed_rank=comm.get_rank(), name="densepose") return cfg
def __init__(self, size: int): """ Args: size (int): the total number of data of the underlying dataset to sample from """ self._size = size assert size > 0 self._rank = comm.get_rank() self._world_size = comm.get_world_size() shard_size = (self._size - 1) // self._world_size + 1 begin = shard_size * self._rank end = min(shard_size * (self._rank + 1), self._size) self._local_indices = range(begin, end)
def default_setup(cfg, args): """ Perform some basic common setups at the beginning of a job, including: 1. Set up the mydl logger 2. Log basic information about environment, cmdline arguments, and config 3. Backup the config to the output directory Args: cfg (CfgNode): the full config to be used args (argparse.NameSpace): the command line arguments to be logged """ output_dir = cfg.OUTPUT_DIR if comm.is_main_process() and output_dir: PathManager.mkdirs(output_dir) rank = comm.get_rank() setup_logger(output_dir, distributed_rank=rank, name="fvcore") logger = setup_logger(output_dir, distributed_rank=rank) logger.info("Rank of current process: {}. World size: {}".format( rank, comm.get_world_size())) logger.info("Environment info:\n" + collect_env_info()) logger.info("Command line arguments: " + str(args)) if hasattr(args, "config_file") and args.config_file != "": logger.info("Contents of args.config_file={}:\n{}".format( args.config_file, PathManager.open(args.config_file, "r").read())) logger.info("Running with full config:\n{}".format(cfg)) if comm.is_main_process() and output_dir: # Note: some of our scripts may expect the existence of # config.yaml in output directory path = os.path.join(output_dir, "config.yaml") with PathManager.open(path, "w") as f: f.write(cfg.dump()) logger.info("Full config saved to {}".format(os.path.abspath(path))) if cfg.VERSION == 2: # make sure each worker has a different, yet deterministic seed if specified seed_all_rng(None if cfg.SEED < 0 else cfg.SEED + rank) # cudnn benchmark has large overhead. It shouldn't be used considering the small size of # typical validation set. if not (hasattr(args, "eval_only") and args.eval_only): torch.backends.cudnn.benchmark = cfg.CUDNN_BENCHMARK
def evaluate(self): """ Returns: dict: has a key "segm", whose value is a dict of "AP" and "AP50". """ comm.synchronize() if comm.get_rank() > 0: return os.environ["CITYSCAPES_DATASET"] = os.path.abspath( os.path.join(self._metadata.gt_dir, "..", "..") ) # Load the Cityscapes eval script *after* setting the required env var, # since the script reads CITYSCAPES_DATASET into global variables at load time. import cityscapesscripts.evaluation.evalInstanceLevelSemanticLabeling as cityscapes_eval self._logger.info("Evaluating results under {} ...".format(self._temp_dir)) # set some global states in cityscapes evaluation API, before evaluating cityscapes_eval.args.predictionPath = os.path.abspath(self._temp_dir) cityscapes_eval.args.predictionWalk = None cityscapes_eval.args.JSONOutput = False cityscapes_eval.args.colorized = False cityscapes_eval.args.gtInstancesFile = os.path.join(self._temp_dir, "gtInstances.json") # These lines are adopted from # https://github.com/mcordts/cityscapesScripts/blob/master/cityscapesscripts/evaluation/evalInstanceLevelSemanticLabeling.py # noqa groundTruthImgList = glob.glob(cityscapes_eval.args.groundTruthSearch) assert len( groundTruthImgList ), "Cannot find any ground truth images to use for evaluation. Searched for: {}".format( cityscapes_eval.args.groundTruthSearch ) predictionImgList = [] for gt in groundTruthImgList: predictionImgList.append(cityscapes_eval.getPrediction(gt, cityscapes_eval.args)) results = cityscapes_eval.evaluateImgLists( predictionImgList, groundTruthImgList, cityscapes_eval.args )["averages"] ret = OrderedDict() ret["segm"] = {"AP": results["allAp"] * 100, "AP50": results["allAp50%"] * 100} self._working_dir.cleanup() return ret
def build_evaluator(cls, cfg, dataset_name, output_folder=None): """ Create evaluator(s) for a given dataset. This uses the special metadata "evaluator_type" associated with each builtin dataset. For your own dataset, you can simply create an evaluator manually in your script and do not have to worry about the hacky if-else logic here. """ if output_folder is None: output_folder = os.path.join(cfg.OUTPUT_DIR, "inference") evaluator_list = [] evaluator_type = MetadataCatalog.get(dataset_name).evaluator_type if evaluator_type in ["sem_seg", "coco_panoptic_seg"]: evaluator_list.append( SemSegEvaluator( dataset_name, distributed=True, num_classes=cfg.MODEL.SEM_SEG_HEAD.NUM_CLASSES, ignore_label=cfg.MODEL.SEM_SEG_HEAD.IGNORE_VALUE, output_dir=output_folder, )) if evaluator_type in ["coco", "coco_panoptic_seg"]: evaluator_list.append( COCOEvaluator(dataset_name, cfg, True, output_folder)) if evaluator_type == "coco_panoptic_seg": evaluator_list.append( COCOPanopticEvaluator(dataset_name, output_folder)) elif evaluator_type == "cityscapes": assert ( torch.cuda.device_count() >= comm.get_rank() ), "CityscapesEvaluator currently do not work with multiple machines." return CityscapesEvaluator(dataset_name) elif evaluator_type == "pascal_voc": return PascalVOCDetectionEvaluator(dataset_name) elif evaluator_type == "lvis": return LVISEvaluator(dataset_name, cfg, True, output_folder) if len(evaluator_list) == 0: raise NotImplementedError( "no Evaluator for the dataset {} with the type {}".format( dataset_name, evaluator_type)) elif len(evaluator_list) == 1: return evaluator_list[0] return DatasetEvaluators(evaluator_list)
def __init__(self, size: int, shuffle: bool = True, seed: Optional[int] = None): """ Args: size (int): the total number of data of the underlying dataset to sample from shuffle (bool): whether to shuffle the indices or not seed (int): the initial seed of the shuffle. Must be the same across all workers. If None, will use a random seed shared among workers (require synchronization among all workers). """ self._size = size assert size > 0 self._shuffle = shuffle if seed is None: seed = comm.shared_random_seed() self._seed = int(seed) self._rank = comm.get_rank() self._world_size = comm.get_world_size()
def __init__(self, dataset_dicts, repeat_thresh, shuffle=True, seed=None): """ Args: dataset_dicts (list[dict]): annotations in mydl dataset format. repeat_thresh (float): frequency threshold below which data is repeated. shuffle (bool): whether to shuffle the indices or not seed (int): the initial seed of the shuffle. Must be the same across all workers. If None, will use a random seed shared among workers (require synchronization among all workers). """ self._shuffle = shuffle if seed is None: seed = comm.shared_random_seed() self._seed = int(seed) self._rank = comm.get_rank() self._world_size = comm.get_world_size() # Get fractional repeat factors and split into whole number (_int_part) # and fractional (_frac_part) parts. rep_factors = self._get_repeat_factors(dataset_dicts, repeat_thresh) self._int_part = torch.trunc(rep_factors) self._frac_part = rep_factors - self._int_part
def train(cfg, local_rank, distributed): model = build_detection_model(cfg) device = torch.device(cfg.MODEL.DEVICE) model.to(device) optimizer = make_optimizer(cfg, model) scheduler = make_lr_scheduler(cfg, optimizer) # Initialize mixed-precision training use_mixed_precision = cfg.DTYPE == "float16" amp_opt_level = 'O1' if use_mixed_precision else 'O0' model, optimizer = amp.initialize(model, optimizer, opt_level=amp_opt_level) if distributed: model = torch.nn.parallel.DistributedDataParallel( model, device_ids=[local_rank], output_device=local_rank, # this should be removed if we update BatchNorm stats broadcast_buffers=False, ) arguments = {} arguments["iteration"] = 0 output_dir = cfg.OUTPUT_DIR save_to_disk = get_rank() == 0 checkpointer = DetectronCheckpointer(cfg, model, optimizer, scheduler, output_dir, save_to_disk) extra_checkpoint_data = checkpointer.load(cfg.MODEL.WEIGHT) arguments.update(extra_checkpoint_data) data_loader = make_data_loader( cfg, is_train=True, is_distributed=distributed, start_iter=arguments["iteration"], ) test_period = cfg.SOLVER.TEST_PERIOD if test_period > 0: data_loader_val = make_data_loader(cfg, is_train=False, is_distributed=distributed, is_for_period=True) else: data_loader_val = None checkpoint_period = cfg.SOLVER.CHECKPOINT_PERIOD do_train( cfg, model, data_loader, data_loader_val, optimizer, scheduler, checkpointer, device, checkpoint_period, test_period, arguments, ) return model
def main(): parser = argparse.ArgumentParser( description="PyTorch Object Detection Training") parser.add_argument( "--config-file", default="", metavar="FILE", help="path to config file", type=str, ) parser.add_argument("--local_rank", type=int, default=0) parser.add_argument( "--skip-test", dest="skip_test", help="Do not test the final model", action="store_true", ) parser.add_argument( "opts", help="Modify config options using the command-line", default=None, nargs=argparse.REMAINDER, ) args = parser.parse_args() num_gpus = int( os.environ["WORLD_SIZE"]) if "WORLD_SIZE" in os.environ else 1 args.distributed = num_gpus > 1 if args.distributed: torch.cuda.set_device(args.local_rank) torch.distributed.init_process_group(backend="nccl", init_method="env://") synchronize() cfg.merge_from_file(args.config_file) cfg.merge_from_list(args.opts) cfg.freeze() output_dir = cfg.OUTPUT_DIR if output_dir: mkdir(output_dir) logger = setup_logger("mydl", output_dir, get_rank()) logger.info("Using {} GPUs".format(num_gpus)) logger.info(args) logger.info("Collecting env info (might take some time)") logger.info("\n" + collect_env_info()) logger.info("Loaded configuration file {}".format(args.config_file)) with open(args.config_file, "r") as cf: config_str = "\n" + cf.read() logger.info(config_str) logger.info("Running with config:\n{}".format(cfg)) output_config_path = os.path.join(cfg.OUTPUT_DIR, 'config.yml') logger.info("Saving config into: {}".format(output_config_path)) # save overloaded model config in the output directory save_config(cfg, output_config_path) model = train(cfg, args.local_rank, args.distributed) if not args.skip_test: run_test(cfg, model, args.distributed)
def main(): parser = argparse.ArgumentParser( description="PyTorch Object Detection Inference") parser.add_argument( "--config-file", default= "/private/home/fmassa/github/detectron.pytorch_v2/configs/e2e_faster_rcnn_R_50_C4_1x_caffe2.yaml", metavar="FILE", help="path to config file", ) parser.add_argument("--local_rank", type=int, default=0) parser.add_argument( "--ckpt", help= "The path to the checkpoint for test, default is the latest checkpoint.", default=None, ) parser.add_argument( "opts", help="Modify config options using the command-line", default=None, nargs=argparse.REMAINDER, ) args = parser.parse_args() num_gpus = int( os.environ["WORLD_SIZE"]) if "WORLD_SIZE" in os.environ else 1 distributed = num_gpus > 1 if distributed: torch.cuda.set_device(args.local_rank) torch.distributed.init_process_group(backend="nccl", init_method="env://") synchronize() cfg.merge_from_file(args.config_file) cfg.merge_from_list(args.opts) cfg.freeze() save_dir = "" logger = setup_logger("mydl", save_dir, get_rank()) logger.info("Using {} GPUs".format(num_gpus)) logger.info(cfg) logger.info("Collecting env info (might take some time)") logger.info("\n" + collect_env_info()) model = build_detection_model(cfg) model.to(cfg.MODEL.DEVICE) # Initialize mixed-precision if necessary use_mixed_precision = cfg.DTYPE == 'float16' amp_handle = amp.init(enabled=use_mixed_precision, verbose=cfg.AMP_VERBOSE) output_dir = cfg.OUTPUT_DIR checkpointer = DetectronCheckpointer(cfg, model, save_dir=output_dir) ckpt = cfg.MODEL.WEIGHT if args.ckpt is None else args.ckpt _ = checkpointer.load(ckpt, use_latest=args.ckpt is None) iou_types = ("bbox", ) if cfg.MODEL.MASK_ON: iou_types = iou_types + ("segm", ) if cfg.MODEL.KEYPOINT_ON: iou_types = iou_types + ("keypoints", ) output_folders = [None] * len(cfg.DATASETS.TEST) dataset_names = cfg.DATASETS.TEST if cfg.OUTPUT_DIR: for idx, dataset_name in enumerate(dataset_names): output_folder = os.path.join(cfg.OUTPUT_DIR, "inference", dataset_name) mkdir(output_folder) output_folders[idx] = output_folder data_loaders_val = make_data_loader(cfg, is_train=False, is_distributed=distributed) for output_folder, dataset_name, data_loader_val in zip( output_folders, dataset_names, data_loaders_val): inference( model, data_loader_val, dataset_name=dataset_name, iou_types=iou_types, box_only=False if cfg.MODEL.RETINANET_ON else cfg.MODEL.RPN_ONLY, bbox_aug=cfg.TEST.BBOX_AUG.ENABLED, device=cfg.MODEL.DEVICE, expected_results=cfg.TEST.EXPECTED_RESULTS, expected_results_sigma_tol=cfg.TEST.EXPECTED_RESULTS_SIGMA_TOL, output_folder=output_folder, ) synchronize()