code
stringlengths
101
5.91M
def choose_devices(target_abi, target_ids): device_clazz = device_class(target_abi) devices = device_clazz.list_devices() if (target_ids == 'all'): run_devices = devices elif (target_ids == 'random'): unlocked_devices = [dev for dev in devices if (not util.is_device_locked(dev))] if unlocked_devices: run_devices = [random.choice(unlocked_devices)] else: run_devices = [random.choice(devices)] else: device_id_list = [dev.strip() for dev in target_ids.split(',')] run_devices = [dev for dev in device_id_list if (dev in devices)] return run_devices
def set_ro(ro): from ..util._optional_deps import _APY_LOADED if _APY_LOADED: from astropy import units if (_APY_LOADED and isinstance(ro, units.Quantity)): ro = ro.to(units.kpc).value __config__.set('normalization', 'ro', str(ro))
class ExternalProcess(object): _ACTION = 1 _RESET = 2 _CLOSE = 3 _ATTRIBUTE = 4 _TRANSITION = 5 _OBSERV = 6 _EXCEPTION = 7 _VALUE = 8 def __init__(self, constructor): (self._conn, conn) = multiprocessing.Pipe() self._process = multiprocessing.Process(target=self._worker, args=(constructor, conn)) atexit.register(self.close) self._process.start() self._observ_space = None self._action_space = None def observation_space(self): if (not self._observ_space): self._observ_space = self.__getattr__('observation_space') return self._observ_space def action_space(self): if (not self._action_space): self._action_space = self.__getattr__('action_space') return self._action_space def __getattr__(self, name): self._conn.send((self._ATTRIBUTE, name)) return self._receive(self._VALUE) def step(self, action, blocking=True): self._conn.send((self._ACTION, action)) if blocking: return self._receive(self._TRANSITION) else: return functools.partial(self._receive, self._TRANSITION) def reset(self, blocking=True): self._conn.send((self._RESET, None)) if blocking: return self._receive(self._OBSERV) else: return functools.partial(self._receive, self._OBSERV) def close(self): try: self._conn.send((self._CLOSE, None)) self._conn.close() except IOError: pass self._process.join() def _receive(self, expected_message): (message, payload) = self._conn.recv() if (message == self._EXCEPTION): stacktrace = payload raise Exception(stacktrace) if (message == expected_message): return payload raise KeyError('Received message of unexpected type {}'.format(message)) def _worker(self, constructor, conn): try: env = constructor() while True: try: if (not conn.poll(0.1)): continue (message, payload) = conn.recv() except (EOFError, KeyboardInterrupt): break if (message == self._ACTION): action = payload conn.send((self._TRANSITION, env.step(action))) continue if (message == self._RESET): assert (payload is None) conn.send((self._OBSERV, env.reset())) continue if (message == self._ATTRIBUTE): name = payload conn.send((self._VALUE, getattr(env, name))) continue if (message == self._CLOSE): assert (payload is None) break raise KeyError('Received message of unknown type {}'.format(message)) except Exception: stacktrace = ''.join(traceback.format_exception(*sys.exc_info())) conn.send((self._EXCEPTION, stacktrace)) tf.logging.error('Error in environment process: {}'.format(stacktrace)) conn.close()
def test(config, test_dataset, testloader, model, sv_dir='./', sv_pred=True): model.eval() with torch.no_grad(): for (_, batch) in enumerate(tqdm(testloader)): (image, size, name) = batch size = size[0] pred = test_dataset.single_scale_inference(config, model, image.cuda()) if ((pred.size()[(- 2)] != size[0]) or (pred.size()[(- 1)] != size[1])): pred = F.interpolate(pred, size[(- 2):], mode='bilinear', align_corners=config.MODEL.ALIGN_CORNERS) if sv_pred: sv_path = os.path.join(sv_dir, 'test_results') if (not os.path.exists(sv_path)): os.mkdir(sv_path) test_dataset.save_pred(pred, sv_path, name)
def get_xdensenet_cifar(num_classes, blocks, growth_rate, bottleneck, expand_ratio=2, model_name=None, pretrained=False, root=os.path.join('~', '.torch', 'models'), **kwargs): assert (num_classes in [10, 100]) if bottleneck: assert (((blocks - 4) % 6) == 0) layers = ([((blocks - 4) // 6)] * 3) else: assert (((blocks - 4) % 3) == 0) layers = ([((blocks - 4) // 3)] * 3) init_block_channels = (2 * growth_rate) from functools import reduce channels = reduce((lambda xi, yi: (xi + [reduce((lambda xj, yj: (xj + [(xj[(- 1)] + yj)])), ([growth_rate] * yi), [(xi[(- 1)][(- 1)] // 2)])[1:]])), layers, [[(init_block_channels * 2)]])[1:] net = CIFARXDenseNet(channels=channels, init_block_channels=init_block_channels, num_classes=num_classes, bottleneck=bottleneck, expand_ratio=expand_ratio, **kwargs) if pretrained: if ((model_name is None) or (not model_name)): raise ValueError('Parameter `model_name` should be properly initialized for loading pretrained model.') from .model_store import download_model download_model(net=net, model_name=model_name, local_model_store_dir_path=root) return net
class IntensiveReader(BaseReader): name: str = 'intensive' def postprocess(self, output: EvalLoopOutput, eval_examples: datasets.Dataset, eval_dataset: datasets.Dataset, log_level: int=logging.WARNING, mode: str='evaluate') -> Union[(List[Dict[(str, Any)]], EvalPrediction)]: (predictions, nbest_json, scores_diff_json) = self.compute_predictions(eval_examples, eval_dataset, output.predictions, version_2_with_negative=self.data_args.version_2_with_negative, n_best_size=self.data_args.n_best_size, max_answer_length=self.data_args.max_answer_length, null_score_diff_threshold=self.data_args.null_score_diff_threshold, output_dir=self.args.output_dir, log_level=log_level, n_tops=(self.data_args.start_n_top, self.data_args.end_n_top)) if (mode == 'retro_inference'): return (nbest_json, scores_diff_json) if self.data_args.version_2_with_negative: formatted_predictions = [{'id': k, 'prediction_text': v, 'no_answer_probability': scores_diff_json[k]} for (k, v) in predictions.items()] else: formatted_predictions = [{'id': k, 'prediction_text': v} for (k, v) in predictions.items()] if (mode == 'predict'): return formatted_predictions else: references = [{'id': ex[C.ID_COLUMN_NAME], 'answers': ex[C.ANSWER_COLUMN_NAME]} for ex in eval_examples] return EvalPrediction(predictions=formatted_predictions, label_ids=references) def compute_predictions(self, examples: datasets.Dataset, features: datasets.Dataset, predictions: Tuple[(np.ndarray, np.ndarray)], version_2_with_negative: bool=False, n_best_size: int=20, max_answer_length: int=30, null_score_diff_threshold: float=0.0, output_dir: Optional[str]=None, log_level: Optional[int]=logging.WARNING, n_tops: Tuple[(int, int)]=((- 1), (- 1)), use_choice_logits: bool=False): if (len(predictions) not in [2, 3]): raise ValueError('`predictions` should be a tuple with two or three elements (start_logits, end_logits, choice_logits).') (all_start_logits, all_end_logits) = predictions[:2] all_choice_logits = None if (len(predictions) == 3): all_choice_logits = predictions[(- 1)] example_id_to_index = {k: i for (i, k) in enumerate(examples[C.ID_COLUMN_NAME])} features_per_example = collections.defaultdict(list) for (i, feature) in enumerate(features): features_per_example[example_id_to_index[feature['example_id']]].append(i) all_predictions = collections.OrderedDict() all_nbest_json = collections.OrderedDict() scores_diff_json = (collections.OrderedDict() if version_2_with_negative else None) logger.setLevel(log_level) logger.info(f'Post-processing {len(examples)} example predictions split into {len(features)} features.') for (example_index, example) in enumerate(tqdm(examples)): feature_indices = features_per_example[example_index] min_null_prediction = None prelim_predictions = [] for feature_index in feature_indices: start_logits = all_start_logits[feature_index] end_logits = all_end_logits[feature_index] feature_null_score = (start_logits[0] + end_logits[0]) if (all_choice_logits is not None): choice_logits = all_choice_logits[feature_index] if use_choice_logits: feature_null_score = choice_logits[1] offset_mapping = features[feature_index]['offset_mapping'] token_is_max_context = features[feature_index].get('token_is_max_context', None) if ((min_null_prediction is None) or (min_null_prediction['score'] > feature_null_score)): min_null_prediction = {'offsets': (0, 0), 'score': feature_null_score, 'start_logit': start_logits[0], 'end_logit': end_logits[0]} start_indexes = np.argsort(start_logits)[(- 1):((- n_best_size) - 1):(- 1)].tolist() end_indexes = np.argsort(end_logits)[(- 1):((- n_best_size) - 1):(- 1)].tolist() for start_index in start_indexes: for end_index in end_indexes: if ((start_index >= len(offset_mapping)) or (end_index >= len(offset_mapping)) or (not offset_mapping[start_index]) or (not offset_mapping[end_index])): continue if ((end_index < start_index) or (((end_index - start_index) + 1) > max_answer_length)): continue if ((token_is_max_context is not None) and (not token_is_max_context.get(str(start_index), False))): continue prelim_predictions.append({'offsets': (offset_mapping[start_index][0], offset_mapping[end_index][1]), 'score': (start_logits[start_index] + end_logits[end_index]), 'start_logit': start_logits[start_index], 'end_logit': end_logits[end_index]}) if version_2_with_negative: prelim_predictions.append(min_null_prediction) null_score = min_null_prediction['score'] predictions = sorted(prelim_predictions, key=(lambda x: x['score']), reverse=True)[:n_best_size] if (version_2_with_negative and (not any(((p['offsets'] == (0, 0)) for p in predictions)))): predictions.append(min_null_prediction) context = example['context'] for pred in predictions: offsets = pred.pop('offsets') pred['text'] = context[offsets[0]:offsets[1]] if ((len(predictions) == 0) or ((len(predictions) == 1) and (predictions[0]['text'] == ''))): predictions.insert(0, {'text': '', 'start_logit': 0.0, 'end_logit': 0.0, 'score': 0.0}) scores = np.array([pred.pop('score') for pred in predictions]) exp_scores = np.exp((scores - np.max(scores))) probs = (exp_scores / exp_scores.sum()) for (prob, pred) in zip(probs, predictions): pred['probability'] = prob if (not version_2_with_negative): all_predictions[example[C.ID_COLUMN_NAME]] = predictions[0]['text'] else: i = 0 try: while (predictions[i]['text'] == ''): i += 1 except: i = 0 best_non_null_pred = predictions[i] score_diff = ((null_score - best_non_null_pred['start_logit']) - best_non_null_pred['end_logit']) scores_diff_json[example[C.ID_COLUMN_NAME]] = float(score_diff) if (score_diff > null_score_diff_threshold): all_predictions[example[C.ID_COLUMN_NAME]] = '' else: all_predictions[example[C.ID_COLUMN_NAME]] = best_non_null_pred['text'] all_nbest_json[example[C.ID_COLUMN_NAME]] = [{k: (float(v) if isinstance(v, (np.float16, np.float32, np.float64)) else v) for (k, v) in pred.items()} for pred in predictions] if (output_dir is not None): if (not os.path.isdir(output_dir)): raise EnvironmentError(f'{output_dir} is not a directory.') prediction_file = os.path.join(output_dir, C.INTENSIVE_PRED_FILE_NAME) nbest_file = os.path.join(output_dir, C.NBEST_PRED_FILE_NAME) if version_2_with_negative: null_odds_file = os.path.join(output_dir, C.SCORE_DIFF_FILE_NAME) logger.info(f'Saving predictions to {prediction_file}.') with open(prediction_file, 'w') as writer: writer.write((json.dumps(all_predictions, indent=4) + '\n')) logger.info(f'Saving nbest_preds to {nbest_file}.') with open(nbest_file, 'w') as writer: writer.write((json.dumps(all_nbest_json, indent=4) + '\n')) if version_2_with_negative: logger.info(f'Saving null_odds to {null_odds_file}.') with open(null_odds_file, 'w') as writer: writer.write((json.dumps(scores_diff_json, indent=4) + '\n')) return (all_predictions, all_nbest_json, scores_diff_json)
def gen_CNN(channels, conv=nn.Conv1d, bias=True, activation=nn.ReLU, batch_norm=None, instance_norm=None): layers = [] for i in range((len(channels) - 1)): (in_size, out_size) = channels[i:(i + 2)] layers.append(conv(in_size, out_size, 1, bias=bias)) if (batch_norm is not None): layers.append(batch_norm(out_size)) if (activation is not None): layers.append(activation(inplace=True)) if (instance_norm is not None): layers.append(instance_norm(out_size, affine=False, track_running_stats=False)) return nn.Sequential(*layers)
def main(args): if (args.seed is not None): print('* absolute seed: {}'.format(args.seed)) random.seed(args.seed) torch.manual_seed(args.seed) torch.cuda.manual_seed(args.seed) cudnn.deterministic = True warnings.warn('You have chosen to seed training. This will turn on the CUDNN deterministic setting, which can slow down your training considerably! You may see unexpected behavior when restarting from checkpoints.') is_train = (True if (not args.evaluate) else False) (train_loader, val_loader, num_classes) = make_data_loader(args, is_train=is_train) model = get_model(num_classes, args) criterion = torch.nn.MultiLabelSoftMarginLoss() trainer = Trainer(model, criterion, train_loader, val_loader, args) if is_train: trainer.train() else: trainer.validate()
def ControlTypeChange(choice): if (choice == 'pos'): return gr.update(visible=False) elif (choice == 'sentiment'): return gr.update(visible=True)
def test_scatter(): input = torch.zeros([1, 3, 3, 3]) output = scatter(input=input, devices=[(- 1)]) assert torch.allclose(input, output) inputs = [torch.zeros([1, 3, 3, 3]), torch.zeros([1, 4, 4, 4])] outputs = scatter(input=inputs, devices=[(- 1)]) for (input, output) in zip(inputs, outputs): assert torch.allclose(input, output) if torch.cuda.is_available(): input = torch.zeros([1, 3, 3, 3]) output = scatter(input=input, devices=[0]) assert torch.allclose(input.cuda(), output) inputs = [torch.zeros([1, 3, 3, 3]), torch.zeros([1, 4, 4, 4])] outputs = scatter(input=inputs, devices=[0]) for (input, output) in zip(inputs, outputs): assert torch.allclose(input.cuda(), output) with pytest.raises(Exception): scatter(5, [(- 1)])
def short_hash(name): if (name not in _model_sha256): raise ValueError('Pretrained model for {name} is not available.'.format(name=name)) return _model_sha256[name][:8]
class TFWorkerWrapper(Worker): def __init__(self): self._inner_worker = None self._sess = None self._sess_entered = None self.worker_init() def worker_init(self): self._sess = tf.compat.v1.get_default_session() if (not self._sess): self._sess = tf.compat.v1.Session() self._sess_entered = True self._sess.__enter__() def shutdown(self): self._inner_worker.shutdown() if (tf.compat.v1.get_default_session() and self._sess_entered): self._sess_entered = False self._sess.__exit__(None, None, None) def agent(self): return self._inner_worker.agent def agent(self, agent): self._inner_worker.agent = agent def env(self): return self._inner_worker.env def env(self, env): self._inner_worker.env = env def update_agent(self, agent_update): self._inner_worker.update_agent(agent_update) def update_env(self, env_update): self._inner_worker.update_env(env_update) def rollout(self): return self._inner_worker.rollout() def start_rollout(self): self._inner_worker.start_rollout() def step_rollout(self): return self._inner_worker.step_rollout() def collect_rollout(self): return self._inner_worker.collect_rollout()
def get_latest_scene(s3_scene_jsons): scenes = [open_remote_pb_object(scene_json, Scene) for scene_json in s3_scene_jsons] creation_ts = [_s.creation_date.ToMicroseconds() for _s in scenes] index = creation_ts.index(max(creation_ts)) return (scenes[index], s3_scene_jsons[index])
class SparkEvaluator(Evaluator[S]): def __init__(self, processes: int=8): self.spark_conf = SparkConf().setAppName('jmetalpy').setMaster(f'local[{processes}]') self.spark_context = SparkContext(conf=self.spark_conf) logger = self.spark_context._jvm.org.apache.log4j logger.LogManager.getLogger('org').setLevel(logger.Level.WARN) def evaluate(self, solution_list: List[S], problem: Problem) -> List[S]: solutions_to_evaluate = self.spark_context.parallelize(solution_list) return solutions_to_evaluate.map((lambda s: problem.evaluate(s))).collect()
class BasicBlock(nn.Module): expansion = 1 def __init__(self, inplanes, planes, stride=1, downsample=None, cardinality=1, base_width=64, reduce_first=1, dilation=1, first_dilation=None, act_layer=nn.ReLU, norm_layer=nn.BatchNorm2d, attn_layer=None, aa_layer=None, drop_block=None, drop_path=None): super(BasicBlock, self).__init__() assert (cardinality == 1), 'BasicBlock only supports cardinality of 1' first_planes = (planes // reduce_first) outplanes = (planes * self.expansion) first_dilation = (first_dilation or dilation) use_aa = ((aa_layer is not None) and ((stride == 2) or (first_dilation != dilation))) self.conv1 = nn.Conv2d(inplanes, first_planes, kernel_size=3, stride=(1 if use_aa else stride), padding=first_dilation, dilation=first_dilation, bias=False) self.bn1 = norm_layer(first_planes) self.act1 = act_layer(inplace=True) self.aa = (aa_layer(channels=first_planes, stride=stride) if use_aa else None) self.conv2 = nn.Conv2d(first_planes, outplanes, kernel_size=3, padding=dilation, dilation=dilation, bias=False) self.bn2 = norm_layer(outplanes) self.attn_layer = attn_layer if (attn_layer == 'our_se'): self.se = SE_Our(outplanes, outplanes) else: self.se = create_attn(attn_layer, outplanes) self.act2 = act_layer(inplace=True) self.downsample = downsample self.stride = stride self.dilation = dilation self.drop_block = drop_block self.drop_path = drop_path def zero_init_last_bn(self): nn.init.zeros_(self.bn2.weight) def forward(self, x): residual = x x = self.conv1(x) x = self.bn1(x) if (self.drop_block is not None): x = self.drop_block(x) x = self.act1(x) if (self.aa is not None): x = self.aa(x) x = self.conv2(x) x = self.bn2(x) if (self.drop_block is not None): x = self.drop_block(x) if (self.downsample is not None): residual = self.downsample(residual) if (self.attn_layer == 'our_se'): residual = (residual * self.se(residual)) elif (self.se is not None): x = self.se(x) if (self.drop_path is not None): x = self.drop_path(x) x += residual x = self.act2(x) return x
def _fused_batch_norm(inputs, decay=0.999, center=True, scale=False, epsilon=0.001, activation_fn=None, param_initializers=None, param_regularizers=None, updates_collections=ops.GraphKeys.UPDATE_OPS, is_training=True, reuse=None, variables_collections=None, outputs_collections=None, trainable=True, data_format=DATA_FORMAT_NHWC, zero_debias_moving_mean=False, scope=None): if (data_format not in (DATA_FORMAT_NCHW, DATA_FORMAT_NHWC)): raise ValueError('data_format has to be either NCHW or NHWC.') with variable_scope.variable_scope(scope, 'BatchNorm', [inputs], reuse=reuse) as sc: inputs = ops.convert_to_tensor(inputs) original_shape = inputs.get_shape() original_inputs = inputs original_rank = original_shape.ndims if (original_rank is None): raise ValueError(('Inputs %s has undefined rank' % inputs.name)) elif (original_rank not in [2, 4]): raise ValueError(('Inputs %s has unsupported rank. Expected 2 or 4 but got %d' % (inputs.name, original_rank))) if (original_rank == 2): channels = inputs.get_shape()[(- 1)].value if (channels is None): raise ValueError('`C` dimension must be known but is None') new_shape = [(- 1), 1, 1, channels] if (data_format == DATA_FORMAT_NCHW): new_shape = [(- 1), channels, 1, 1] inputs = array_ops.reshape(inputs, new_shape) inputs_shape = inputs.get_shape() if (data_format == DATA_FORMAT_NHWC): params_shape = inputs_shape[(- 1):] else: params_shape = inputs_shape[1:2] if (not params_shape.is_fully_defined()): raise ValueError(('Inputs %s has undefined `C` dimension %s.' % (inputs.name, params_shape))) beta_collections = utils.get_variable_collections(variables_collections, 'beta') variable_dtype = dtypes.float32 if (not param_initializers): param_initializers = {} if (not param_regularizers): param_regularizers = {} beta_regularizer = param_regularizers.get('beta') gamma_regularizer = param_regularizers.get('gamma') if center: beta_initializer = param_initializers.get('beta', init_ops.zeros_initializer()) beta = variables.model_variable('beta', shape=params_shape, dtype=variable_dtype, initializer=beta_initializer, regularizer=beta_regularizer, collections=beta_collections, trainable=trainable) else: beta = array_ops.constant(0.0, dtype=variable_dtype, shape=params_shape) if scale: gamma_collections = utils.get_variable_collections(variables_collections, 'gamma') gamma_initializer = param_initializers.get('gamma', init_ops.ones_initializer()) gamma = variables.model_variable('gamma', shape=params_shape, dtype=variable_dtype, initializer=gamma_initializer, regularizer=gamma_regularizer, collections=gamma_collections, trainable=trainable) else: gamma = array_ops.constant(1.0, dtype=variable_dtype, shape=params_shape) with variable_scope.variable_scope(variable_scope.get_variable_scope()) as local_scope: local_scope.set_partitioner(None) moving_mean_collections = utils.get_variable_collections(variables_collections, 'moving_mean') moving_mean_initializer = param_initializers.get('moving_mean', init_ops.zeros_initializer()) moving_mean = variables.model_variable('moving_mean', shape=params_shape, dtype=variable_dtype, initializer=moving_mean_initializer, trainable=False, collections=moving_mean_collections) moving_variance_collections = utils.get_variable_collections(variables_collections, 'moving_variance') moving_variance_initializer = param_initializers.get('moving_variance', init_ops.ones_initializer()) moving_variance = variables.model_variable('moving_variance', shape=params_shape, dtype=variable_dtype, initializer=moving_variance_initializer, trainable=False, collections=moving_variance_collections) def _fused_batch_norm_training(): return nn.fused_batch_norm(inputs, gamma, beta, epsilon=epsilon, data_format=data_format) def _fused_batch_norm_inference(): return nn.fused_batch_norm(inputs, gamma, beta, mean=moving_mean, variance=moving_variance, epsilon=epsilon, is_training=False, data_format=data_format) (outputs, mean, variance) = utils.smart_cond(is_training, _fused_batch_norm_training, _fused_batch_norm_inference) is_training_value = utils.constant_value(is_training) need_updates = ((is_training_value is None) or is_training_value) if need_updates: if (updates_collections is None): no_updates = (lambda : outputs) def _force_updates(): update_moving_mean = moving_averages.assign_moving_average(moving_mean, mean, decay, zero_debias=zero_debias_moving_mean) update_moving_variance = moving_averages.assign_moving_average(moving_variance, variance, decay, zero_debias=False) with ops.control_dependencies([update_moving_mean, update_moving_variance]): return array_ops.identity(outputs) outputs = utils.smart_cond(is_training, _force_updates, no_updates) else: moving_vars_fn = (lambda : (moving_mean, moving_variance)) def _delay_updates(): update_moving_mean = moving_averages.assign_moving_average(moving_mean, mean, decay, zero_debias=zero_debias_moving_mean) update_moving_variance = moving_averages.assign_moving_average(moving_variance, variance, decay, zero_debias=False) return (update_moving_mean, update_moving_variance) (update_mean, update_variance) = utils.smart_cond(is_training, _delay_updates, moving_vars_fn) ops.add_to_collections(updates_collections, update_mean) ops.add_to_collections(updates_collections, update_variance) outputs.set_shape(inputs_shape) if (original_shape.ndims == 2): outputs = array_ops.reshape(outputs, array_ops.shape(original_inputs)) if (activation_fn is not None): outputs = activation_fn(outputs) return utils.collect_named_outputs(outputs_collections, sc.name, outputs)
def _create_local(name, shape, collections=None, validate_shape=True, dtype=tf.float32): collections = list((collections or [])) collections += [ops.GraphKeys.LOCAL_VARIABLES] return variables.Variable(initial_value=array_ops.zeros(shape, dtype=dtype), name=name, trainable=False, collections=collections, validate_shape=validate_shape)
def rank_by_significance(embeddings, class_embeddings): similarities = cosine_similarity_embeddings(embeddings, class_embeddings) significance_score = [np.max(softmax(similarity)) for similarity in similarities] significance_ranking = {i: r for (r, i) in enumerate(np.argsort((- np.array(significance_score))))} return significance_ranking
def is_torchdynamo_available(): if (not is_torch_available()): return False try: import torch._dynamo as dynamo return True except Exception: return False
def blend_images(image, image2, should_blend=0, alpha=0.5, scope=None): with tf.name_scope(scope, 'blend_images', [image, image2]): if (should_blend == 0): return image else: image = tf.py_func(blend_images_np, [image, image2, alpha]) return image
def parse_args(input_args=None): parser = argparse.ArgumentParser(description='Simple example of a training script.') parser.add_argument('--pretrained_model_name_or_path', type=str, default=None, required=True, help='Path to pretrained model or model identifier from huggingface.co/models.') parser.add_argument('--revision', type=str, default=None, required=False, help='Revision of pretrained model identifier from huggingface.co/models.') parser.add_argument('--tokenizer_name', type=str, default=None, help='Pretrained tokenizer name or path if not the same as model_name') parser.add_argument('--instance_data_dir', type=str, default=None, required=True, help='A folder containing the training data of instance images.') parser.add_argument('--class_data_dir', type=str, default=None, required=False, help='A folder containing the training data of class images.') parser.add_argument('--instance_prompt', type=str, default='a photo of sks dog', required=False, help='The prompt with identifier specifying the instance') parser.add_argument('--class_prompt', type=str, default=None, help='The prompt to specify images in the same class as provided instance images.') parser.add_argument('--with_prior_preservation', default=False, action='store_true', help='Flag to add prior preservation loss.') parser.add_argument('--prior_loss_weight', type=float, default=1.0, help='The weight of prior preservation loss.') parser.add_argument('--num_class_images', type=int, default=100, help='Minimal class images for prior preservation loss. If there are not enough images already present in class_data_dir, additional images will be sampled with class_prompt.') parser.add_argument('--output_dir', type=str, default='text-inversion-model', help='The output directory where the model predictions and checkpoints will be written.') parser.add_argument('--seed', type=int, default=None, help='A seed for reproducible training.') parser.add_argument('--resolution', type=int, default=512, help='The resolution for input images, all the images in the train/validation dataset will be resized to this resolution') parser.add_argument('--placement', type=str, default='cpu', help='Placement Policy for Gemini. Valid when using colossalai as dist plan.') parser.add_argument('--center_crop', default=False, action='store_true', help='Whether to center crop the input images to the resolution. If not set, the images will be randomly cropped. The images will be resized to the resolution first before cropping.') parser.add_argument('--train_batch_size', type=int, default=4, help='Batch size (per device) for the training dataloader.') parser.add_argument('--sample_batch_size', type=int, default=4, help='Batch size (per device) for sampling images.') parser.add_argument('--num_train_epochs', type=int, default=1) parser.add_argument('--max_train_steps', type=int, default=None, help='Total number of training steps to perform. If provided, overrides num_train_epochs.') parser.add_argument('--save_steps', type=int, default=500, help='Save checkpoint every X updates steps.') parser.add_argument('--gradient_checkpointing', action='store_true', help='Whether or not to use gradient checkpointing to save memory at the expense of slower backward pass.') parser.add_argument('--learning_rate', type=float, default=5e-06, help='Initial learning rate (after the potential warmup period) to use.') parser.add_argument('--scale_lr', action='store_true', default=False, help='Scale the learning rate by the number of GPUs, gradient accumulation steps, and batch size.') parser.add_argument('--lr_scheduler', type=str, default='constant', help='The scheduler type to use. Choose between ["linear", "cosine", "cosine_with_restarts", "polynomial", "constant", "constant_with_warmup"]') parser.add_argument('--lr_warmup_steps', type=int, default=500, help='Number of steps for the warmup in the lr scheduler.') parser.add_argument('--use_8bit_adam', action='store_true', help='Whether or not to use 8-bit Adam from bitsandbytes.') parser.add_argument('--max_grad_norm', default=1.0, type=float, help='Max gradient norm.') parser.add_argument('--push_to_hub', action='store_true', help='Whether or not to push the model to the Hub.') parser.add_argument('--hub_token', type=str, default=None, help='The token to use to push to the Model Hub.') parser.add_argument('--hub_model_id', type=str, default=None, help='The name of the repository to keep in sync with the local `output_dir`.') parser.add_argument('--logging_dir', type=str, default='logs', help='[TensorBoard]( log directory. Will default to *output_dir/runs/**CURRENT_DATETIME_HOSTNAME***.') parser.add_argument('--mixed_precision', type=str, default=None, choices=['no', 'fp16', 'bf16'], help='Whether to use mixed precision. Choose between fp16 and bf16 (bfloat16). Bf16 requires PyTorch >= 1.10.and an Nvidia Ampere GPU. Default to the value of accelerate config of the current system or the flag passed with the `accelerate.launch` command. Use this argument to override the accelerate config.') parser.add_argument('--local_rank', type=int, default=(- 1), help='For distributed training: local_rank') if (input_args is not None): args = parser.parse_args(input_args) else: args = parser.parse_args() env_local_rank = int(os.environ.get('LOCAL_RANK', (- 1))) if ((env_local_rank != (- 1)) and (env_local_rank != args.local_rank)): args.local_rank = env_local_rank if args.with_prior_preservation: if (args.class_data_dir is None): raise ValueError('You must specify a data directory for class images.') if (args.class_prompt is None): raise ValueError('You must specify prompt for class images.') else: if (args.class_data_dir is not None): logger.warning('You need not use --class_data_dir without --with_prior_preservation.') if (args.class_prompt is not None): logger.warning('You need not use --class_prompt without --with_prior_preservation.') return args
def accuracy(logits, labels): (_, indices) = torch.max(logits, dim=1) if (len(indices.shape) > 1): indices = indices.view((- 1)) labels = labels.view((- 1)) correct = torch.sum((indices == labels)) return ((correct.item() * 1.0) / len(labels))
def mobilenet_v2(pretrained: bool=False, include_top: bool=False, freeze: bool=False): model = torchvision.models.mobilenet_v2(pretrained) if freeze: set_parameter_requires_grad(model, 'classifier') if (not include_top): output_size = model.classifier[1].in_features model.classifier = nn.Identity() return BackboneModule(model, output_size) else: return model
class NonMaximaSuppression2d(nn.Module): def __init__(self, kernel_size: Tuple[(int, int)]): super(NonMaximaSuppression2d, self).__init__() self.kernel_size: Tuple[(int, int)] = kernel_size self.padding: Tuple[(int, int, int, int)] = self._compute_zero_padding2d(kernel_size) self.kernel = _get_nms_kernel2d(*kernel_size) def _compute_zero_padding2d(kernel_size: Tuple[(int, int)]) -> Tuple[(int, int, int, int)]: assert isinstance(kernel_size, tuple), type(kernel_size) assert (len(kernel_size) == 2), kernel_size def pad(x): return ((x - 1) // 2) (ky, kx) = kernel_size return (pad(ky), pad(ky), pad(kx), pad(kx)) def forward(self, x: torch.Tensor) -> torch.Tensor: assert (len(x.shape) == 4), x.shape (B, CH, H, W) = x.size() max_non_center = F.conv2d(F.pad(x, list(self.padding)[::(- 1)], mode='replicate'), self.kernel.repeat(CH, 1, 1, 1).to(x.device, x.dtype), stride=1, groups=CH).view(B, CH, (- 1), H, W).max(dim=2)[0] mask = (x > max_non_center) return (x * mask.to(x.dtype))
def _fake_roi_head(with_shared_head=False): if (not with_shared_head): roi_head = Config(dict(type='StandardRoIHead', bbox_roi_extractor=dict(type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=7, sampling_ratio=0), out_channels=1, featmap_strides=[4, 8, 16, 32]), bbox_head=dict(type='Shared2FCBBoxHead', in_channels=1, fc_out_channels=1, num_classes=4), mask_roi_extractor=dict(type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=14, sampling_ratio=0), out_channels=1, featmap_strides=[4, 8, 16, 32]), mask_head=dict(type='FCNMaskHead', num_convs=1, in_channels=1, conv_out_channels=1, num_classes=4), train_cfg=dict(assigner=dict(type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=True, ignore_iof_thr=(- 1)), sampler=dict(type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=(- 1), add_gt_as_proposals=True), mask_size=28, pos_weight=(- 1), debug=False), test_cfg=dict(score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100, mask_thr_binary=0.5))) else: roi_head = Config(dict(type='StandardRoIHead', shared_head=dict(type='ResLayer', depth=50, stage=3, stride=2, dilation=1, style='caffe', norm_cfg=dict(type='BN', requires_grad=False), norm_eval=True), bbox_roi_extractor=dict(type='SingleRoIExtractor', roi_layer=dict(type='RoIAlign', output_size=14, sampling_ratio=0), out_channels=1, featmap_strides=[16]), bbox_head=dict(type='BBoxHead', with_avg_pool=True, in_channels=2048, roi_feat_size=7, num_classes=4), mask_roi_extractor=None, mask_head=dict(type='FCNMaskHead', num_convs=0, in_channels=2048, conv_out_channels=1, num_classes=4), train_cfg=dict(assigner=dict(type='MaxIoUAssigner', pos_iou_thr=0.5, neg_iou_thr=0.5, min_pos_iou=0.5, match_low_quality=False, ignore_iof_thr=(- 1)), sampler=dict(type='RandomSampler', num=512, pos_fraction=0.25, neg_pos_ub=(- 1), add_gt_as_proposals=True), mask_size=14, pos_weight=(- 1), debug=False), test_cfg=dict(score_thr=0.05, nms=dict(type='nms', iou_threshold=0.5), max_per_img=100, mask_thr_binary=0.5))) return roi_head
def resnet50s16(pretrained=False, finetune_layers=(), s16_feats=('layer4',), s8_feats=('layer2',), s4_feats=('layer1',), **kwargs): model = ResNetS16(finetune_layers, s16_feats, s8_feats, s4_feats, Bottleneck, [3, 4, 6, 3], **kwargs) if pretrained: model.load_state_dict(model_zoo.load_url(model_urls['resnet50'], model_dir=config['nn_weights_path'])) return model
def logging(s, log_path, print_=True, log_=True): if print_: print(s) if log_: with open(log_path, 'a+') as f_log: f_log.write((s + '\n'))
class ConfigDense(NamedTuple): seed: int = 0 floatx: Any = 'float64' jitter: float = 1e-06 num_test: int = 128 num_cond: int = 32 num_samples: int = 16384 shard_size: int = 1024 input_dims: int = 3 kernel_variance: float = 0.9 rel_lengthscales_min: float = 0.05 rel_lengthscales_max: float = 0.5 noise_variance: float = 0.01 def error_tol(self): return (4 * (self.num_samples ** (- 0.5)))
class TestVisualization(TestCase): def test_import_should_okay(self): try: from bigdl.nano.automl.hpo.visualization import plot_optimization_history except ImportError: self.fail('cannot import plot_optimization_history from nano.aotoml.hpo.visualization.') try: from bigdl.nano.automl.hpo.visualization import plot_parallel_coordinate except ImportError: self.fail('cannot import plot_parallel_coordinate from nano.aotoml.hpo.visualization.') try: from bigdl.nano.automl.hpo.visualization import plot_intermediate_values except ImportError: self.fail('cannot import plot_intermediate_values from nano.aotoml.hpo.visualization.') try: from bigdl.nano.automl.hpo.visualization import plot_contour except ImportError: self.fail('cannot import plot_contour from nano.aotoml.hpo.visualization.') try: from bigdl.nano.automl.hpo.visualization import plot_param_importances except ImportError: self.fail('cannot import plot_param_importances from nano.aotoml.hpo.visualization.')
class _IndexToTokenDefaultDict(_NamespaceDependentDefaultDict): def __init__(self, non_padded_namespaces: Set[str], padding_token: str, oov_token: str) -> None: super(_IndexToTokenDefaultDict, self).__init__(non_padded_namespaces, (lambda : {0: padding_token, 1: oov_token}), (lambda : {}))
def format_if_possible(format, value): try: return (format % value) except: return ('%s' % value)
def extract_valid_amino_acid(m, amino_acids): ms = SplitMolByPDBResidues(m) valid_ms = [ms[k] for k in ms.keys()] ret_m = None for i in range(len(valid_ms)): if (i == 0): ret_m = valid_ms[0] else: ret_m = CombineMols(ret_m, valid_ms[i]) return ret_m
class TestTPProfile(unittest.TestCase): def test_isothermal(self): profile = Profile(num_profile_heights=130) profile.set_isothermal(1300) self.assertEqual(len(profile.pressures), 130) self.assertEqual(len(profile.temperatures), 130) self.assertTrue(np.all((profile.temperatures == 1300))) def test_parametric(self): profile = Profile() P0 = np.min(profile.pressures) T0 = 1300 P1 = 0.001 alpha1 = 0.3 alpha2 = 0.5 P3 = 10000.0 T3 = 2000 (P2, T2) = profile.set_parametric(T0, P1, alpha1, alpha2, P3, T3) self.assertTrue((abs((P3 - (P2 * np.exp((alpha2 * ((T3 - T2) ** 0.5)))))) < (0.001 * P3))) T1 = (((np.log((P1 / P0)) ** 2) / (alpha1 ** 2)) + T0) self.assertTrue((abs((P1 - (P0 * np.exp((alpha1 * ((T1 - T0) ** 0.5)))))) < (0.001 * P1))) self.assertTrue((abs((P1 - (P2 * np.exp(((- alpha2) * ((T1 - T2) ** 0.5)))))) < (0.001 * P1))) def test_set_opacity(self): p = Profile() p.set_isothermal(1200) calc = EclipseDepthCalculator() (wavelengths, depths, info_dict) = calc.compute_depths(p, R_sun, M_jup, R_jup, 5700, full_output=True) p.set_from_opacity(1700, info_dict) self.assertTrue(np.all((p.temperatures > 0))) self.assertTrue(np.all((~ np.isnan(p.temperatures)))) def test_radiative_solution(self): p = Profile() p.set_from_radiative_solution(5040, (0.756 * R_sun), (0.031 * AU), (0.885 * M_jup), R_jup, 1, np.log10(0.003), np.log10(0.158), np.log10(0.158), 0.5, 100) is_upper_atm = np.logical_and((p.pressures > 0.1), (p.pressures < 1000.0)) self.assertTrue(np.all((p.temperatures[is_upper_atm] > 1000))) self.assertTrue(np.all((p.temperatures[is_upper_atm] < 1100))) is_lower_atm = np.logical_and((p.pressures > 100000.0), (p.pressures < 3000000.0)) self.assertTrue(np.all((p.temperatures[is_lower_atm] > 1600))) self.assertTrue(np.all((p.temperatures[is_lower_atm] < 1700))) self.assertTrue(np.all((np.diff(p.temperatures) > 0)))
_module() class Mask2Former(MaskFormer): 'Implementation of `Masked-attention Mask\n Transformer for Universal Image Segmentation\n < def __init__(self, backbone, neck=None, panoptic_head=None, panoptic_fusion_head=None, train_cfg=None, test_cfg=None, init_cfg=None): super().__init__(backbone, neck=neck, panoptic_head=panoptic_head, panoptic_fusion_head=panoptic_fusion_head, train_cfg=train_cfg, test_cfg=test_cfg, init_cfg=init_cfg)
def uniform(lower: float, upper: float) -> 'tune.sample.Float': return tune.uniform(lower, upper)
class NiftiEvaluator(Evaluator): def __init__(self, *args, **kwargs): self.test_nifti = None self.reference_nifti = None super(NiftiEvaluator, self).__init__(*args, **kwargs) def set_test(self, test): if (test is not None): self.test_nifti = sitk.ReadImage(test) super(NiftiEvaluator, self).set_test(sitk.GetArrayFromImage(self.test_nifti)) else: self.test_nifti = None super(NiftiEvaluator, self).set_test(test) def set_reference(self, reference): if (reference is not None): self.reference_nifti = sitk.ReadImage(reference) super(NiftiEvaluator, self).set_reference(sitk.GetArrayFromImage(self.reference_nifti)) else: self.reference_nifti = None super(NiftiEvaluator, self).set_reference(reference) def evaluate(self, test=None, reference=None, voxel_spacing=None, **metric_kwargs): if (voxel_spacing is None): voxel_spacing = np.array(self.test_nifti.GetSpacing())[::(- 1)] metric_kwargs['voxel_spacing'] = voxel_spacing return super(NiftiEvaluator, self).evaluate(test, reference, **metric_kwargs)
def test_chrono_system_clock(): date1 = m.test_chrono1() date2 = datetime.datetime.today() assert isinstance(date1, datetime.datetime) diff = abs((date1 - date2)) assert (diff.days == 0) assert (diff.seconds == 0) assert (diff.microseconds < 500000)
def reduce_loss_dict(loss_dict): world_size = get_world_size() if (world_size < 2): return loss_dict with torch.no_grad(): loss_names = [] all_losses = [] for k in sorted(loss_dict.keys()): loss_names.append(k) all_losses.append(loss_dict[k]) all_losses = torch.stack(all_losses, dim=0) torch.distributed.reduce(all_losses, dst=0) if (torch.distributed.get_rank() == 0): all_losses /= world_size reduced_losses = {k: v for (k, v) in zip(loss_names, all_losses)} return reduced_losses
class EZ_agent(): def __init__(self, args, logger): self.args = args self.lr = args.lr self.noise_dim = self.args.noise_dim self.state_shape = self.args.state_shape self.policy = Policy(args) self.optimizer = optim.Adam(self.policy.parameters(), lr=self.lr) self.entropy_scaling = args.entropy_scaling self.uniform_distrib = torch.distributions.one_hot_categorical.OneHotCategorical(torch.tensor([(1 / self.args.noise_dim) for _ in range(self.args.noise_dim)]).repeat(self.args.batch_size_run, 1)) self.buffer = deque(maxlen=self.args.bandit_buffer) self.epsilon_floor = args.bandit_epsilon self.logger = logger def sample(self, state, test_mode): if test_mode: return self.uniform_distrib.sample() else: probs = self.policy(state) m = torch.distributions.one_hot_categorical.OneHotCategorical(probs) action = m.sample().cpu() return action def update_returns(self, states, actions, returns, test_mode, t): if test_mode: return for (s, a, r) in zip(states, actions, returns): self.buffer.append((s, a, torch.tensor(r, dtype=torch.float))) for _ in range(self.args.bandit_iters): idxs = np.random.randint(0, len(self.buffer), size=self.args.bandit_batch) batch_elems = [self.buffer[i] for i in idxs] states_ = torch.stack([x[0] for x in batch_elems]).to(states.device) actions_ = torch.stack([x[1] for x in batch_elems]).to(states.device) returns_ = torch.stack([x[2] for x in batch_elems]).to(states.device) probs = self.policy(states_) m = torch.distributions.one_hot_categorical.OneHotCategorical(probs) log_probs = m.log_prob(actions_.to(probs.device)) self.optimizer.zero_grad() policy_loss = ((- torch.dot(log_probs, torch.tensor(returns_, device=log_probs.device).float())) + (self.entropy_scaling * log_probs.sum())) policy_loss.backward() self.optimizer.step() mean_entropy = m.entropy().mean() self.logger.log_stat('bandit_entropy', mean_entropy.item(), t) def cuda(self): self.policy.cuda() def save_model(self, path): torch.save(self.policy.state_dict(), '{}/ez_bandit_policy.th'.format(path))
def create_bio_labels(text, opinions): offsets = [l[0] for l in tk.span_tokenize(text)] columns = ['Source', 'Target', 'Polar_expression'] labels = {c: (['O'] * len(offsets)) for c in columns} anns = {c: [] for c in columns} for o in opinions: try: anns['Source'].extend(get_bio_holder(o)) except: pass try: anns['Target'].extend(get_bio_target(o)) except: pass try: anns['Polar_expression'].extend(get_bio_expression(o)) except: pass for c in columns: for (bidx, tags) in anns[c]: labels[c] = replace_with_labels(labels[c], offsets, bidx, tags) labels[c] = restart_orphans(labels[c]) return labels
class SingletonMeter(meter.Meter): def __init__(self, maxlen=1): super(SingletonMeter, self).__init__() self.__val = None def reset(self): old_val = self.__val self.__val = None return old_val def add(self, value): self.__val = value def value(self): return self.__val
def mouse_release(event): global fixed_left, fixed_right, fixed_top, fixed_bottom, root currentx = (root.winfo_pointerx() - root.winfo_rootx()) currenty = (root.winfo_pointery() - root.winfo_rooty()) for frame in root.winfo_children(): frame.grid_forget() if (move_fixed_area == True): width_move = (currentx - locationx) height_move = (currenty - locationy) global step if (step == 0): width_move = 143 height_move = (- 130) elif (step == 1): width_move = (341 - 334) height_move = (303 - 127) elif (step == 2): width_move = (80 - 341) height_move = ((101 - 303) - 20) step += 1 fixed_left += width_move fixed_right += width_move fixed_top += height_move fixed_bottom += height_move main()
def ze_grad(pred, target, classes, gpu): pred_grad = torch.zeros_like(pred).to(gpu) target_grad = torch.zeros_like(target).to(gpu) pred_argm = torch.argmax(pred, 1) target_argm = torch.argmax(target, 1) for i in range(target.shape[0]): pred_grad[(i, pred_argm[i])] += 1.0 target_grad[(i, target_argm[i])] += 1.0 return (pred_grad - target_grad)
def test_audio_dataset_init_reproducible(fs, mocker): dataset_a = audio_dataset(fs, mocker) dataset_b = AudioDataset(TEST_DATA_DIR, TEST_META_FILE, TEST_SAMPLE_RATE, TEST_NUM_SAMPLES) assert (dataset_a.file_list == dataset_b.file_list)
def _variable_assign(var, new_value): return state_ops.assign(var, new_value, name=(var.op.name + '_assign'))
def main(): args = parse_args() cfg = Config.fromfile(args.config) if (args.cfg_options is not None): cfg.merge_from_dict(args.cfg_options) if cfg.get('custom_imports', None): from mmcv.utils import import_modules_from_strings import_modules_from_strings(**cfg['custom_imports']) if cfg.get('cudnn_benchmark', False): torch.backends.cudnn.benchmark = True if (args.work_dir is not None): cfg.work_dir = args.work_dir elif (cfg.get('work_dir', None) is None): cfg.work_dir = osp.join('./work_dirs', osp.splitext(osp.basename(args.config))[0]) if (args.resume_from is not None): cfg.resume_from = args.resume_from if (args.gpu_ids is not None): cfg.gpu_ids = args.gpu_ids else: cfg.gpu_ids = (range(1) if (args.gpus is None) else range(args.gpus)) if (args.launcher == 'none'): distributed = False else: distributed = True init_dist(args.launcher, **cfg.dist_params) (_, world_size) = get_dist_info() cfg.gpu_ids = range(world_size) mmcv.mkdir_or_exist(osp.abspath(cfg.work_dir)) cfg.dump(osp.join(cfg.work_dir, osp.basename(args.config))) timestamp = time.strftime('%Y%m%d_%H%M%S', time.localtime()) log_file = osp.join(cfg.work_dir, f'{timestamp}.log') logger = get_root_logger(log_file=log_file, log_level=cfg.log_level) meta = dict() env_info_dict = collect_env() env_info = '\n'.join([f'{k}: {v}' for (k, v) in env_info_dict.items()]) dash_line = (('-' * 60) + '\n') logger.info((((('Environment info:\n' + dash_line) + env_info) + '\n') + dash_line)) meta['env_info'] = env_info meta['config'] = cfg.pretty_text logger.info(f'Distributed training: {distributed}') logger.info(f'''Config: {cfg.pretty_text}''') if (args.seed is not None): logger.info(f'Set random seed to {args.seed}, deterministic: {args.deterministic}') set_random_seed(args.seed, deterministic=args.deterministic) cfg.seed = args.seed meta['seed'] = args.seed meta['exp_name'] = osp.basename(args.config) model = build_detector(cfg.model, train_cfg=cfg.train_cfg, test_cfg=cfg.test_cfg) pretrained_dict = torch.load(args.pre_train) model_dict = model.state_dict() pretrained_dict = pretrained_dict['state_dict'] model_dict.update(pretrained_dict) model.load_state_dict(model_dict) print('pretrained model loaded from {}.'.format(args.pre_train)) for p in model.neck.parameters(): p.requires_grad = False for p in model.bbox_head.parameters(): p.requires_grad = False print(' trained param ') for (name, param) in model.named_parameters(): if param.requires_grad: print(name) print(' trained param END') datasets = [build_dataset(cfg.data.train)] if (len(cfg.workflow) == 2): val_dataset = copy.deepcopy(cfg.data.val) val_dataset.pipeline = cfg.data.train.pipeline datasets.append(build_dataset(val_dataset)) if (cfg.checkpoint_config is not None): cfg.checkpoint_config.meta = dict(mmdet_version=(__version__ + get_git_hash()[:7]), CLASSES=datasets[0].CLASSES) model.CLASSES = datasets[0].CLASSES train_detector(model, datasets, cfg, distributed=distributed, validate=(not args.no_validate), timestamp=timestamp, meta=meta)
def save_model(model, args, save_dir, model_name, should_print=True): save_path = ('%s/model_%s' % (save_dir, str(model_name))) save_dict = {'model': model.state_dict(), 'args': args} torch.save(save_dict, save_path) if should_print: print(('Model saved to: %s' % save_path))
class MockS3Client(): def __init__(self, enable_mc=True): self.enable_mc = enable_mc def Get(self, filepath): with open(filepath, 'rb') as f: content = f.read() return content
class CLAM_SB(nn.Module): def __init__(self, gate=True, size_arg='small', dropout=False, k_sample=8, n_classes=2, instance_loss_fn=nn.CrossEntropyLoss(), subtyping=False): super(CLAM_SB, self).__init__() self.size_dict = {'small': [1024, 512, 256], 'big': [1024, 512, 384]} size = self.size_dict[size_arg] fc = [nn.Linear(size[0], size[1]), nn.ReLU()] if dropout: fc.append(nn.Dropout(0.25)) if gate: attention_net = Attn_Net_Gated(L=size[1], D=size[2], dropout=dropout, n_classes=1) else: attention_net = Attn_Net(L=size[1], D=size[2], dropout=dropout, n_classes=1) fc.append(attention_net) self.attention_net = nn.Sequential(*fc) self.classifiers = nn.Linear(size[1], n_classes) instance_classifiers = [nn.Linear(size[1], 2) for i in range(n_classes)] self.instance_classifiers = nn.ModuleList(instance_classifiers) self.k_sample = k_sample self.instance_loss_fn = instance_loss_fn self.n_classes = n_classes self.subtyping = subtyping initialize_weights(self) def relocate(self): device = torch.device(('cuda' if torch.cuda.is_available() else 'cpu')) self.attention_net = self.attention_net.to(device) self.classifiers = self.classifiers.to(device) self.instance_classifiers = self.instance_classifiers.to(device) def create_positive_targets(length, device): return torch.full((length,), 1, device=device, dtype=torch.long) def create_negative_targets(length, device): return torch.full((length,), 0, device=device, dtype=torch.long) def inst_eval(self, A, h, classifier): device = h.device if (len(A.shape) == 1): A = A.view(1, (- 1)) top_p_ids = torch.topk(A, self.k_sample)[1][(- 1)] top_p = torch.index_select(h, dim=0, index=top_p_ids) top_n_ids = torch.topk((- A), self.k_sample, dim=1)[1][(- 1)] top_n = torch.index_select(h, dim=0, index=top_n_ids) p_targets = self.create_positive_targets(self.k_sample, device) n_targets = self.create_negative_targets(self.k_sample, device) all_targets = torch.cat([p_targets, n_targets], dim=0) all_instances = torch.cat([top_p, top_n], dim=0) logits = classifier(all_instances) all_preds = torch.topk(logits, 1, dim=1)[1].squeeze(1) instance_loss = self.instance_loss_fn(logits, all_targets) return (instance_loss, all_preds, all_targets) def inst_eval_out(self, A, h, classifier): device = h.device if (len(A.shape) == 1): A = A.view(1, (- 1)) top_p_ids = torch.topk(A, self.k_sample)[1][(- 1)] top_p = torch.index_select(h, dim=0, index=top_p_ids) p_targets = self.create_negative_targets(self.k_sample, device) logits = classifier(top_p) p_preds = torch.topk(logits, 1, dim=1)[1].squeeze(1) instance_loss = self.instance_loss_fn(logits, p_targets) return (instance_loss, p_preds, p_targets) def forward(self, h, label=None, instance_eval=False, return_features=False, attention_only=False, train_epoch=(- 1), testing=False, instance_mask=None): (A, h) = self.attention_net(h) A = torch.transpose(A, 1, 0) if attention_only: return A A_raw = A A = F.softmax(A, dim=1) if instance_eval: total_inst_loss = 0.0 all_preds = [] all_targets = [] inst_labels = F.one_hot(label, num_classes=self.n_classes).squeeze() for i in range(len(self.instance_classifiers)): inst_label = inst_labels[i].item() classifier = self.instance_classifiers[i] if (inst_label == 1): (instance_loss, preds, targets) = self.inst_eval(A, h, classifier) all_preds.extend(preds.cpu().numpy()) all_targets.extend(targets.cpu().numpy()) elif self.subtyping: (instance_loss, preds, targets) = self.inst_eval_out(A, h, classifier) all_preds.extend(preds.cpu().numpy()) all_targets.extend(targets.cpu().numpy()) else: continue total_inst_loss += instance_loss if self.subtyping: total_inst_loss /= len(self.instance_classifiers) M = torch.mm(A, h) logits = self.classifiers(M) Y_hat = torch.topk(logits, 1, dim=1)[1] Y_prob = F.softmax(logits, dim=1) if instance_eval: results_dict = {'instance_loss': total_inst_loss, 'inst_labels': np.array(all_targets), 'inst_preds': np.array(all_preds)} else: results_dict = {'instance_loss': torch.tensor([0]).cuda(), 'inst_labels': np.array([0]), 'inst_preds': np.array([0])} if return_features: results_dict.update({'features': M}) return (logits, Y_prob, Y_hat, A_raw, results_dict)
class ShuffleUnit(nn.Module): def __init__(self, in_channels, out_channels, groups=3, first_block=True, combine='add', conv_cfg=None, norm_cfg=dict(type='BN'), act_cfg=dict(type='ReLU'), with_cp=False): norm_cfg = copy.deepcopy(norm_cfg) act_cfg = copy.deepcopy(act_cfg) super().__init__() self.in_channels = in_channels self.out_channels = out_channels self.first_block = first_block self.combine = combine self.groups = groups self.bottleneck_channels = (self.out_channels // 4) self.with_cp = with_cp if (self.combine == 'add'): self.depthwise_stride = 1 self._combine_func = self._add assert (in_channels == out_channels), 'in_channels must be equal to out_channels when combine is add' elif (self.combine == 'concat'): self.depthwise_stride = 2 self._combine_func = self._concat self.out_channels -= self.in_channels self.avgpool = nn.AvgPool2d(kernel_size=3, stride=2, padding=1) else: raise ValueError(f'Cannot combine tensors with {self.combine}. Only "add" and "concat" are supported') self.first_1x1_groups = (1 if first_block else self.groups) self.g_conv_1x1_compress = ConvModule(in_channels=self.in_channels, out_channels=self.bottleneck_channels, kernel_size=1, groups=self.first_1x1_groups, conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=act_cfg) self.depthwise_conv3x3_bn = ConvModule(in_channels=self.bottleneck_channels, out_channels=self.bottleneck_channels, kernel_size=3, stride=self.depthwise_stride, padding=1, groups=self.bottleneck_channels, conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=None) self.g_conv_1x1_expand = ConvModule(in_channels=self.bottleneck_channels, out_channels=self.out_channels, kernel_size=1, groups=self.groups, conv_cfg=conv_cfg, norm_cfg=norm_cfg, act_cfg=None) self.act = build_activation_layer(act_cfg) def _add(x, out): return (x + out) def _concat(x, out): return torch.cat((x, out), 1) def forward(self, x): def _inner_forward(x): residual = x out = self.g_conv_1x1_compress(x) out = self.depthwise_conv3x3_bn(out) if (self.groups > 1): out = channel_shuffle(out, self.groups) out = self.g_conv_1x1_expand(out) if (self.combine == 'concat'): residual = self.avgpool(residual) out = self.act(out) out = self._combine_func(residual, out) else: out = self._combine_func(residual, out) out = self.act(out) return out if (self.with_cp and x.requires_grad): out = cp.checkpoint(_inner_forward, x) else: out = _inner_forward(x) return out
def make_line_magic(flow_: 'NotebookFlow'): line_magic_names = [name for (name, val) in globals().items() if inspect.isfunction(val)] def _handle(cmd, line): cmd = cmd.replace('-', '_') if (cmd in ('enable', 'disable', 'on', 'off')): return toggle_dataflow(cmd) elif (cmd in ('deps', 'show_deps', 'show_dependency', 'show_dependencies')): return show_deps(line) elif (cmd in ('code', 'get_code')): return get_code(line) elif (cmd in ('waiting', 'show_waiting')): return show_waiting(line) elif (cmd == 'trace_messages'): return trace_messages(line) elif (cmd in ('hls', 'nohls', 'highlight', 'highlights')): return set_highlights(cmd, line) elif (cmd in ('dag', 'make_dag', 'cell_dag', 'make_cell_dag')): return json.dumps(create_dag_metadata(), indent=2) elif (cmd in ('slice', 'make_slice', 'gather_slice')): return make_slice(line) elif (cmd == 'tag'): return tag(line) elif (cmd == 'show_tags'): return show_tags(line) elif (cmd in ('mode', 'exec_mode')): return set_exec_mode(line) elif (cmd in ('schedule', 'exec_schedule', 'execution_schedule')): return set_exec_schedule(line) elif (cmd in ('direction', 'flow_direction', 'order', 'flow_order', 'semantics', 'flow_semantics')): return set_flow_direction(line) elif (cmd == 'reactivity'): return set_reactivity(line) elif (cmd in ('register', 'register_tracer')): return register_tracer(line) elif (cmd in ('deregister', 'deregister_tracer')): return deregister_tracer(line) elif (cmd == 'clear'): flow_.min_timestamp = flow_.cell_counter() return None elif cmd.endswith('warn_ooo'): flow_.mut_settings.warn_out_of_order_usages = (not cmd.startswith('no')) return None elif cmd.endswith('lint_ooo'): flow_.mut_settings.lint_out_of_order_usages = (not cmd.startswith('no')) return None elif (cmd == 'syntax_transforms'): is_on = line.endswith(('enabled', 'on')) is_off = line.endswith(('disabled', 'off')) if (is_on or is_off): flow_.mut_settings.syntax_transforms_enabled = is_on return None elif (cmd == 'syntax_transforms_only'): flow_.mut_settings.syntax_transforms_only = True return None elif cmd.startswith('register_annotation'): return register_annotations(line) elif (cmd == 'toggle_reactivity'): flow_.toggle_reactivity() return None elif (cmd == 'bump_min_forced_reactive_counter'): flow_.bump_min_forced_reactive_counter() return None elif (cmd in line_magic_names): warn(f'We have a magic for {cmd}, but have not yet registered it') return None else: warn(_USAGE) return None def _flow_magic(line: str): try: (cmd, line) = line.split(' ', 1) if (cmd in ('slice', 'make_slice', 'gather_slice')): line = re.sub("--tag +<class '(\\w+)'>", '--tag $\\1', line) except ValueError: (cmd, line) = (line, '') try: (line, fname) = line.split('>', 1) except ValueError: (line, fname) = (line, None) line = line.strip() if (fname is not None): fname = fname.strip() outstr = _handle(cmd, line) if (outstr is None): return if (fname is None): print_(outstr) else: with open(fname, 'w') as f: f.write(outstr) _flow_magic.__name__ = _FLOW_LINE_MAGIC return register_line_magic(_flow_magic)
def get_model_from_config(model_config: ConfigDict, nelec: Array, ion_pos: Array, ion_charges: Array, dtype=jnp.float32) -> Module: spin_split = get_spin_split(nelec) compute_input_streams = get_compute_input_streams_from_config(model_config.input_streams, ion_pos) backflow = get_backflow_from_config(model_config.backflow, spin_split, dtype=dtype) (kernel_init_constructor, bias_init_constructor) = _get_dtype_init_constructors(dtype) ferminet_model_types = ['ferminet', 'embedded_particle_ferminet', 'extended_orbital_matrix_ferminet'] if (model_config.type in ferminet_model_types): determinant_fn = None resnet_config = model_config.det_resnet if model_config.use_det_resnet: determinant_fn = get_resnet_determinant_fn_for_ferminet(resnet_config.ndense, resnet_config.nlayers, _get_named_activation_fn(resnet_config.activation), kernel_init_constructor(resnet_config.kernel_init), bias_init_constructor(resnet_config.bias_init), resnet_config.use_bias) if (model_config.type == 'ferminet'): return FermiNet(spin_split, compute_input_streams, backflow, model_config.ndeterminants, kernel_initializer_orbital_linear=kernel_init_constructor(model_config.kernel_init_orbital_linear), kernel_initializer_envelope_dim=kernel_init_constructor(model_config.kernel_init_envelope_dim), kernel_initializer_envelope_ion=kernel_init_constructor(model_config.kernel_init_envelope_ion), envelope_softening=model_config.envelope_softening, bias_initializer_orbital_linear=bias_init_constructor(model_config.bias_init_orbital_linear), orbitals_use_bias=model_config.orbitals_use_bias, isotropic_decay=model_config.isotropic_decay, determinant_fn=determinant_fn, determinant_fn_mode=DeterminantFnMode[resnet_config.mode.upper()], full_det=model_config.full_det) elif (model_config.type == 'embedded_particle_ferminet'): total_nelec = (jnp.array(model_config.nhidden_fermions_per_spin) + nelec) total_spin_split = get_spin_split(total_nelec) backflow = get_backflow_from_config(model_config.backflow, total_spin_split, dtype=dtype) invariance_config = model_config.invariance invariance_compute_input_streams = get_compute_input_streams_from_config(invariance_config.input_streams, ion_pos) invariance_backflow = get_backflow_from_config(invariance_config.backflow, spin_split, dtype=dtype) return EmbeddedParticleFermiNet(spin_split, compute_input_streams, backflow, model_config.ndeterminants, kernel_initializer_orbital_linear=kernel_init_constructor(model_config.kernel_init_orbital_linear), kernel_initializer_envelope_dim=kernel_init_constructor(model_config.kernel_init_envelope_dim), kernel_initializer_envelope_ion=kernel_init_constructor(model_config.kernel_init_envelope_ion), envelope_softening=model_config.envelope_softening, bias_initializer_orbital_linear=bias_init_constructor(model_config.bias_init_orbital_linear), orbitals_use_bias=model_config.orbitals_use_bias, isotropic_decay=model_config.isotropic_decay, determinant_fn=determinant_fn, determinant_fn_mode=DeterminantFnMode[resnet_config.mode.upper()], full_det=model_config.full_det, nhidden_fermions_per_spin=model_config.nhidden_fermions_per_spin, invariance_compute_input_streams=invariance_compute_input_streams, invariance_backflow=invariance_backflow, invariance_kernel_initializer=kernel_init_constructor(invariance_config.kernel_initializer), invariance_bias_initializer=bias_init_constructor(invariance_config.bias_initializer), invariance_use_bias=invariance_config.use_bias) elif (model_config.type == 'extended_orbital_matrix_ferminet'): invariance_config = model_config.invariance if model_config.use_separate_invariance_backflow: invariance_backflow = get_backflow_from_config(invariance_config.backflow, spin_split, dtype=dtype) else: invariance_backflow = None return ExtendedOrbitalMatrixFermiNet(spin_split, compute_input_streams, backflow, model_config.ndeterminants, kernel_initializer_orbital_linear=kernel_init_constructor(model_config.kernel_init_orbital_linear), kernel_initializer_envelope_dim=kernel_init_constructor(model_config.kernel_init_envelope_dim), kernel_initializer_envelope_ion=kernel_init_constructor(model_config.kernel_init_envelope_ion), envelope_softening=model_config.envelope_softening, bias_initializer_orbital_linear=bias_init_constructor(model_config.bias_init_orbital_linear), orbitals_use_bias=model_config.orbitals_use_bias, isotropic_decay=model_config.isotropic_decay, determinant_fn=determinant_fn, determinant_fn_mode=DeterminantFnMode[resnet_config.mode.upper()], full_det=model_config.full_det, nhidden_fermions_per_spin=model_config.nhidden_fermions_per_spin, invariance_backflow=invariance_backflow, invariance_kernel_initializer=kernel_init_constructor(invariance_config.kernel_initializer), invariance_bias_initializer=bias_init_constructor(invariance_config.bias_initializer), invariance_use_bias=invariance_config.use_bias) else: raise ValueError('FermiNet model type {} requested, but the only supported types are: {}'.format(model_config.type, ferminet_model_types)) elif (model_config.type in ['orbital_cofactor_net', 'per_particle_dets_net']): if (model_config.type == 'orbital_cofactor_net'): antieq_layer: Callable[([Array, Array], ArrayList)] = antiequivariance.OrbitalCofactorAntiequivarianceLayer(spin_split, kernel_initializer_orbital_linear=kernel_init_constructor(model_config.kernel_init_orbital_linear), kernel_initializer_envelope_dim=kernel_init_constructor(model_config.kernel_init_envelope_dim), kernel_initializer_envelope_ion=kernel_init_constructor(model_config.kernel_init_envelope_ion), bias_initializer_orbital_linear=bias_init_constructor(model_config.bias_init_orbital_linear), orbitals_use_bias=model_config.orbitals_use_bias, isotropic_decay=model_config.isotropic_decay) elif (model_config.type == 'per_particle_dets_net'): antieq_layer = antiequivariance.PerParticleDeterminantAntiequivarianceLayer(spin_split, kernel_initializer_orbital_linear=kernel_init_constructor(model_config.kernel_init_orbital_linear), kernel_initializer_envelope_dim=kernel_init_constructor(model_config.kernel_init_envelope_dim), kernel_initializer_envelope_ion=kernel_init_constructor(model_config.kernel_init_envelope_ion), bias_initializer_orbital_linear=bias_init_constructor(model_config.bias_init_orbital_linear), orbitals_use_bias=model_config.orbitals_use_bias, isotropic_decay=model_config.isotropic_decay) array_list_sign_covariance = get_sign_covariance_from_config(model_config, spin_split, kernel_init_constructor, dtype) return AntiequivarianceNet(spin_split, compute_input_streams, backflow, antieq_layer, array_list_sign_covariance, multiply_by_eq_features=model_config.multiply_by_eq_features) elif (model_config.type == 'explicit_antisym'): jastrow_config = model_config.jastrow def _get_two_body_decay_jastrow(): return get_two_body_decay_scaled_for_chargeless_molecules(ion_pos, ion_charges, init_ee_strength=jastrow_config.two_body_decay.init_ee_strength, trainable=jastrow_config.two_body_decay.trainable) def _get_backflow_based_jastrow(): if jastrow_config.backflow_based.use_separate_jastrow_backflow: jastrow_backflow = get_backflow_from_config(jastrow_config.backflow_based.backflow, spin_split, dtype=dtype) else: jastrow_backflow = None return BackflowJastrow(backflow=jastrow_backflow) if (jastrow_config.type == 'one_body_decay'): jastrow: Jastrow = OneBodyExpDecay(kernel_initializer=kernel_init_constructor(jastrow_config.one_body_decay.kernel_init)) elif (jastrow_config.type == 'two_body_decay'): jastrow = _get_two_body_decay_jastrow() elif (jastrow_config.type == 'backflow_based'): jastrow = _get_backflow_based_jastrow() elif (jastrow_config.type == 'two_body_decay_and_backflow_based'): two_body_decay_jastrow = _get_two_body_decay_jastrow() backflow_jastrow = _get_backflow_based_jastrow() jastrow = AddedModel([two_body_decay_jastrow, backflow_jastrow]) else: raise ValueError('Unsupported jastrow type; {} was requested, but the only supported types are: {}'.format(jastrow_config.type, ', '.join(VALID_JASTROW_TYPES))) if (model_config.antisym_type == 'factorized'): return FactorizedAntisymmetry(spin_split, compute_input_streams, backflow, jastrow, rank=model_config.rank, ndense_resnet=model_config.ndense_resnet, nlayers_resnet=model_config.nlayers_resnet, kernel_initializer_resnet=kernel_init_constructor(model_config.kernel_init_resnet), bias_initializer_resnet=bias_init_constructor(model_config.bias_init_resnet), activation_fn_resnet=_get_named_activation_fn(model_config.activation_fn_resnet), resnet_use_bias=model_config.resnet_use_bias) elif (model_config.antisym_type == 'generic'): return GenericAntisymmetry(spin_split, compute_input_streams, backflow, jastrow, ndense_resnet=model_config.ndense_resnet, nlayers_resnet=model_config.nlayers_resnet, kernel_initializer_resnet=kernel_init_constructor(model_config.kernel_init_resnet), bias_initializer_resnet=bias_init_constructor(model_config.bias_init_resnet), activation_fn_resnet=_get_named_activation_fn(model_config.activation_fn_resnet), resnet_use_bias=model_config.resnet_use_bias) else: raise ValueError('Unsupported explicit antisymmetry type; {} was requested'.format(model_config.antisym_type)) else: raise ValueError('Unsupported model type; {} was requested'.format(model_config.type))
class ReplayBuffer(Dataset): def __init__(self, observation_space: gym.spaces.Box, action_dim: int, capacity: int): observations = np.empty((capacity, *observation_space.shape), dtype=observation_space.dtype) actions = np.empty((capacity, action_dim), dtype=np.float32) rewards = np.empty((capacity,), dtype=np.float32) masks = np.empty((capacity,), dtype=np.float32) dones_float = np.empty((capacity,), dtype=np.float32) next_observations = np.empty((capacity, *observation_space.shape), dtype=observation_space.dtype) super().__init__(observations=observations, actions=actions, rewards=rewards, masks=masks, dones_float=dones_float, next_observations=next_observations, size=0) self.size = 0 self.insert_index = 0 self.capacity = capacity def initialize_with_dataset(self, dataset: Dataset, num_samples: Optional[int]): assert (self.insert_index == 0), 'Can insert a batch online in an empty replay buffer.' dataset_size = len(dataset.observations) if (num_samples is None): num_samples = dataset_size else: num_samples = min(dataset_size, num_samples) assert (self.capacity >= num_samples), 'Dataset cannot be larger than the replay buffer capacity.' if (num_samples < dataset_size): perm = np.random.permutation(dataset_size) indices = perm[:num_samples] else: indices = np.arange(num_samples) self.observations[:num_samples] = dataset.observations[indices] self.actions[:num_samples] = dataset.actions[indices] self.rewards[:num_samples] = dataset.rewards[indices] self.masks[:num_samples] = dataset.masks[indices] self.dones_float[:num_samples] = dataset.dones_float[indices] self.next_observations[:num_samples] = dataset.next_observations[indices] self.insert_index = num_samples self.size = num_samples def insert(self, observation: np.ndarray, action: np.ndarray, reward: float, mask: float, done_float: float, next_observation: np.ndarray): self.observations[self.insert_index] = observation self.actions[self.insert_index] = action self.rewards[self.insert_index] = reward self.masks[self.insert_index] = mask self.dones_float[self.insert_index] = done_float self.next_observations[self.insert_index] = next_observation self.insert_index = ((self.insert_index + 1) % self.capacity) self.size = min((self.size + 1), self.capacity)
def collate(samples, pad_idx, eos_idx, left_pad_source=True, left_pad_target=False, input_feeding=True): if (len(samples) == 0): return {} def merge(key, left_pad, move_eos_to_beginning=False): return data_utils.collate_tokens([s[key] for s in samples], pad_idx, eos_idx, left_pad, move_eos_to_beginning) id = torch.LongTensor([s['id'] for s in samples]) src_tokens = merge('source', left_pad=left_pad_source) src_lengths = torch.LongTensor([s['source'].numel() for s in samples]) (src_lengths, sort_order) = src_lengths.sort(descending=True) id = id.index_select(0, sort_order) src_tokens = src_tokens.index_select(0, sort_order) prev_output_tokens = None target = None if (samples[0].get('target', None) is not None): target = merge('target', left_pad=left_pad_target) target = target.index_select(0, sort_order) ntokens = sum((len(s['target']) for s in samples)) if input_feeding: prev_output_tokens = merge('target', left_pad=left_pad_target, move_eos_to_beginning=True) prev_output_tokens = prev_output_tokens.index_select(0, sort_order) else: ntokens = sum((len(s['source']) for s in samples)) batch = {'id': id, 'nsentences': len(samples), 'ntokens': ntokens, 'net_input': {'src_tokens': src_tokens, 'src_lengths': src_lengths}, 'target': target} if (prev_output_tokens is not None): batch['net_input']['prev_output_tokens'] = prev_output_tokens return batch
def evaluations_scipy(ty, pv): if (not ((scipy != None) and isinstance(ty, scipy.ndarray) and isinstance(pv, scipy.ndarray))): raise TypeError('type of ty and pv must be ndarray') if (len(ty) != len(pv)): raise ValueError('len(ty) must be equal to len(pv)') ACC = (100.0 * (ty == pv).mean()) MSE = ((ty - pv) ** 2).mean() l = len(ty) sumv = pv.sum() sumy = ty.sum() sumvy = (pv * ty).sum() sumvv = (pv * pv).sum() sumyy = (ty * ty).sum() with scipy.errstate(all='raise'): try: SCC = ((((l * sumvy) - (sumv * sumy)) * ((l * sumvy) - (sumv * sumy))) / (((l * sumvv) - (sumv * sumv)) * ((l * sumyy) - (sumy * sumy)))) except: SCC = float('nan') return (float(ACC), float(MSE), float(SCC))
class Annotator(object): def __init__(self, config, filenames, current_mode, args): self.current_mode = current_mode self.current_num = None self.search_term = '' self.partial_typing = '' self.cfilename = (- 1) self.filename = None self.filenames = filenames self.datum = None self.view = None self.window = None self.config = config self.args = args self.action_to_function = {'delete-query-char': self.delete_typing_char, 'leave-query-mode': self.leave_typing_mode, 'enter-query-mode': self.enter_typing_mode, 'clear-query': self.clear_query, 'add-to-query': self.add_to_typing, 'delete-label-char': self.delete_typing_char, 'assign-text-label': self.assign_text, 'enter-label-mode': self.enter_typing_mode, 'add-to-label': self.add_to_typing, 'toggle-line-numbers': self.toggle_line_numbers, 'move-up': self.move, 'move-down': self.move, 'move-left': self.move, 'move-right': self.move, 'move-link-up': self.move, 'move-link-down': self.move, 'move-link-left': self.move, 'move-link-right': self.move, 'jump-up': self.move, 'jump-down': self.move, 'jump-left': self.move, 'jump-right': self.move, 'extend-up': self.change_span, 'extend-down': self.change_span, 'extend-left': self.change_span, 'extend-right': self.change_span, 'contract-up': self.change_span, 'contract-down': self.change_span, 'contract-left': self.change_span, 'contract-right': self.change_span, 'extend-link-up': self.change_span, 'extend-link-down': self.change_span, 'extend-link-left': self.change_span, 'extend-link-right': self.change_span, 'contract-link-up': self.change_span, 'contract-link-down': self.change_span, 'contract-link-left': self.change_span, 'contract-link-right': self.change_span, 'search-previous': self.search, 'search-next': self.search, 'search-link-previous': self.search, 'search-link-next': self.search, 'page-up': self.shift_view, 'page-down': self.shift_view, 'toggle-help': self.modify_display, 'toggle-progress': self.modify_display, 'toggle-legend': self.modify_display, 'toggle-current-mark': self.modify_display, 'next-file': self.change_file, 'previous-file': self.change_file, 'quit': self.save_or_quit, 'save-and-quit': self.save_or_quit, 'save': self.save_or_quit, 'create-link': self.create_link, 'create-link-and-move': self.create_link, 'edit-annotation': self.edit_annotation, 'remove-annotation': self.remove_annotation, 'update-num': self.update_number} def move(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return direction = action.split('-')[(- 1)] jump = ('jump' in action) link = ('link' in action) num = 1 if (self.current_num == 0): jump = True self.current_num = None elif (self.current_num is not None): num = self.current_num self.current_num = None self.view.move(direction, num, jump, link) def toggle_line_numbers(self, user_input, action): self.view.line_numbers = (not self.view.line_numbers) def change_span(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return change = action.split('-')[0] direction = action.split('-')[(- 1)] link = ('link' in action) num = 1 jump = False if (self.current_num == 0): jump = True self.current_num = None elif (self.current_num is not None): num = self.current_num self.current_num = None self.view.adjust(direction, num, change, jump, link) def delete_typing_char(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if (self.current_mode[(- 1)] == 'write_query'): self.search_term = self.search_term[:(- 1)] else: self.partial_typing = self.partial_typing[:(- 1)] def leave_typing_mode(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if (len(self.current_mode) > 1): self.current_mode.pop() def assign_text(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if (len(self.current_mode) > 1): self.current_mode.pop() self.datum.modify_annotation([self.view.cursor], self.partial_typing) self.partial_typing = '' def enter_typing_mode(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if ('query' in action): self.current_mode.append('write_query') else: self.current_mode.append('write_label') self.partial_typing = '' def clear_query(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return self.search_term = '' def add_to_typing(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return char = user_input[0] if (user_input[0] == 'SPACE'): char = ' ' if (self.current_mode[(- 1)] == 'write_query'): self.search_term += char else: self.partial_typing += char def change_file(self, user_input, action): if (self.current_mode[(- 1)] != 'no_file'): self.save_or_quit(None, 'save') direction = (1 if ('next' in action) else (- 1)) if (self.current_mode[(- 1)] == 'no_file'): if ((self.cfilename < 0) == (direction > 0)): self.current_mode.pop() self.cfilename += direction elif (0 <= (self.cfilename + direction) < len(self.filenames)): self.cfilename += direction (self.filename, start_pos, output_file, annotation_files) = self.filenames[self.cfilename] self.datum = Datum(self.filename, self.config, output_file, annotation_files) self.get_view(self.config, self.cfilename, len(self.filenames), start_pos, self.view) elif (self.current_mode != 'no_file'): self.cfilename += direction self.current_mode.append('no_file') def modify_display(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if ('help' in action): self.view.toggle_help() elif ('progress' in action): self.view.toggle_progress() elif ('legend' in action): self.view.toggle_legend() elif ('current-mark' in action): self.view.toggle_current_mark() def shift_view(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if ('up' in action): self.view.shift_view() else: self.view.shift_view(True) def update_number(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return num = int(user_input[0]) if (self.current_num is None): self.current_num = 0 else: self.current_num *= 10 self.current_num += num def remove_annotation(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if (self.current_mode[(- 1)] != 'read'): spans = [self.view.cursor] if (self.current_mode[(- 1)] == 'link'): spans = [self.view.linking_pos] self.datum.remove_annotation(spans) def edit_annotation(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return if (self.current_mode[(- 1)] == 'category'): label = self.config.get_label_for_input(user_input) self.datum.modify_annotation([self.view.cursor], label) def create_link(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return self.datum.modify_annotation([self.view.cursor, self.view.linking_pos]) if ('and-move' in action): if (self.config.annotation == 'line'): self.view.move('down', 1, False, True) self.view.put_cursor_beside_link() else: self.view.move('right', 1, False, True) self.view.put_cursor_beside_link() self.view.must_show_linking_pos = True def save_or_quit(self, user_input, action): if ('save' in action): if (self.current_mode[(- 1)] != 'read'): self.datum.write_out() if (0 <= self.cfilename < len(self.filenames)): cur = self.filenames[self.cfilename] pos = self.view.cursor if (self.config.annotation_type == 'link'): pos = self.view.linking_pos self.filenames[self.cfilename] = (cur[0], pos, cur[2], cur[3]) if ('quit' in action): if ('save' not in action): pass return 'quit' def search(self, user_input, action): if (self.current_mode[(- 1)] == 'no_file'): return direction = action.split('-')[(- 1)] jump = False link = ('link' in action) num = 1 if (self.current_num == 0): jump = True self.current_num = None elif (self.current_num is not None): num = self.current_num self.current_num = None if (len(self.search_term) > 0): self.view.search(self.search_term, direction, num, jump, link) else: self.view.search(None, direction, num, jump, link) def input_to_symbol(self, num): if (num in key_to_symbol): return key_to_symbol[num] else: return 'UNKNOWN' def get_view(self, config, file_num, total_files, position, prev_view=None): cursor = position link = (position if (self.config.annotation_type == 'link') else None) self.view = View(self.window, cursor, link, self.datum, self.config, file_num, total_files, prev_view) def annotate(self, window_in): self.window = window_in curses.use_default_colors() for (num, fore, back) in COLORS: curses.init_pair(num, fore, back) curses.curs_set(0) self.cfilename = 0 (self.filename, start_pos, output_file, annotation_files) = self.filenames[self.cfilename] self.datum = Datum(self.filename, self.config, output_file, annotation_files) self.get_view(self.config, self.cfilename, len(self.filenames), start_pos) if self.args.show_help: self.view.toggle_help() if self.args.show_progress: self.view.toggle_progress() if self.args.show_legend: self.view.toggle_legend() if self.args.show_mark: self.view.toggle_current_mark() last_num = None at_end = None nsteps = 0 user_input = [] while True: if (self.current_mode[(- 1)] == 'no_file'): self.view.render_edgecase((self.cfilename >= 0)) else: tmp_term = self.search_term if (self.current_mode[(- 1)] == 'write_query'): tmp_term = ('\\' + tmp_term) self.view.render(tmp_term, self.partial_typing) self.view.must_show_linking_pos = False ch = self.window.getch() next_user_input = self.input_to_symbol(ch) logging.debug('Input {} converted to {} in mode {}'.format(ch, next_user_input, self.current_mode)) user_input.append(next_user_input) tuser_input = tuple(user_input) if ((self.current_mode[(- 1)], tuser_input) not in self.config.valid_prefixes): if ((None, tuser_input) not in self.config.valid_prefixes): if ((self.current_mode[(- 1)], tuser_input) not in self.config.input_to_action): if ((None, tuser_input) not in self.config.input_to_action): user_input = [next_user_input] tuser_input = (next_user_input,) nsteps += 1 if (((nsteps % 100) == 0) and (self.current_mode[(- 1)] == 'category')): self.datum.write_out() action = None function = None if ((self.current_mode[(- 1)], tuser_input) in self.config.input_to_action): action = self.config.input_to_action[(self.current_mode[(- 1)], tuser_input)] if (action in self.action_to_function): function = self.action_to_function[action] elif ((None, tuser_input) in self.config.input_to_action): action = self.config.input_to_action[(None, tuser_input)] if (action in self.action_to_function): function = self.action_to_function[action] logging.debug('{} {} -> {} {}'.format(self.current_mode, tuser_input, action, function)) if (function is not None): outcome = function(tuser_input, action) user_input = [] if (outcome == 'quit'): break self.window.clear() out_filename = (self.args.log_prefix + '.todo') out = open(out_filename, 'w') for (fname, start_pos, output_file, annotation_files) in self.filenames: parts = [fname, output_file, str(start_pos), ' '.join(annotation_files)] print(' '.join(parts), file=out) out.close()
class dataset_iitnn(Dataset): def __init__(self, data_dir, input1, input2, augmentation1, normalize_1, normalize_2, sup=True, num_images=None, **kwargs): super(dataset_iitnn, self).__init__() img_paths_1 = [] img_paths_2 = [] mask_paths = [] image_dir_1 = ((data_dir + '/') + input1) image_dir_2 = ((data_dir + '/') + input2) if sup: mask_dir = (data_dir + '/mask') for image in os.listdir(image_dir_1): image_path_1 = os.path.join(image_dir_1, image) img_paths_1.append(image_path_1) image_path_2 = os.path.join(image_dir_2, image) img_paths_2.append(image_path_2) if sup: mask_path = os.path.join(mask_dir, image) mask_paths.append(mask_path) assert (len(img_paths_1) == len(img_paths_2)) if sup: assert (len(img_paths_1) == len(mask_paths)) if (num_images is not None): len_img_paths = len(img_paths_1) quotient = (num_images // len_img_paths) remainder = (num_images % len_img_paths) if (num_images <= len_img_paths): img_paths_1 = img_paths_1[:num_images] img_paths_2 = img_paths_2[:num_images] else: rand_indices = torch.randperm(len_img_paths).tolist() new_indices = rand_indices[:remainder] img_paths_1 = (img_paths_1 * quotient) img_paths_1 += [img_paths_1[i] for i in new_indices] img_paths_2 = (img_paths_2 * quotient) img_paths_2 += [img_paths_2[i] for i in new_indices] if sup: mask_paths = (mask_paths * quotient) mask_paths += [mask_paths[i] for i in new_indices] self.img_paths_1 = img_paths_1 self.img_paths_2 = img_paths_2 self.mask_paths = mask_paths self.augmentation_1 = augmentation1 self.normalize_1 = normalize_1 self.normalize_2 = normalize_2 self.sup = sup self.kwargs = kwargs def __getitem__(self, index): img_path_1 = self.img_paths_1[index] img_1 = Image.open(img_path_1) img_1 = np.array(img_1) img_path_2 = self.img_paths_2[index] img_2 = Image.open(img_path_2) img_2 = np.array(img_2) if self.sup: mask_path = self.mask_paths[index] mask = Image.open(mask_path) mask = np.array(mask) augment_1 = self.augmentation_1(image=img_1, image2=img_2, mask=mask) img_1 = augment_1['image'] img_2 = augment_1['image2'] mask = augment_1['mask'] normalize_1 = self.normalize_1(image=img_1, mask=mask) img_1 = normalize_1['image'] mask = normalize_1['mask'] mask = mask.long() normalize_2 = self.normalize_2(image=img_2) img_2 = normalize_2['image'] sampel = {'image': img_1, 'image_2': img_2, 'mask': mask, 'ID': os.path.split(mask_path)[1]} else: augment_1 = self.augmentation_1(image=img_1, image2=img_2) img_1 = augment_1['image'] img_2 = augment_1['image2'] normalize_1 = self.normalize_1(image=img_1) img_1 = normalize_1['image'] normalize_2 = self.normalize_2(image=img_2) img_2 = normalize_2['image'] sampel = {'image': img_1, 'image_2': img_2, 'ID': os.path.split(img_path_1)[1]} return sampel def __len__(self): return len(self.img_paths_1)
def describe_graph(graph: str, expert_description='an expert statistician and data scientist.', y_axis_description='', special_task_description='', dataset_description='', include_assistant_response=True): prompt = (('{{#system~}}\n' + f'''You are {expert_description} You interpret global explanations produced by a generalized additive model (GAM). GAMs produce explanations in the form of graphs that contain the effect of a specific input feature. {('You will be given graphs from the model, and the user will ask you questions about the graphs.' if ((dataset_description is None) or (dataset_description == '')) else 'The user will first provide a general description of the dataset. Then you will be given graphs from the model, and the user will ask you questions about the graphs.')} Answer all questions to the best of your ability, combining both the data contained in the graph{(', the data set description you were given, and your knowledge about the real world.' if ((dataset_description is not None) and (len(dataset_description) > 0)) else ' and your knowledge about the real world.')} Graphs will be presented as a JSON object with keys representing the x-axis and values representing the y-axis. For continuous features, the keys are intervals that represent ranges where the function predicts the same value. For categorical features, each key represents a possible value that the feature can take. {(y_axis_description if ((y_axis_description is not None) and (len(dataset_description) > 0)) else '')} The user will provide graphs in the following format: - The name of the feature depicted in the graph - The type of the feature (continuous, categorical, or boolean) - Mean values - Lower bounds of confidence interval - Upper bounds of confidence interval {special_task_description} ''') + '{{~/system}}\n') if ((dataset_description is not None) and (len(dataset_description) > 0)): prompt += (('\n{{#user~}}\n' + dataset_description) + '\n{{~/user}}\n\n{{#assistant~}}\nThanks for this general description of the data set. Please continue and provide more information, for example about the graphs from the model.\n{{~/assistant}}\n') prompt += (('\n{{#user~}}\nConsider the following graph from the model. ' + graph) + '\nPlease describe the general pattern of the graph.\n{{~/user}}\n\n') if include_assistant_response: prompt += "{{#assistant~}}{{gen 'graph_description' temperature=0.7 max_tokens=2000}}{{~/assistant}}" return prompt
def kaiming_init(module, mode='fan_out', nonlinearity='relu', bias=0, distribution='normal'): assert (distribution in ['uniform', 'normal']) if (distribution == 'uniform'): nn.init.kaiming_uniform_(module.weight, mode=mode, nonlinearity=nonlinearity) else: nn.init.kaiming_normal_(module.weight, mode=mode, nonlinearity=nonlinearity) if hasattr(module, 'bias'): nn.init.constant_(module.bias, bias)
def ibn_pre_conv1x1_block(in_channels, out_channels, stride=1, use_ibn=False, return_preact=False): return IBNPreConvBlock(in_channels=in_channels, out_channels=out_channels, kernel_size=1, stride=stride, padding=0, use_ibn=use_ibn, return_preact=return_preact)
_module() class PISARetinaHead(RetinaHead): _fp32(apply_to=('cls_scores', 'bbox_preds')) def loss(self, cls_scores, bbox_preds, gt_bboxes, gt_labels, img_metas, gt_bboxes_ignore=None): featmap_sizes = [featmap.size()[(- 2):] for featmap in cls_scores] assert (len(featmap_sizes) == self.prior_generator.num_levels) device = cls_scores[0].device (anchor_list, valid_flag_list) = self.get_anchors(featmap_sizes, img_metas, device=device) label_channels = (self.cls_out_channels if self.use_sigmoid_cls else 1) cls_reg_targets = self.get_targets(anchor_list, valid_flag_list, gt_bboxes, img_metas, gt_bboxes_ignore_list=gt_bboxes_ignore, gt_labels_list=gt_labels, label_channels=label_channels, return_sampling_results=True) if (cls_reg_targets is None): return None (labels_list, label_weights_list, bbox_targets_list, bbox_weights_list, num_total_pos, num_total_neg, sampling_results_list) = cls_reg_targets num_total_samples = ((num_total_pos + num_total_neg) if self.sampling else num_total_pos) num_level_anchors = [anchors.size(0) for anchors in anchor_list[0]] concat_anchor_list = [] for i in range(len(anchor_list)): concat_anchor_list.append(torch.cat(anchor_list[i])) all_anchor_list = images_to_levels(concat_anchor_list, num_level_anchors) num_imgs = len(img_metas) flatten_cls_scores = [cls_score.permute(0, 2, 3, 1).reshape(num_imgs, (- 1), label_channels) for cls_score in cls_scores] flatten_cls_scores = torch.cat(flatten_cls_scores, dim=1).reshape((- 1), flatten_cls_scores[0].size((- 1))) flatten_bbox_preds = [bbox_pred.permute(0, 2, 3, 1).reshape(num_imgs, (- 1), 4) for bbox_pred in bbox_preds] flatten_bbox_preds = torch.cat(flatten_bbox_preds, dim=1).view((- 1), flatten_bbox_preds[0].size((- 1))) flatten_labels = torch.cat(labels_list, dim=1).reshape((- 1)) flatten_label_weights = torch.cat(label_weights_list, dim=1).reshape((- 1)) flatten_anchors = torch.cat(all_anchor_list, dim=1).reshape((- 1), 4) flatten_bbox_targets = torch.cat(bbox_targets_list, dim=1).reshape((- 1), 4) flatten_bbox_weights = torch.cat(bbox_weights_list, dim=1).reshape((- 1), 4) isr_cfg = self.train_cfg.get('isr', None) if (isr_cfg is not None): all_targets = (flatten_labels, flatten_label_weights, flatten_bbox_targets, flatten_bbox_weights) with torch.no_grad(): all_targets = isr_p(flatten_cls_scores, flatten_bbox_preds, all_targets, flatten_anchors, sampling_results_list, bbox_coder=self.bbox_coder, loss_cls=self.loss_cls, num_class=self.num_classes, **self.train_cfg.isr) (flatten_labels, flatten_label_weights, flatten_bbox_targets, flatten_bbox_weights) = all_targets losses_cls = self.loss_cls(flatten_cls_scores, flatten_labels, flatten_label_weights, avg_factor=num_total_samples) losses_bbox = self.loss_bbox(flatten_bbox_preds, flatten_bbox_targets, flatten_bbox_weights, avg_factor=num_total_samples) loss_dict = dict(loss_cls=losses_cls, loss_bbox=losses_bbox) carl_cfg = self.train_cfg.get('carl', None) if (carl_cfg is not None): loss_carl = carl_loss(flatten_cls_scores, flatten_labels, flatten_bbox_preds, flatten_bbox_targets, self.loss_bbox, **self.train_cfg.carl, avg_factor=num_total_pos, sigmoid=True, num_class=self.num_classes) loss_dict.update(loss_carl) return loss_dict
def make_data_config(config: ml_collections.ConfigDict, mode: str, num_res: int) -> Tuple[(ml_collections.ConfigDict, List[str])]: cfg = copy.deepcopy(config) mode_cfg = cfg[mode] with cfg.unlocked(): if (mode_cfg.crop_size is None): mode_cfg.crop_size = num_res feature_names = cfg.common.unsupervised_features if cfg.common.use_templates: feature_names += cfg.common.template_features if cfg[mode].supervised: feature_names += cfg.supervised.supervised_features return (cfg, feature_names)
def default_auto_wrap_policy(module: nn.Module, recurse: bool, unwrapped_params: int, min_num_params: int=int(.0), force_leaf_modules: Optional[Set[Type[nn.Module]]]=None, exclude_wrap_modules: Optional[Set[Type[nn.Module]]]=None) -> bool: force_leaf_modules = (default_auto_wrap_policy.FORCE_LEAF_MODULES if (force_leaf_modules is None) else force_leaf_modules) exclude_wrap_modules = (default_auto_wrap_policy.EXCLUDE_WRAP_MODULES if (exclude_wrap_modules is None) else exclude_wrap_modules) is_large = (unwrapped_params >= min_num_params) if recurse: return (is_large and (not isinstance(module, tuple(force_leaf_modules)))) else: return (is_large and (not isinstance(module, tuple(exclude_wrap_modules))))
def get_config(FLAGS): config = Config for (k, v) in FLAGS.__flags.items(): if hasattr(config, k): setattr(config, k, v.value) return config
class DataLoader(): def __init__(self, csv_file='data/nyu2_train.csv', DEBUG=False): self.shape_rgb = (480, 640, 3) self.shape_depth = (240, 320, 1) self.read_nyu_data(csv_file, DEBUG=DEBUG) def nyu_resize(self, img, resolution=480, padding=6): from skimage.transform import resize return resize(img, (resolution, int(((resolution * 4) / 3))), preserve_range=True, mode='reflect', anti_aliasing=True) def read_nyu_data(self, csv_file, DEBUG=False): csv = open(csv_file, 'r').read() nyu2_train = list((row.split(',') for row in csv.split('\n') if (len(row) > 0))) nyu2_train = shuffle(nyu2_train, random_state=0) if DEBUG: nyu2_train = nyu2_train[:10] self.filenames = [i[0] for i in nyu2_train] self.labels = [i[1] for i in nyu2_train] self.length = len(self.filenames) def _parse_function(self, filename, label): image_decoded = tf.image.decode_jpeg(tf.io.read_file(filename)) depth_resized = tf.image.resize(tf.image.decode_jpeg(tf.io.read_file(label)), [self.shape_depth[0], self.shape_depth[1]]) rgb = tf.image.convert_image_dtype(image_decoded, dtype=tf.float32) depth = tf.image.convert_image_dtype((depth_resized / 255.0), dtype=tf.float32) depth = (1000 / tf.clip_by_value((depth * 1000), 10, 1000)) return (rgb, depth) def get_batched_dataset(self, batch_size): self.dataset = tf.data.Dataset.from_tensor_slices((self.filenames, self.labels)) self.dataset = self.dataset.shuffle(buffer_size=len(self.filenames), reshuffle_each_iteration=True) self.dataset = self.dataset.repeat() self.dataset = self.dataset.map(map_func=self._parse_function, num_parallel_calls=tf.data.experimental.AUTOTUNE) self.dataset = self.dataset.batch(batch_size=batch_size) return self.dataset
def build_lr_scheduler(config, optimizer): assert isinstance(config, dict) lr_type = config['lr_type'].upper() warmup_type = config.get('warmup_type', 'NO') warmup_iters = config.get('warmup_iters', 0) warmup_factor = config.get('warmup_factor', 0.1) if (lr_type not in _ALLOWED_LR_TYPES): raise ValueError(f'Invalid learning rate scheduler type `{lr_type}`!Allowed types: {_ALLOWED_LR_TYPES}.') if (lr_type == 'FIXED'): return FixedWarmUpLR(optimizer=optimizer, warmup_type=warmup_type, warmup_iters=warmup_iters, warmup_factor=warmup_factor) if (lr_type == 'STEP'): return StepWarmUpLR(optimizer=optimizer, decay_step=config['decay_step'], decay_factor=config.get('decay_factor', 0.1), warmup_type=warmup_type, warmup_iters=warmup_iters, warmup_factor=warmup_factor) if (lr_type == 'EXPSTEP'): return EXPStepWarmUpLR(optimizer=optimizer, decay_step=config['decay_step'], decay_factor=config.get('decay_factor', 0.1), warmup_type=warmup_type, warmup_iters=warmup_iters, warmup_factor=warmup_factor) raise NotImplementedError(f'Not implemented scheduler type `{lr_type}`!')
def add_jitter(models, sd=0.1): for a_model in models: a_model.kernel = add_jitter_k([a_model.kernel], sd=sd)[0] return models
class _SimpleSegmentationModel(nn.Module): def __init__(self, backbone, classifier, im_num, ex_num): super(_SimpleSegmentationModel, self).__init__() self.backbone = backbone self.classifier = classifier self.bat_low = _bound_learner(hidden_features=128, im_num=im_num, ex_num=ex_num) def forward(self, x): input_shape = x.shape[(- 2):] features = self.backbone(x) (features, point_pre1, point_pre2, point_pre3) = self.bat_low(features) x = self.classifier(features) x = F.interpolate(x, size=input_shape, mode='bilinear', align_corners=False) return (x, point_pre1, point_pre2, point_pre3)
def build_and_train(slot_affinity_code, log_dir, run_ID, config_key): affinity = affinity_from_code(slot_affinity_code) config = configs[config_key] variant = load_variant(log_dir) config = update_config(config, variant) config['eval_env']['game'] = config['env']['game'] sampler = AsyncSerialSampler(EnvCls=AtariEnv, env_kwargs=config['env'], CollectorCls=DbCpuResetCollector, TrajInfoCls=AtariTrajInfo, eval_env_kwargs=config['eval_env'], **config['sampler']) algo = DQN(optim_kwargs=config['optim'], **config['algo']) agent = AtariDqnAgent(model_kwargs=config['model'], **config['agent']) runner = AsyncRlEval(algo=algo, agent=agent, sampler=sampler, affinity=affinity, **config['runner']) name = ('async_serial_' + config['env']['game']) with logger_context(log_dir, run_ID, name, config): runner.train()
class BaseDataset(): def __init__(self): pass def get_pair(self, cls, shuffle): raise NotImplementedError
def learn(env, policy_func, reward_giver, expert_dataset, rank, pretrained, pretrained_weight, *, g_step, d_step, entcoeff, save_per_iter, ckpt_dir, log_dir, timesteps_per_batch, task_name, gamma, lam, max_kl, cg_iters, cg_damping=0.01, vf_stepsize=0.0003, d_stepsize=0.0003, vf_iters=3, max_timesteps=0, max_episodes=0, max_iters=0, callback=None): nworkers = MPI.COMM_WORLD.Get_size() rank = MPI.COMM_WORLD.Get_rank() np.set_printoptions(precision=3) ob_space = env.observation_space ac_space = env.action_space pi = policy_func('pi', ob_space, ac_space, reuse=(pretrained_weight != None)) oldpi = policy_func('oldpi', ob_space, ac_space) atarg = tf.placeholder(dtype=tf.float32, shape=[None]) ret = tf.placeholder(dtype=tf.float32, shape=[None]) ob = U.get_placeholder_cached(name='ob') ac = pi.pdtype.sample_placeholder([None]) kloldnew = oldpi.pd.kl(pi.pd) ent = pi.pd.entropy() meankl = tf.reduce_mean(kloldnew) meanent = tf.reduce_mean(ent) entbonus = (entcoeff * meanent) vferr = tf.reduce_mean(tf.square((pi.vpred - ret))) ratio = tf.exp((pi.pd.logp(ac) - oldpi.pd.logp(ac))) surrgain = tf.reduce_mean((ratio * atarg)) optimgain = (surrgain + entbonus) losses = [optimgain, meankl, entbonus, surrgain, meanent] loss_names = ['optimgain', 'meankl', 'entloss', 'surrgain', 'entropy'] dist = meankl all_var_list = pi.get_trainable_variables() var_list = [v for v in all_var_list if (v.name.startswith('pi/pol') or v.name.startswith('pi/logstd'))] vf_var_list = [v for v in all_var_list if v.name.startswith('pi/vff')] assert (len(var_list) == (len(vf_var_list) + 1)) d_adam = MpiAdam(reward_giver.get_trainable_variables()) vfadam = MpiAdam(vf_var_list) get_flat = U.GetFlat(var_list) set_from_flat = U.SetFromFlat(var_list) klgrads = tf.gradients(dist, var_list) flat_tangent = tf.placeholder(dtype=tf.float32, shape=[None], name='flat_tan') shapes = [var.get_shape().as_list() for var in var_list] start = 0 tangents = [] for shape in shapes: sz = U.intprod(shape) tangents.append(tf.reshape(flat_tangent[start:(start + sz)], shape)) start += sz gvp = tf.add_n([tf.reduce_sum((g * tangent)) for (g, tangent) in zipsame(klgrads, tangents)]) fvp = U.flatgrad(gvp, var_list) assign_old_eq_new = U.function([], [], updates=[tf.assign(oldv, newv) for (oldv, newv) in zipsame(oldpi.get_variables(), pi.get_variables())]) compute_losses = U.function([ob, ac, atarg], losses) compute_lossandgrad = U.function([ob, ac, atarg], (losses + [U.flatgrad(optimgain, var_list)])) compute_fvp = U.function([flat_tangent, ob, ac, atarg], fvp) compute_vflossandgrad = U.function([ob, ret], U.flatgrad(vferr, vf_var_list)) def timed(msg): if (rank == 0): print(colorize(msg, color='magenta')) tstart = time.time() (yield) print(colorize(('done in %.3f seconds' % (time.time() - tstart)), color='magenta')) else: (yield) def allmean(x): assert isinstance(x, np.ndarray) out = np.empty_like(x) MPI.COMM_WORLD.Allreduce(x, out, op=MPI.SUM) out /= nworkers return out U.initialize() th_init = get_flat() MPI.COMM_WORLD.Bcast(th_init, root=0) set_from_flat(th_init) d_adam.sync() vfadam.sync() if (rank == 0): print('Init param sum', th_init.sum(), flush=True) seg_gen = traj_segment_generator(pi, env, reward_giver, timesteps_per_batch, stochastic=True) episodes_so_far = 0 timesteps_so_far = 0 iters_so_far = 0 tstart = time.time() lenbuffer = deque(maxlen=40) rewbuffer = deque(maxlen=40) true_rewbuffer = deque(maxlen=40) assert (sum([(max_iters > 0), (max_timesteps > 0), (max_episodes > 0)]) == 1) g_loss_stats = stats(loss_names) d_loss_stats = stats(reward_giver.loss_name) ep_stats = stats(['True_rewards', 'Rewards', 'Episode_length']) if (pretrained_weight is not None): U.load_state(pretrained_weight, var_list=pi.get_variables()) while True: if callback: callback(locals(), globals()) if (max_timesteps and (timesteps_so_far >= max_timesteps)): break elif (max_episodes and (episodes_so_far >= max_episodes)): break elif (max_iters and (iters_so_far >= max_iters)): break if ((rank == 0) and ((iters_so_far % save_per_iter) == 0) and (ckpt_dir is not None)): fname = os.path.join(ckpt_dir, task_name) os.makedirs(os.path.dirname(fname), exist_ok=True) saver = tf.train.Saver() saver.save(tf.get_default_session(), fname) logger.log((' Iteration %i ' % iters_so_far)) def fisher_vector_product(p): return (allmean(compute_fvp(p, *fvpargs)) + (cg_damping * p)) logger.log('Optimizing Policy...') for _ in range(g_step): with timed('sampling'): seg = seg_gen.__next__() add_vtarg_and_adv(seg, gamma, lam) (ob, ac, atarg, tdlamret) = (seg['ob'], seg['ac'], seg['adv'], seg['tdlamret']) vpredbefore = seg['vpred'] atarg = ((atarg - atarg.mean()) / atarg.std()) if hasattr(pi, 'ob_rms'): pi.ob_rms.update(ob) args = (seg['ob'], seg['ac'], atarg) fvpargs = [arr[::5] for arr in args] assign_old_eq_new() with timed('computegrad'): (*lossbefore, g) = compute_lossandgrad(*args) lossbefore = allmean(np.array(lossbefore)) g = allmean(g) if np.allclose(g, 0): logger.log('Got zero gradient. not updating') else: with timed('cg'): stepdir = cg(fisher_vector_product, g, cg_iters=cg_iters, verbose=(rank == 0)) assert np.isfinite(stepdir).all() shs = (0.5 * stepdir.dot(fisher_vector_product(stepdir))) lm = np.sqrt((shs / max_kl)) fullstep = (stepdir / lm) expectedimprove = g.dot(fullstep) surrbefore = lossbefore[0] stepsize = 1.0 thbefore = get_flat() for _ in range(10): thnew = (thbefore + (fullstep * stepsize)) set_from_flat(thnew) meanlosses = (surr, kl, *_) = allmean(np.array(compute_losses(*args))) improve = (surr - surrbefore) logger.log(('Expected: %.3f Actual: %.3f' % (expectedimprove, improve))) if (not np.isfinite(meanlosses).all()): logger.log('Got non-finite value of losses -- bad!') elif (kl > (max_kl * 1.5)): logger.log('violated KL constraint. shrinking step.') elif (improve < 0): logger.log("surrogate didn't improve. shrinking step.") else: logger.log('Stepsize OK!') break stepsize *= 0.5 else: logger.log("couldn't compute a good step") set_from_flat(thbefore) if ((nworkers > 1) and ((iters_so_far % 20) == 0)): paramsums = MPI.COMM_WORLD.allgather((thnew.sum(), vfadam.getflat().sum())) assert all((np.allclose(ps, paramsums[0]) for ps in paramsums[1:])) with timed('vf'): for _ in range(vf_iters): for (mbob, mbret) in dataset.iterbatches((seg['ob'], seg['tdlamret']), include_final_partial_batch=False, batch_size=128): if hasattr(pi, 'ob_rms'): pi.ob_rms.update(mbob) g = allmean(compute_vflossandgrad(mbob, mbret)) vfadam.update(g, vf_stepsize) g_losses = meanlosses for (lossname, lossval) in zip(loss_names, meanlosses): logger.record_tabular(lossname, lossval) logger.record_tabular('ev_tdlam_before', explained_variance(vpredbefore, tdlamret)) logger.log('Optimizing Discriminator...') logger.log(fmt_row(13, reward_giver.loss_name)) (ob_expert, ac_expert) = expert_dataset.get_next_batch(len(ob)) batch_size = (len(ob) // d_step) d_losses = [] for (ob_batch, ac_batch) in dataset.iterbatches((ob, ac), include_final_partial_batch=False, batch_size=batch_size): (ob_expert, ac_expert) = expert_dataset.get_next_batch(len(ob_batch)) if hasattr(reward_giver, 'obs_rms'): reward_giver.obs_rms.update(np.concatenate((ob_batch, ob_expert), 0)) (*newlosses, g) = reward_giver.lossandgrad(ob_batch, ac_batch, ob_expert, ac_expert) d_adam.update(allmean(g), d_stepsize) d_losses.append(newlosses) logger.log(fmt_row(13, np.mean(d_losses, axis=0))) lrlocal = (seg['ep_lens'], seg['ep_rets'], seg['ep_true_rets']) listoflrpairs = MPI.COMM_WORLD.allgather(lrlocal) (lens, rews, true_rets) = map(flatten_lists, zip(*listoflrpairs)) true_rewbuffer.extend(true_rets) lenbuffer.extend(lens) rewbuffer.extend(rews) logger.record_tabular('EpLenMean', np.mean(lenbuffer)) logger.record_tabular('EpRewMean', np.mean(rewbuffer)) logger.record_tabular('EpTrueRewMean', np.mean(true_rewbuffer)) logger.record_tabular('EpThisIter', len(lens)) episodes_so_far += len(lens) timesteps_so_far += sum(lens) iters_so_far += 1 logger.record_tabular('EpisodesSoFar', episodes_so_far) logger.record_tabular('TimestepsSoFar', timesteps_so_far) logger.record_tabular('TimeElapsed', (time.time() - tstart)) if (rank == 0): logger.dump_tabular()
def get_delta(pca, latent, idx, strength): w_centered = (latent - pca['mean'].to('cuda')) lat_comp = pca['comp'].to('cuda') lat_std = pca['std'].to('cuda') w_coord = (torch.sum((w_centered[0].reshape((- 1)) * lat_comp[idx].reshape((- 1)))) / lat_std[idx]) delta = (((strength - w_coord) * lat_comp[idx]) * lat_std[idx]) return delta
def train(model, device, dataset, fold, restart, seed): params_bb = ({param for param in model.learner.image_encoder.parameters()} - {param for param in model.learner.image_encoder.terminal_module_dict.parameters()}) params_new = ({param for param in model.parameters()} - params_bb) parameters = [{'params': [param for param in params_new if param.requires_grad]}, {'params': [param for param in params_bb if param.requires_grad], 'lr': 1e-06}] optimizer = optim.AdamW(parameters, lr=5e-05, weight_decay=0.001) lr_sched = optim.lr_scheduler.LambdaLR(optimizer, (lambda n: (1.0 if (n <= 30) else 0.1))) num_epochs = 40 train_transform = Compose([RandomHorizontalFlip(), Resize((512, 512)), ToTensor(), Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])]) val_transform = Compose([Resize((512, 512)), ToTensor(), Normalize(mean=[0.485, 0.456, 0.406], std=[0.229, 0.224, 0.225])]) training_data = coco.DatasetCOCO(datapath=config['coco_path'], fold=fold, transform=train_transform, split='train', shot=5, mode='training', data_list_path=os.path.join(config['workspace_path'], 'data_splits', 'coco'), class_balance=False) validation_data = coco.DatasetCOCO(datapath=config['coco_path'], fold=fold, transform=val_transform, split='val', shot=5, mode='training', data_list_path=os.path.join(config['workspace_path'], 'data_splits', 'coco'), class_balance=False) print('Loaded training set with', len(training_data), 'samples') print('Loaded validation set with', len(validation_data), 'samples') training_sampler = torch.utils.data.RandomSampler(training_data, num_samples=8000, replacement=True) validation_sampler = torch.utils.data.RandomSampler(validation_data, num_samples=2000, replacement=True) training_loader = DataLoader(training_data, sampler=training_sampler, batch_size=8, num_workers=16) validation_loader = DataLoader(validation_data, sampler=validation_sampler, batch_size=20, num_workers=16) trainer = fss_trainer.FSSTrainer(model=model, optimizer=optimizer, lr_sched=lr_sched, train_loader=training_loader, val_loaders=[validation_loader], checkpoint_path=config['checkpoint_path'], visualization_path=os.path.join(config['visualization_path']), save_name=f'{os.path.splitext(os.path.basename(__file__))[0]}_{dataset}_{fold}_{seed}', device=device, checkpoint_epochs=[num_epochs], print_interval=100, visualization_epochs=[2, num_epochs]) if (not restart): trainer.load_checkpoint() trainer.train(num_epochs)
class InducedNormConv2d(nn.Module): def __init__(self, in_channels, out_channels, kernel_size, stride, padding, bias=True, coeff=0.97, domain=2, codomain=2, n_iterations=None, atol=None, rtol=None, **unused_kwargs): del unused_kwargs super(InducedNormConv2d, self).__init__() self.in_channels = in_channels self.out_channels = out_channels self.kernel_size = _pair(kernel_size) self.stride = _pair(stride) self.padding = _pair(padding) self.coeff = coeff self.n_iterations = n_iterations self.domain = domain self.codomain = codomain self.atol = atol self.rtol = rtol self.weight = nn.Parameter(torch.Tensor(out_channels, in_channels, *self.kernel_size)) if bias: self.bias = nn.Parameter(torch.Tensor(out_channels)) else: self.register_parameter('bias', None) self.reset_parameters() self.register_buffer('initialized', torch.tensor(0)) self.register_buffer('spatial_dims', torch.tensor([1.0, 1.0])) self.register_buffer('scale', torch.tensor(0.0)) self.register_buffer('u', self.weight.new_empty(self.out_channels)) self.register_buffer('v', self.weight.new_empty(self.in_channels)) def compute_domain_codomain(self): if torch.is_tensor(self.domain): domain = asym_squash(self.domain) codomain = asym_squash(self.codomain) else: (domain, codomain) = (self.domain, self.codomain) return (domain, codomain) def reset_parameters(self): init.kaiming_uniform_(self.weight, a=math.sqrt(5)) if (self.bias is not None): (fan_in, _) = init._calculate_fan_in_and_fan_out(self.weight) bound = (1 / math.sqrt(fan_in)) init.uniform_(self.bias, (- bound), bound) def _initialize_u_v(self): with torch.no_grad(): (domain, codomain) = self.compute_domain_codomain() if (self.kernel_size == (1, 1)): self.u.resize_(self.out_channels).normal_(0, 1) self.u.copy_(normalize_u(self.u, codomain)) self.v.resize_(self.in_channels).normal_(0, 1) self.v.copy_(normalize_v(self.v, domain)) else: (c, h, w) = (self.in_channels, int(self.spatial_dims[0].item()), int(self.spatial_dims[1].item())) with torch.no_grad(): num_input_dim = ((c * h) * w) self.v.resize_(num_input_dim).normal_(0, 1) self.v.copy_(normalize_v(self.v, domain)) u = F.conv2d(self.v.view(1, c, h, w), self.weight, stride=self.stride, padding=self.padding, bias=None) num_output_dim = (((u.shape[0] * u.shape[1]) * u.shape[2]) * u.shape[3]) self.u.resize_(num_output_dim).normal_(0, 1) self.u.copy_(normalize_u(self.u, codomain)) self.initialized.fill_(1) self.compute_weight(True) best_scale = self.scale.clone() (best_u, best_v) = (self.u.clone(), self.v.clone()) if (not ((domain == 2) and (codomain == 2))): for _ in range(10): if (self.kernel_size == (1, 1)): self.u.copy_(normalize_u(self.weight.new_empty(self.out_channels).normal_(0, 1), codomain)) self.v.copy_(normalize_v(self.weight.new_empty(self.in_channels).normal_(0, 1), domain)) else: self.u.copy_(normalize_u(torch.randn(num_output_dim).to(self.weight), codomain)) self.v.copy_(normalize_v(torch.randn(num_input_dim).to(self.weight), domain)) self.compute_weight(True, n_iterations=200) if (self.scale > best_scale): (best_u, best_v) = (self.u.clone(), self.v.clone()) self.u.copy_(best_u) self.v.copy_(best_v) def compute_one_iter(self): if (not self.initialized): raise ValueError('Layer needs to be initialized first.') (domain, codomain) = self.compute_domain_codomain() if (self.kernel_size == (1, 1)): u = self.u.detach() v = self.v.detach() weight = self.weight.detach().view(self.out_channels, self.in_channels) u = normalize_u(torch.mv(weight, v), codomain) v = normalize_v(torch.mv(weight.t(), u), domain) return torch.dot(u, torch.mv(weight, v)) else: u = self.u.detach() v = self.v.detach() weight = self.weight.detach() (c, h, w) = (self.in_channels, int(self.spatial_dims[0].item()), int(self.spatial_dims[1].item())) u_s = F.conv2d(v.view(1, c, h, w), weight, stride=self.stride, padding=self.padding, bias=None) out_shape = u_s.shape u = normalize_u(u_s.view((- 1)), codomain) v_s = F.conv_transpose2d(u.view(out_shape), weight, stride=self.stride, padding=self.padding, output_padding=0) v = normalize_v(v_s.view((- 1)), domain) weight_v = F.conv2d(v.view(1, c, h, w), weight, stride=self.stride, padding=self.padding, bias=None) return torch.dot(u.view((- 1)), weight_v.view((- 1))) def compute_weight(self, update=True, n_iterations=None, atol=None, rtol=None): if (not self.initialized): self._initialize_u_v() if (self.kernel_size == (1, 1)): return self._compute_weight_1x1(update, n_iterations, atol, rtol) else: return self._compute_weight_kxk(update, n_iterations, atol, rtol) def _compute_weight_1x1(self, update=True, n_iterations=None, atol=None, rtol=None): n_iterations = (self.n_iterations if (n_iterations is None) else n_iterations) atol = (self.atol if (atol is None) else atol) rtol = (self.rtol if (rtol is None) else atol) if ((n_iterations is None) and ((atol is None) or (rtol is None))): raise ValueError('Need one of n_iteration or (atol, rtol).') max_itrs = 200 if (n_iterations is not None): max_itrs = n_iterations u = self.u v = self.v weight = self.weight.view(self.out_channels, self.in_channels) if update: with torch.no_grad(): (domain, codomain) = self.compute_domain_codomain() itrs_used = 0 for _ in range(max_itrs): old_v = v.clone() old_u = u.clone() u = normalize_u(torch.mv(weight, v), codomain, out=u) v = normalize_v(torch.mv(weight.t(), u), domain, out=v) itrs_used = (itrs_used + 1) if ((n_iterations is None) and (atol is not None) and (rtol is not None)): err_u = (torch.norm((u - old_u)) / (u.nelement() ** 0.5)) err_v = (torch.norm((v - old_v)) / (v.nelement() ** 0.5)) tol_u = (atol + (rtol * torch.max(u))) tol_v = (atol + (rtol * torch.max(v))) if ((err_u < tol_u) and (err_v < tol_v)): break if (itrs_used > 0): if ((domain != 1) and (domain != 2)): self.v.copy_(v) if ((codomain != 2) and (codomain != float('inf'))): self.u.copy_(u) u = u.clone() v = v.clone() sigma = torch.dot(u, torch.mv(weight, v)) with torch.no_grad(): self.scale.copy_(sigma) factor = torch.max(torch.ones(1).to(weight.device), (sigma / self.coeff)) weight = (weight / factor) return weight.view(self.out_channels, self.in_channels, 1, 1) def _compute_weight_kxk(self, update=True, n_iterations=None, atol=None, rtol=None): n_iterations = (self.n_iterations if (n_iterations is None) else n_iterations) atol = (self.atol if (atol is None) else atol) rtol = (self.rtol if (rtol is None) else atol) if ((n_iterations is None) and ((atol is None) or (rtol is None))): raise ValueError('Need one of n_iteration or (atol, rtol).') max_itrs = 200 if (n_iterations is not None): max_itrs = n_iterations u = self.u v = self.v weight = self.weight (c, h, w) = (self.in_channels, int(self.spatial_dims[0].item()), int(self.spatial_dims[1].item())) if update: with torch.no_grad(): (domain, codomain) = self.compute_domain_codomain() itrs_used = 0 for _ in range(max_itrs): old_u = u.clone() old_v = v.clone() u_s = F.conv2d(v.view(1, c, h, w), weight, stride=self.stride, padding=self.padding, bias=None) out_shape = u_s.shape u = normalize_u(u_s.view((- 1)), codomain, out=u) v_s = F.conv_transpose2d(u.view(out_shape), weight, stride=self.stride, padding=self.padding, output_padding=0) v = normalize_v(v_s.view((- 1)), domain, out=v) itrs_used = (itrs_used + 1) if ((n_iterations is None) and (atol is not None) and (rtol is not None)): err_u = (torch.norm((u - old_u)) / (u.nelement() ** 0.5)) err_v = (torch.norm((v - old_v)) / (v.nelement() ** 0.5)) tol_u = (atol + (rtol * torch.max(u))) tol_v = (atol + (rtol * torch.max(v))) if ((err_u < tol_u) and (err_v < tol_v)): break if (itrs_used > 0): if (domain != 2): self.v.copy_(v) if (codomain != 2): self.u.copy_(u) v = v.clone() u = u.clone() weight_v = F.conv2d(v.view(1, c, h, w), weight, stride=self.stride, padding=self.padding, bias=None) weight_v = weight_v.view((- 1)) sigma = torch.dot(u.view((- 1)), weight_v) with torch.no_grad(): self.scale.copy_(sigma) factor = torch.max(torch.ones(1).to(weight.device), (sigma / self.coeff)) weight = (weight / factor) return weight def forward(self, input): if (not self.initialized): self.spatial_dims.copy_(torch.tensor(input.shape[2:4]).to(self.spatial_dims)) weight = self.compute_weight(update=False) return F.conv2d(input, weight, self.bias, self.stride, self.padding, 1, 1) def extra_repr(self): (domain, codomain) = self.compute_domain_codomain() s = '{in_channels}, {out_channels}, kernel_size={kernel_size}, stride={stride}' if (self.padding != ((0,) * len(self.padding))): s += ', padding={padding}' if (self.bias is None): s += ', bias=False' s += ', coeff={}, domain={:.2f}, codomain={:.2f}, n_iters={}, atol={}, rtol={}, learnable_ord={}'.format(self.coeff, domain, codomain, self.n_iterations, self.atol, self.rtol, torch.is_tensor(self.domain)) return s.format(**self.__dict__)
def flatten_observation_spaces(observation_spaces, observation_excluded=()): if (not isinstance(observation_excluded, (list, tuple))): observation_excluded = [observation_excluded] lower_bound = [] upper_bound = [] for (key, value) in observation_spaces.spaces.items(): if (key not in observation_excluded): lower_bound.append(np.asarray(value.low).flatten()) upper_bound.append(np.asarray(value.high).flatten()) lower_bound = np.concatenate(lower_bound) upper_bound = np.concatenate(upper_bound) observation_space = spaces.Box(np.array(lower_bound), np.array(upper_bound), dtype=np.float32) if (not observation_excluded): return observation_space else: observation_spaces_after_flatten = {'other': observation_space} for key in observation_excluded: observation_spaces_after_flatten[key] = observation_spaces[key] return spaces.Dict(observation_spaces_after_flatten)
class TestSequeneceGenerator(TestSequenceGeneratorBase): def setUp(self): (self.tgt_dict, self.w1, self.w2, src_tokens, src_lengths, self.model) = test_utils.sequence_generator_setup() self.sample = {'net_input': {'src_tokens': src_tokens, 'src_lengths': src_lengths}} def test_with_normalization(self): generator = SequenceGenerator([self.model], self.tgt_dict, beam_size=2) hypos = generator.forward(self.sample) (eos, w1, w2) = (self.tgt_dict.eos(), self.w1, self.w2) self.assertHypoTokens(hypos[0][0], [w1, eos]) self.assertHypoScore(hypos[0][0], [0.9, 1.0]) self.assertHypoTokens(hypos[0][1], [w2, w1, w2, eos]) self.assertHypoScore(hypos[0][1], [0.1, 0.9, 0.9, 1.0]) self.assertHypoTokens(hypos[1][0], [w1, w2, w1, eos]) self.assertHypoScore(hypos[1][0], [0.7, 0.4, 0.4, 1.0]) self.assertHypoTokens(hypos[1][1], [w1, w2, eos]) self.assertHypoScore(hypos[1][1], [0.7, 0.4, 0.6]) def test_without_normalization(self): generator = SequenceGenerator([self.model], self.tgt_dict, beam_size=2, normalize_scores=False) hypos = generator.forward(self.sample) (eos, w1, w2) = (self.tgt_dict.eos(), self.w1, self.w2) self.assertHypoTokens(hypos[0][0], [w1, eos]) self.assertHypoScore(hypos[0][0], [0.9, 1.0], normalized=False) self.assertHypoTokens(hypos[0][1], [w2, w1, w2, eos]) self.assertHypoScore(hypos[0][1], [0.1, 0.9, 0.9, 1.0], normalized=False) self.assertHypoTokens(hypos[1][0], [w1, w2, eos]) self.assertHypoScore(hypos[1][0], [0.7, 0.4, 0.6], normalized=False) self.assertHypoTokens(hypos[1][1], [w1, w2, w1, eos]) self.assertHypoScore(hypos[1][1], [0.7, 0.4, 0.4, 1.0], normalized=False) def test_with_lenpen_favoring_short_hypos(self): lenpen = 0.6 generator = SequenceGenerator([self.model], self.tgt_dict, beam_size=2, len_penalty=lenpen) hypos = generator.forward(self.sample) (eos, w1, w2) = (self.tgt_dict.eos(), self.w1, self.w2) self.assertHypoTokens(hypos[0][0], [w1, eos]) self.assertHypoScore(hypos[0][0], [0.9, 1.0], lenpen=lenpen) self.assertHypoTokens(hypos[0][1], [w2, w1, w2, eos]) self.assertHypoScore(hypos[0][1], [0.1, 0.9, 0.9, 1.0], lenpen=lenpen) self.assertHypoTokens(hypos[1][0], [w1, w2, eos]) self.assertHypoScore(hypos[1][0], [0.7, 0.4, 0.6], lenpen=lenpen) self.assertHypoTokens(hypos[1][1], [w1, w2, w1, eos]) self.assertHypoScore(hypos[1][1], [0.7, 0.4, 0.4, 1.0], lenpen=lenpen) def test_with_lenpen_favoring_long_hypos(self): lenpen = 5.0 generator = SequenceGenerator([self.model], self.tgt_dict, beam_size=2, len_penalty=lenpen) hypos = generator.forward(self.sample) (eos, w1, w2) = (self.tgt_dict.eos(), self.w1, self.w2) self.assertHypoTokens(hypos[0][0], [w2, w1, w2, eos]) self.assertHypoScore(hypos[0][0], [0.1, 0.9, 0.9, 1.0], lenpen=lenpen) self.assertHypoTokens(hypos[0][1], [w1, eos]) self.assertHypoScore(hypos[0][1], [0.9, 1.0], lenpen=lenpen) self.assertHypoTokens(hypos[1][0], [w1, w2, w1, eos]) self.assertHypoScore(hypos[1][0], [0.7, 0.4, 0.4, 1.0], lenpen=lenpen) self.assertHypoTokens(hypos[1][1], [w1, w2, eos]) self.assertHypoScore(hypos[1][1], [0.7, 0.4, 0.6], lenpen=lenpen) def test_maxlen(self): generator = SequenceGenerator([self.model], self.tgt_dict, beam_size=2, max_len_b=2) hypos = generator.forward(self.sample) (eos, w1, w2) = (self.tgt_dict.eos(), self.w1, self.w2) self.assertHypoTokens(hypos[0][0], [w1, eos]) self.assertHypoScore(hypos[0][0], [0.9, 1.0]) self.assertHypoTokens(hypos[0][1], [w2, w2, eos]) self.assertHypoScore(hypos[0][1], [0.1, 0.1, 0.6]) self.assertHypoTokens(hypos[1][0], [w1, w2, eos]) self.assertHypoScore(hypos[1][0], [0.7, 0.4, 0.6]) self.assertHypoTokens(hypos[1][1], [w2, w2, eos]) self.assertHypoScore(hypos[1][1], [0.3, 0.9, 0.01]) def test_encoder_with_different_output_len(self): args = self.model.encoder.args task = test_utils.TestTranslationTask.setup_task(args, self.tgt_dict, self.tgt_dict) reshaping_model = test_utils.TestReshapingModel.build_model(args, task) generator = SequenceGenerator([reshaping_model], self.tgt_dict, beam_size=2, max_len_b=2) hypos = generator.forward(self.sample) for sent in [0, 1]: for beam in [0, 1]: assert (hypos[sent][beam]['attention'] is not None) def test_generation_with_additional_input(self): args = self.model.encoder.args task = test_utils.TestTranslationTask.setup_task(args, self.tgt_dict, self.tgt_dict) add_input_model = test_utils.TestAdditionalInputModel.build_model(args, task) generator = SequenceGenerator([add_input_model], self.tgt_dict, beam_size=2) sample = self.sample.copy() sample['net_input']['fancy_other_input'] = sample['net_input']['src_tokens'] hypos = generator.forward(self.sample) (eos, w1, w2) = (self.tgt_dict.eos(), self.w1, self.w2) self.assertHypoTokens(hypos[0][0], [w1, eos]) self.assertHypoScore(hypos[0][0], [0.9, 1.0])
def val(test_loader, model, epoch, save_path, writer): global best_mae, best_epoch model.eval() with torch.no_grad(): mae_sum = 0 for i in range(test_loader.size): (images, shorts, gt, name, scene) = test_loader.load_data() inputs = torch.cat([images, shorts], 2) preds = model(inputs) gt = np.asarray(gt, np.float32) gt /= (gt.max() + 1e-08) images = [x.cuda() for x in images] res = F.upsample(res, size=gt.shape, mode='bilinear', align_corners=False) res = res.sigmoid().data.cpu().numpy().squeeze() res = ((res - res.min()) / ((res.max() - res.min()) + 1e-08)) mae_sum += ((np.sum(np.abs((res - gt))) * 1.0) / (gt.shape[0] * gt.shape[1])) mae = (mae_sum / test_loader.size) writer.add_scalar('MAE', torch.tensor(mae), global_step=epoch) print('Epoch: {}, MAE: {}, bestMAE: {}, bestEpoch: {}.'.format(epoch, mae, best_mae, best_epoch)) if (epoch == 1): best_mae = mae elif (mae < best_mae): best_mae = mae best_epoch = epoch torch.save(model.state_dict(), (save_path + 'Net_epoch_best.pth')) print('Save state_dict successfully! Best epoch:{}.'.format(epoch)) logging.info('[Val Info]:Epoch:{} MAE:{} bestEpoch:{} bestMAE:{}'.format(epoch, mae, best_epoch, best_mae))
def check_labels(labels, estimator): label_encoder = None if (estimator._estimator_type == 'classifier'): if np.issubdtype(type(labels[0]), np.str_): label_encoder = LabelEncoder() label_encoder.fit(labels) labels = label_encoder.transform(labels) y_type = type_of_target(labels) if (y_type not in ['binary', 'multiclass']): raise ValueError(('Unknown label type: %r' % y_type)) elif (type(labels[0]) not in (np.float32, np.float64)): logger.info('The labels have been converted in float64') labels = labels.astype('float64') _assert_all_finite(labels) if ((estimator._estimator_type == 'classifier') and (len(np.unique(labels)) == 1)): raise ValueError("Classifier can't train when only one class is present.") return (labels, label_encoder)
def test_nested_typechange(conf_scope): cfg = conf_scope({'f': {'a': 10}}) assert (cfg.typechanged == {'f.a': (type('a'), int)})
def image_label(loader, model, threshold=0.9, out_dir=None): out_path = osp.join(out_dir, 'pseudo_label.txt') print('Pseudo Labeling to ', out_path) iter_label = iter(loader['target_label']) with torch.no_grad(): with open(out_path, 'w') as f: for i in range(len(loader['target_label'])): (inputs, labels, paths) = iter_label.next() inputs = inputs.cuda() (_, outputs) = model(inputs) softmax_outputs = nn.Softmax(dim=1)(outputs) (maxpred, pseudo_labels) = torch.max(softmax_outputs, dim=1) pseudo_labels[(maxpred < threshold)] = (- 1) for (path, label) in zip(paths, pseudo_labels): f.write((((path + ' ') + str(label.item())) + '\n')) return out_path
class TFDebertaModel(metaclass=DummyObject): _backends = ['tf'] def __init__(self, *args, **kwargs): requires_backends(self, ['tf'])
class MultiEdgeGraphPairwiseFormatter(BaseGraphFormatter): def __init__(self, config, name='MultiEdgeGraphPairwiseFormatter'): self.name = name self.disable_tqdm = config.disable_tqdm self.config = config BaseGraphFormatter.__init__(self, config, name) def format(self, item_json, vocab_dicts): (token_vd, node_vd, target_vd, word_vd) = vocab_dicts datapoint = self.datapoint_class() datapoint.tgt = item_json['target'] dgl_graph_one = self.get_graph(item_json['item_1'], token_vd, node_vd) datapoint.function_one_graph = dgl_graph_one dgl_graph_two = self.get_graph(item_json['item_2'], token_vd, node_vd) datapoint.function_two_graph = dgl_graph_two return datapoint def get_graph(self, cur_item, token_vd, node_vd): graph = self._convert_to_multi_edge_dglgraph(cur_item['jsgraph'], token_vd, node_vd) return graph
class DecentralizedDistributedMixin(): def _get_advantages_distributed(self, rollouts: RolloutStorage) -> torch.Tensor: advantages = (rollouts.returns[:(- 1)] - rollouts.value_preds[:(- 1)]) if (not self.use_normalized_advantage): return advantages (mean, var) = distributed_mean_and_var(advantages) return ((advantages - mean) / (var.sqrt() + EPS_PPO)) def init_distributed(self, find_unused_params: bool=True) -> None: class Guard(): def __init__(self, model, device): if torch.cuda.is_available(): self.ddp = torch.nn.parallel.DistributedDataParallel(model, device_ids=[device], output_device=device) else: self.ddp = torch.nn.parallel.DistributedDataParallel(model) self._ddp_hooks = Guard(self.actor_critic, self.device) self.get_advantages = self._get_advantages_distributed self.reducer = self._ddp_hooks.ddp.reducer self.find_unused_params = find_unused_params def before_backward(self, loss): super().before_backward(loss) if self.find_unused_params: self.reducer.prepare_for_backward([loss]) else: self.reducer.prepare_for_backward([])
def prepare_keys_div2k(folder_path): print('Reading image path list ...') img_path_list = sorted(list(scandir(folder_path, suffix='png', recursive=False))) keys = [img_path.split('.png')[0] for img_path in sorted(img_path_list)] return (img_path_list, keys)
class ReversibleSequence(nn.Module): def __init__(self, blocks, args_route={}): super().__init__() self.args_route = args_route self.blocks = nn.ModuleList([ReversibleBlock(f=f, g=g) for (f, g) in blocks]) def forward(self, x, **kwargs): x = torch.cat([x, x], dim=(- 1)) blocks = self.blocks args = route_args(self.args_route, kwargs, len(blocks)) args = list(map((lambda x: {'f_args': x[0], 'g_args': x[1]}), args)) out = _ReversibleFunction.apply(x, blocks, args) return torch.stack(out.chunk(2, dim=(- 1))).mean(dim=0)
class ResContextBlock(nn.Module): def __init__(self, in_filters, out_filters, kernel_size=(3, 3, 3), stride=1, indice_key=None): super(ResContextBlock, self).__init__() self.conv1 = conv1x3(in_filters, out_filters, indice_key=(indice_key + 'bef')) self.bn0 = nn.BatchNorm1d(out_filters) self.act1 = nn.LeakyReLU() self.conv1_2 = conv3x1(out_filters, out_filters, indice_key=(indice_key + 'bef')) self.bn0_2 = nn.BatchNorm1d(out_filters) self.act1_2 = nn.LeakyReLU() self.conv2 = conv3x1(in_filters, out_filters, indice_key=(indice_key + 'bef')) self.act2 = nn.LeakyReLU() self.bn1 = nn.BatchNorm1d(out_filters) self.conv3 = conv1x3(out_filters, out_filters, indice_key=(indice_key + 'bef')) self.act3 = nn.LeakyReLU() self.bn2 = nn.BatchNorm1d(out_filters) self.weight_initialization() def weight_initialization(self): for m in self.modules(): if isinstance(m, nn.BatchNorm1d): nn.init.constant_(m.weight, 1) nn.init.constant_(m.bias, 0) def forward(self, x): shortcut = self.conv1(x) shortcut.features = self.act1(shortcut.features) shortcut.features = self.bn0(shortcut.features) shortcut = self.conv1_2(shortcut) shortcut.features = self.act1_2(shortcut.features) shortcut.features = self.bn0_2(shortcut.features) resA = self.conv2(x) resA.features = self.act2(resA.features) resA.features = self.bn1(resA.features) resA = self.conv3(resA) resA.features = self.act3(resA.features) resA.features = self.bn2(resA.features) resA.features = (resA.features + shortcut.features) return resA
class Device(): du = Device_Util() speed_distri = None try: with open('speed_distri.json', 'r') as f: speed_distri = json.load(f) except FileNotFoundError as e: speed_distri = None logger.warn("no user's network speed trace was found, set all communication time to 0.0s") def __init__(self, cfg, model_size=0): self.device_model = None self.cfg = cfg self.sampled_train_time_per_batch = 0 self.sampled_download_time = 0 self.sampled_upload_time = 0 self.sample_count = 10 self.model_size = (model_size / 1024) if ((cfg.behav_hete == False) and (cfg.hard_hete == False)): self.model_size = 0 if (Device.speed_distri == None): self.model_size = 0 self.upload_speed_u = 1.0 self.upload_speed_sigma = 0.0 self.download_speed_u = 1.0 self.download_speed_sigma = 0.0 else: if (cfg.hard_hete == False): guid = list(Device.speed_distri.keys())[(cfg.seed % len(Device.speed_distri))] else: guid = random.sample(list(Device.speed_distri.keys()), 1)[0] self.download_speed_u = Device.speed_distri[guid]['down_u'] self.download_speed_sigma = Device.speed_distri[guid]['down_sigma'] self.upload_speed_u = Device.speed_distri[guid]['up_u'] self.upload_speed_sigma = Device.speed_distri[guid]['up_sigma'] Device.du.set_model(cfg.model) Device.du.set_dataset(cfg.dataset) def set_device_model(self, real_device_model, client_id): device_train_times_per_batch = [] self.device_model = Device.du.transfer(real_device_model) for i in range(self.sample_count): device_train_times_per_batch.append(Device.du.get_train_time_per_batch(self.device_model)) self.sampled_train_time_per_batch = np.mean(device_train_times_per_batch) return device_train_times_per_batch def set_device_model_weakDeviceToCertainClass(self, real_device_model, label): self.device_model = Device.du.unknown_weakDeviceToCertainClass(label) def get_upload_time(self): if (self.model_size == 0.0): return 0.0 upload_speed = np.random.normal(self.upload_speed_u, self.upload_speed_sigma) while (upload_speed < 0): upload_speed = np.random.normal(self.upload_speed_u, self.upload_speed_sigma) upload_time = (self.model_size / upload_speed) return float(upload_time) def get_download_time(self): if (self.model_size == 0.0): return 0.0 download_speed = np.random.normal(self.download_speed_u, self.download_speed_sigma) while (download_speed < 0): download_speed = np.random.normal(self.download_speed_u, self.download_speed_sigma) download_time = (self.model_size / download_speed) return float(download_time) def get_expected_download_time(self): if (self.model_size == 0.0): return 0.0 download_speed = self.download_speed_u while (download_speed < 0): download_speed = self.download_speed_u download_time = (self.model_size / download_speed) return float(download_time) def get_expected_upload_time(self): if (self.model_size == 0.0): return 0.0 upload_speed = self.upload_speed_u while (upload_speed < 0): upload_speed = self.upload_speed_u upload_time = (self.model_size / upload_speed) return float(upload_time) def get_train_time_and_train_time_per_batch(self, num_sample, batch_size, num_epoch): if (self.device_model == None): assert False return Device.du.get_train_time_and_train_time_per_batch_and_train_time_per_epoch(self.device_model, num_sample, batch_size, num_epoch)[:(- 1)] def get_train_time_and_train_time_per_batch_and_train_time_per_epoch(self, num_sample, batch_size, num_epoch): if (self.device_model == None): assert False return Device.du.get_train_time_and_train_time_per_batch_and_train_time_per_epoch(self.device_model, num_sample, batch_size, num_epoch)
class ESPNet(nn.Module): def __init__(self, in_channels, num_classes): super().__init__() self.input1 = InputProjectionA(1) self.input2 = InputProjectionA(1) initial = 16 config = [32, 128, 256, 256] reps = [2, 2, 3] self.level0 = CBR(in_channels, initial, 7, 2) self.level1 = nn.ModuleList() for i in range(reps[0]): if (i == 0): self.level1.append(DilatedParllelResidualBlockB1(initial, config[0])) else: self.level1.append(DilatedParllelResidualBlockB1(config[0], config[0])) self.level2 = DilatedParllelResidualBlockB1(config[0], config[1], stride=2) self.level_2 = nn.ModuleList() for i in range(0, reps[1]): self.level_2.append(DilatedParllelResidualBlockB1(config[1], config[1])) self.level3_0 = DilatedParllelResidualBlockB1(config[1], config[2], stride=2) self.level_3 = nn.ModuleList() for i in range(0, reps[2]): self.level_3.append(DilatedParllelResidualBlockB1(config[2], config[2])) self.up_l3_l2 = UpSampler(config[2], config[1]) self.merge_l2 = DilatedParllelResidualBlockB1((2 * config[1]), config[1]) self.dec_l2 = nn.ModuleList() for i in range(0, reps[0]): self.dec_l2.append(DilatedParllelResidualBlockB1(config[1], config[1])) self.up_l2_l1 = UpSampler(config[1], config[0]) self.merge_l1 = DilatedParllelResidualBlockB1((2 * config[0]), config[0]) self.dec_l1 = nn.ModuleList() for i in range(0, reps[0]): self.dec_l1.append(DilatedParllelResidualBlockB1(config[0], config[0])) self.dec_l1.append(CBR(config[0], num_classes, 3, 1)) self.dec_l1.append(ASPBlock(num_classes, num_classes)) self.pspModules = nn.ModuleList() scales = [0.2, 0.4, 0.6, 0.8] for sc in scales: self.pspModules.append(PSPDec(num_classes, num_classes, sc)) self.classifier = self.classifier = nn.Sequential(CBR(((len(scales) + 1) * num_classes), num_classes, 3, 1), ASPBlock(num_classes, num_classes), nn.Upsample(scale_factor=2, mode='trilinear', align_corners=True), CBR(num_classes, num_classes, 7, 1), C(num_classes, num_classes, 1, 1)) for m in self.modules(): if isinstance(m, nn.Conv3d): n = (((m.kernel_size[0] * m.kernel_size[1]) * m.kernel_size[2]) * m.out_channels) m.weight.data.normal_(0, math.sqrt((2.0 / n))) if isinstance(m, nn.ConvTranspose3d): n = (((m.kernel_size[0] * m.kernel_size[1]) * m.kernel_size[2]) * m.out_channels) m.weight.data.normal_(0, math.sqrt((2.0 / n))) elif isinstance(m, nn.BatchNorm3d): m.weight.data.fill_(1) m.bias.data.zero_() def forward(self, input1, inp_res=(128, 128, 128), inpSt2=False): dim0 = input1.size(2) dim1 = input1.size(3) dim2 = input1.size(4) if (self.training or (inp_res is None)): inp_res = ((math.ceil((dim0 / 8)) * 8), (math.ceil((dim1 / 8)) * 8), (math.ceil((dim2 / 8)) * 8)) if inp_res: input1 = F.adaptive_avg_pool3d(input1, output_size=inp_res) out_l0 = self.level0(input1) for (i, layer) in enumerate(self.level1): if (i == 0): out_l1 = layer(out_l0) else: out_l1 = layer(out_l1) out_l2_down = self.level2(out_l1) for (i, layer) in enumerate(self.level_2): if (i == 0): out_l2 = layer(out_l2_down) else: out_l2 = layer(out_l2) del out_l2_down out_l3_down = self.level3_0(out_l2) for (i, layer) in enumerate(self.level_3): if (i == 0): out_l3 = layer(out_l3_down) else: out_l3 = layer(out_l3) del out_l3_down dec_l3_l2 = self.up_l3_l2(out_l3) merge_l2 = self.merge_l2(torch.cat([dec_l3_l2, out_l2], 1)) for (i, layer) in enumerate(self.dec_l2): if (i == 0): dec_l2 = layer(merge_l2) else: dec_l2 = layer(dec_l2) dec_l2_l1 = self.up_l2_l1(dec_l2) merge_l1 = self.merge_l1(torch.cat([dec_l2_l1, out_l1], 1)) for (i, layer) in enumerate(self.dec_l1): if (i == 0): dec_l1 = layer(merge_l1) else: dec_l1 = layer(dec_l1) psp_outs = dec_l1.clone() for layer in self.pspModules: out_psp = layer(dec_l1) psp_outs = torch.cat([psp_outs, out_psp], 1) decoded = self.classifier(psp_outs) return F.upsample(decoded, size=(dim0, dim1, dim2), mode='trilinear', align_corners=True)
def get_checkpoint_from_config_class(config_class): checkpoint = None config_source = inspect.getsource(config_class) checkpoints = _re_checkpoint.findall(config_source) for (ckpt_name, ckpt_link) in checkpoints: if ckpt_link.endswith('/'): ckpt_link = ckpt_link[:(- 1)] ckpt_link_from_name = f' if (ckpt_link == ckpt_link_from_name): checkpoint = ckpt_name break return checkpoint
def prepare_keys_vimeo90k(folder_path, train_list_path, mode): print('Reading image path list ...') with open(train_list_path, 'r') as fin: train_list = [line.strip() for line in fin] img_path_list = [] keys = [] for line in train_list: (folder, sub_folder) = line.split('/') img_path_list.extend([osp.join(folder, sub_folder, f'im{(j + 1)}.png') for j in range(7)]) keys.extend([f'{folder}/{sub_folder}/im{(j + 1)}' for j in range(7)]) if (mode == 'gt'): print('Only keep the 4th frame for the gt mode.') img_path_list = [v for v in img_path_list if v.endswith('im4.png')] keys = [v for v in keys if v.endswith('/im4')] return (img_path_list, keys)
class TextSampler(Sampler): def __init__(self, lengths, batch_size, n_buckets, shuffle=False): self.lengths = lengths self.batch_size = batch_size self.shuffle = shuffle (self.sizes, self.buckets) = kmeans(x=lengths, k=n_buckets) self.chunks = [max(((size * len(bucket)) // self.batch_size), 1) for (size, bucket) in zip(self.sizes, self.buckets)] def __iter__(self): range_fn = (torch.randperm if self.shuffle else torch.arange) for i in range_fn(len(self.buckets)): for batch in range_fn(len(self.buckets[i])).chunk(self.chunks[i]): (yield [self.buckets[i][j] for j in batch.tolist()]) def __len__(self): return sum(self.chunks)
def cal_acc(zeros, ones): accuracy = 0.0 for example in zeros: if (not np.isnan(example[0])): if (example[0] < 0.5): accuracy += 1.0 for example in ones: if (not np.isnan(example[0])): if (example[0] > 0.5): accuracy += 1.0 accuracy = (accuracy / float((len(zeros) + len(ones)))) print(('The accuracy of the discriminator is: ' + str(accuracy))) return accuracy
def dump2json(ofn, data, force=False): if (os.path.exists(ofn) and (not force)): return def default(obj): if isinstance(obj, np.int32): return int(obj) elif isinstance(obj, np.int64): return int(obj) elif isinstance(obj, np.float32): return float(obj) elif (isinstance(obj, set) or isinstance(obj, np.ndarray)): return list(obj) else: raise TypeError('Unserializable object {} of type {}'.format(obj, type(obj))) with open(ofn, 'w') as of: json.dump(data, of, default=default)
class EncoderLayer(nn.Module): def __init__(self, d_model, self_attn, feed_forward=None, use_residual=False, dropout=0.1): super(EncoderLayer, self).__init__() self.self_attn = self_attn self.feed_forward = feed_forward self.use_residual = use_residual if use_residual: self.sublayer = nn.ModuleList([SublayerConnection(d_model, dropout) for _ in range(2)]) self.d_model = d_model def forward(self, x, mask): if self.use_residual: x = self.sublayer[0](x, (lambda x: self.self_attn(x, x, x, mask))) if (self.feed_forward is not None): return self.sublayer[1](x, self.feed_forward) else: return x else: return self.self_attn(x, x, x, mask)
class RePU(nn.ReLU): def __init__(self, n): super(RePU, self).__init__() self.n = n def forward(self, x: torch.Tensor): return (torch.relu(x) ** self.n)
def main(): args = parse_args() dataset_path = args.dataset_path if (args.out_dir is None): out_dir = osp.join('data', 'loveDA') else: out_dir = args.out_dir print('Making directories...') mmcv.mkdir_or_exist(out_dir) mmcv.mkdir_or_exist(osp.join(out_dir, 'img_dir')) mmcv.mkdir_or_exist(osp.join(out_dir, 'img_dir', 'train')) mmcv.mkdir_or_exist(osp.join(out_dir, 'img_dir', 'val')) mmcv.mkdir_or_exist(osp.join(out_dir, 'img_dir', 'test')) mmcv.mkdir_or_exist(osp.join(out_dir, 'ann_dir')) mmcv.mkdir_or_exist(osp.join(out_dir, 'ann_dir', 'train')) mmcv.mkdir_or_exist(osp.join(out_dir, 'ann_dir', 'val')) assert ('Train.zip' in os.listdir(dataset_path)), 'Train.zip is not in {}'.format(dataset_path) assert ('Val.zip' in os.listdir(dataset_path)), 'Val.zip is not in {}'.format(dataset_path) assert ('Test.zip' in os.listdir(dataset_path)), 'Test.zip is not in {}'.format(dataset_path) with tempfile.TemporaryDirectory(dir=args.tmp_dir) as tmp_dir: for dataset in ['Train', 'Val', 'Test']: zip_file = zipfile.ZipFile(os.path.join(dataset_path, (dataset + '.zip'))) zip_file.extractall(tmp_dir) data_type = dataset.lower() for location in ['Rural', 'Urban']: for image_type in ['images_png', 'masks_png']: if (image_type == 'images_png'): dst = osp.join(out_dir, 'img_dir', data_type) else: dst = osp.join(out_dir, 'ann_dir', data_type) if ((dataset == 'Test') and (image_type == 'masks_png')): continue else: src_dir = osp.join(tmp_dir, dataset, location, image_type) src_lst = os.listdir(src_dir) for file in src_lst: shutil.move(osp.join(src_dir, file), dst) print('Removing the temporary files...') print('Done!')
def main(args, logger): logger.info(args) fix_seed_for_reproducability(args.seed) t_start = t.time() (model, optimizer, classifier) = get_model_and_optimizer(args, logger) (inv_list, eqv_list) = get_transform_params(args) trainset = get_dataset(args, mode='train', inv_list=inv_list, eqv_list=eqv_list) trainloader = torch.utils.data.DataLoader(trainset, batch_size=args.batch_size_cluster, shuffle=True, num_workers=args.num_workers, pin_memory=True, collate_fn=collate_train_baseline, worker_init_fn=worker_init_fn(args.seed)) testset = get_dataset(args, mode='train_val') testloader = torch.utils.data.DataLoader(testset, batch_size=args.batch_size_test, shuffle=False, num_workers=args.num_workers, pin_memory=True, collate_fn=collate_eval, worker_init_fn=worker_init_fn(args.seed)) (_, _) = evaluate(args, logger, testloader, classifier, model) if (not args.eval_only): for epoch in range(args.start_epoch, args.num_epoch): trainloader.dataset.mode = 'compute' trainloader.dataset.reshuffle() adjust_learning_rate(optimizer, epoch, args) logger.info('\n [Epoch {}] \n'.format(epoch)) logger.info('Start computing centroids.') t1 = t.time() (centroids, kmloss) = run_mini_batch_kmeans(args, logger, trainloader, model, view=1) logger.info('-Centroids ready. [{}]\n'.format(get_datetime((int(t.time()) - int(t1))))) t2 = t.time() weight = compute_labels(args, logger, trainloader, model, centroids, view=1) logger.info('-Cluster labels ready. [{}]\n'.format(get_datetime((int(t.time()) - int(t2))))) criterion = torch.nn.CrossEntropyLoss(weight=weight).cuda() classifier = initialize_classifier(args) if args.nonparametric: classifier.module.weight.data = centroids.unsqueeze((- 1)).unsqueeze((- 1)) freeze_all(classifier) if args.nonparametric: optimizer_loop = None elif (args.optim_type == 'SGD'): optimizer_loop = torch.optim.SGD(filter((lambda x: x.requires_grad), classifier.module.parameters()), lr=args.lr, momentum=args.momentum, weight_decay=args.weight_decay) elif (args.optim_type == 'Adam'): optimizer_loop = torch.optim.Adam(filter((lambda x: x.requires_grad), classifier.module.parameters()), lr=args.lr) trainset.mode = 'baseline_train' trainset.labeldir = args.save_model_path trainloader_loop = torch.utils.data.DataLoader(trainset, batch_size=args.batch_size_train, shuffle=True, num_workers=args.num_workers, pin_memory=True, collate_fn=collate_train_baseline, worker_init_fn=worker_init_fn(args.seed)) logger.info('Start training ...') train_loss = train(args, logger, trainloader_loop, model, classifier, criterion, optimizer, optimizer_loop) (acc, res) = evaluate(args, logger, testloader, classifier, model) logger.info(' Epoch [{}] '.format(epoch)) logger.info(' Time total : [{}].'.format(get_datetime((int(t.time()) - int(t1))))) logger.info(' K-Means loss : {:.5f}.'.format(kmloss)) logger.info(' Training loss : {:.5f}.'.format(train_loss)) logger.info(' ACC: {:.4f} | mIoU: {:.4f}'.format(acc, res['mean_iou'])) logger.info('\n') torch.save({'epoch': (epoch + 1), 'args': args, 'state_dict': model.state_dict(), 'classifier1_state_dict': classifier.state_dict(), 'optimizer': optimizer.state_dict()}, os.path.join(args.save_model_path, 'checkpoint_{}.pth.tar'.format(epoch))) torch.save({'epoch': (epoch + 1), 'args': args, 'state_dict': model.state_dict(), 'classifier1_state_dict': classifier.state_dict(), 'optimizer': optimizer.state_dict()}, os.path.join(args.save_model_path, 'checkpoint.pth.tar')) trainset = get_dataset(args, mode='eval_val') trainloader = torch.utils.data.DataLoader(trainset, batch_size=args.batch_size_cluster, shuffle=True, num_workers=args.num_workers, pin_memory=True, collate_fn=collate_train_baseline, worker_init_fn=worker_init_fn(args.seed)) testset = get_dataset(args, mode='eval_test') testloader = torch.utils.data.DataLoader(testset, batch_size=args.batch_size_test, shuffle=False, num_workers=args.num_workers, pin_memory=True, collate_fn=collate_eval, worker_init_fn=worker_init_fn(args.seed)) acc_list_new = [] res_list_new = [] logger.info('Start computing centroids.') if (args.repeats > 0): for _ in range(args.repeats): t1 = t.time() (centroids, kmloss) = run_mini_batch_kmeans(args, logger, trainloader, model, view=(- 1)) logger.info('-Centroids ready. [Loss: {:.5f}/ Time: {}]\n'.format(kmloss, get_datetime((int(t.time()) - int(t1))))) classifier = initialize_classifier(args) classifier.module.weight.data = centroids.unsqueeze((- 1)).unsqueeze((- 1)) freeze_all(classifier) (acc_new, res_new) = evaluate(args, logger, testloader, classifier, model) acc_list_new.append(acc_new) res_list_new.append(res_new) else: (acc_new, res_new) = evaluate(args, logger, testloader, classifier, model) acc_list_new.append(acc_new) res_list_new.append(res_new) logger.info('Average overall pixel accuracy [NEW] : {} +/- {}.'.format(round(np.mean(acc_list_new), 2), np.std(acc_list_new))) logger.info('Average mIoU [NEW] : {:.3f} +/- {:.3f}. '.format(np.mean([res['mean_iou'] for res in res_list_new]), np.std([res['mean_iou'] for res in res_list_new]))) logger.info('Experiment done. [{}]\n'.format(get_datetime((int(t.time()) - int(t_start)))))
def reduce_param_groups(params: List[Dict[(str, Any)]]) -> List[Dict[(str, Any)]]: params = _expand_param_groups(params) groups = defaultdict(list) for item in params: cur_params = tuple(((x, y) for (x, y) in item.items() if (x != 'params'))) groups[cur_params].extend(item['params']) ret = [] for (param_keys, param_values) in groups.items(): cur = {kv[0]: kv[1] for kv in param_keys} cur['params'] = param_values ret.append(cur) return ret
def add_dmc_args(parser): parser.add_argument('--domain_name', type=str, default='fish') parser.add_argument('--task_name', type=str, default='swim') parser.add_argument('--from_pixels', action='store_true', help='Use image observations') parser.add_argument('--height', type=int, default=84) parser.add_argument('--width', type=int, default=84) parser.add_argument('--camera_id', type=int, default=0) parser.add_argument('--frame_skip', type=int, default=1) parser.add_argument('--frame_stack', type=int, default=3) parser.add_argument('--channels_last', action='store_true') parser.add_argument('--rgb', action='store_true') parser.add_argument('--seed', type=int, default=231)
def dumb_css_parser(data): data += ';' importIndex = data.find('') while (importIndex != (- 1)): data = (data[0:importIndex] + data[(data.find(';', importIndex) + 1):]) importIndex = data.find('') elements = [x.split('{') for x in data.split('}') if ('{' in x.strip())] try: elements = dict([(a.strip(), dumb_property_dict(b)) for (a, b) in elements]) except ValueError: elements = {} return elements