code
stringlengths
101
5.91M
(signature, parallel=True, cache=True, nogil=False) def weighted_average_product_PxP_C(config1, config2, weights, q1, q2): B = config1.shape[0] M = config1.shape[1] N = config2.shape[1] out = np.zeros((M, N, q1, q2), dtype=curr_float) for m in prange(M): for n in prange(N): for b in prange(B): out[(m, n, config1[(b, m)], config2[(b, n)])] += weights[b] out /= weights.sum() return out
def _graphsLoad(gs: List[Graph], add: bool) -> List[Graph]: us = _unwrap(gs) res = [_graphLoad(a, name=None, add=add) for a in us] return res
def unflatten(arr, shape): size = np.prod(shape) head = ((np.uint64(arr[:size]) << 32) | np.uint64(arr[size:(2 * size)])) return (np.reshape(head, shape), (arr[(2 * size):], ()))
def generate_pretrained_model(): ((x_train, y_train), (x_test, y_test)) = cifar10.load_data() input_shape = x_train.shape[1:] x_train = (x_train.astype('float32') / 255) x_test = (x_test.astype('float32') / 255) x_train_mean = np.mean(x_train, axis=0) x_train -= x_train_mean x_test -= x_train_mean print('x_train shape:', x_train.shape) print(x_train.shape[0], 'train samples') print(x_test.shape[0], 'test samples') print('y_train shape:', y_train.shape) y_train = tensorflow.keras.utils.to_categorical(y_train, num_classes) y_test = tensorflow.keras.utils.to_categorical(y_test, num_classes) model = resnet_v2(input_shape=input_shape, depth=depth) model.compile(loss='categorical_crossentropy', optimizer=tensorflow.keras.optimizers.Adam(learning_rate=0.01), metrics=['accuracy']) model.summary() lr_scheduler = LearningRateScheduler(lr_schedule) lr_reducer = ReduceLROnPlateau(factor=np.sqrt(0.1), cooldown=0, patience=5, min_lr=5e-07) callbacks = [lr_reducer, lr_scheduler] model.fit(x_train, y_train, batch_size=batch_size, epochs=epochs, validation_data=(x_test, y_test), shuffle=True, callbacks=callbacks) scores = model.evaluate(x_test, y_test, verbose=1) print('Test loss:', scores[0]) print('Test accuracy:', scores[1]) model.save('baseline_model')
def _find_quantized_op_num(module, op_qcfgs, prefix='', op_count=0): for (name_tmp, child_tmp) in module.named_children(): op_name = (((prefix + '.') + name_tmp) if (prefix != '') else name_tmp) if ((op_name in op_qcfgs.keys()) and (type(child_tmp) != torch.quantization.QuantStub)): op_count += 1 else: op_count = _find_quantized_op_num(child_tmp, op_qcfgs, op_name, op_count) return op_count
class AdditiveKernels(KernelBase): def __init__(self, k1, k2): super().__init__() self.k1 = k1 self.k2 = k2 def __call__(self, x, y): return (self.k1(x, y) + self.k2(x, y))
def num_prompts(data): pmts = set() for row in data: pmts.add(((row[2] + row[3]) + row[4])) return len(pmts)
class Object3d(BEVBox3D): def __init__(self, center, size, yaw, name, box): super().__init__(center, size, yaw, name, (- 1.0)) self.occlusion = box['occlusion'] self.quaternion = box['quaternion'] self.coords_3d = box['3d_coord'] self.coords_2d = box['2d_coord'] def generate_corners3d(self): return self.coords_3d
def points_to_bev(points, voxel_size, coors_range, with_reflectivity=False, density_norm_num=16, max_voxels=40000): if (not isinstance(voxel_size, np.ndarray)): voxel_size = np.array(voxel_size, dtype=points.dtype) if (not isinstance(coors_range, np.ndarray)): coors_range = np.array(coors_range, dtype=points.dtype) voxelmap_shape = ((coors_range[3:] - coors_range[:3]) / voxel_size) voxelmap_shape = tuple(np.round(voxelmap_shape).astype(np.int32).tolist()) voxelmap_shape = voxelmap_shape[::(- 1)] coor_to_voxelidx = (- np.ones(shape=voxelmap_shape, dtype=np.int32)) bev_map_shape = list(voxelmap_shape) bev_map_shape[0] += 1 height_lowers = np.linspace(coors_range[2], coors_range[5], voxelmap_shape[0], endpoint=False) if with_reflectivity: bev_map_shape[0] += 1 bev_map = np.zeros(shape=bev_map_shape, dtype=points.dtype) _points_to_bevmap_reverse_kernel(points, voxel_size, coors_range, coor_to_voxelidx, bev_map, height_lowers, with_reflectivity, max_voxels) return bev_map
def sigmoid_2(x, mu, sd): xn = ((x - mu) / sd) sig = torch.sigmoid(xn) s = sig js = (((1 / sd) * sig) * (1 - sig)) jjs = ((((1 / (sd ** 2)) * sig) * (1 - sig)) * (1 - (2 * sig))) return (s, js, jjs)
def generate_random_targets(labels: Tensor, num_classes: int) -> Tensor: random = torch.rand(len(labels), num_classes, device=labels.device, dtype=torch.float) random.scatter_(1, labels.unsqueeze((- 1)), 0) return random.argmax(1)
def get_local_size() -> int: if (not dist.is_available()): return 1 if (not dist.is_initialized()): return 1 return dist.get_world_size(group=LOCAL_PROCESS_GROUP)
class Prior(nn.Module): def __init__(self): super().__init__() def sample(self, **kwargs): raise NotImplementedError def log_p(self, input, **kwargs): return self.forward(z) def forward(self, input, **kwargs): raise NotImplementedError def __str__(self): raise NotImplementedError
class Server(ABC): def __init__(self, config, network_config, model, test_loader, seed, optimizer_class: Type, optimizer_params: dict, use_adaptive, use_evaluate=True, lr_scheduler_class=None, lr_scheduler_params=None, control=None, control_scheduler=None, resume=False, init_time_offset=0.0): self.config = config self.use_adaptive = use_adaptive self.use_evaluate = use_evaluate self.max_round = (self.config.MAX_ROUND_ADAPTIVE if use_adaptive else self.config.MAX_ROUND_CONVENTIONAL_FL) if use_adaptive: assert (control is not None) assert (control_scheduler is not None) self.control = control self.control_scheduler = control_scheduler self.model = model self.test_loader = test_loader self.socket = ServerSocket(network_config.SERVER_ADDR, network_config.SERVER_PORT, config.NUM_CLIENTS) save_dir_name = self.get_save_dir_name() self.save_path = os.path.join('results', 'exp_{}'.format(config.EXP_NAME), save_dir_name, 'server') exp_config = ExpConfig(self.config.EXP_NAME, save_dir_name, seed, self.config.CLIENT_BATCH_SIZE, self.config.NUM_LOCAL_UPDATES, optimizer_class, optimizer_params, lr_scheduler_class, lr_scheduler_params, use_adaptive) self.list_loss = None self.list_acc = None self.list_time_stamp = None self.list_model_size = None self.start_time = None self.init_time_offset = init_time_offset self.round = None self.eval_thread = None self.client_is_sparse = False self.terminate = False self.initialize(exp_config, resume) def get_init_extra_params(self) -> List[tuple]: pass def initialize(self, exp_config, resume): list_extra_params = self.get_init_extra_params() self.socket.wait_for_connections() if resume: print('Resuming server...') self.list_loss = load(os.path.join(self.save_path, 'loss.pt')) self.list_acc = load(os.path.join(self.save_path, 'accuracy.pt')) self.list_time_stamp = load(os.path.join(self.save_path, 'time.pt')) self.list_model_size = load(os.path.join(self.save_path, 'model_size.pt')) self.model = load(os.path.join(self.save_path, 'model.pt')) num_loss_acc = len(self.list_loss) assert (len(self.list_acc) == num_loss_acc) num_evals = len(self.list_time_stamp) assert (len(self.list_model_size) == num_evals) if ((num_evals - num_loss_acc) == 1): (loss, acc) = self.model.evaluate(self.test_loader) self.list_loss.append(loss) self.list_acc.append(acc) elif (num_evals != num_loss_acc): raise RuntimeError('Cannot resume') self.round = ((num_evals - 1) * self.config.EVAL_DISP_INTERVAL) assert (self.round >= 0) self.start_time = (timer() - self.list_time_stamp[(- 1)]) self.check_client_to_sparse() resume_param = (True, (self.round + 1), self.client_is_sparse) list_params = [(idx, exp_config, self.model, list_extra_params[idx], resume_param) for idx in range(self.config.NUM_CLIENTS)] resume_msgs_to_client = [ServerToClientInitMessage(init_params) for init_params in list_params] self.socket.init_connections(resume_msgs_to_client) self.round += 1 print('Server resumed') print(self) else: self.list_loss = [] self.list_acc = [] self.list_time_stamp = [] self.list_model_size = [] self.start_time = (timer() + self.init_time_offset) self.round = 0 mkdir_save(self.model, os.path.join(self.save_path, 'init_model.pt')) self.model.eval() list_init_params = [(idx, exp_config, self.model, list_extra_params[idx], (False, 0, False)) for idx in range(self.config.NUM_CLIENTS)] init_msgs_to_client = [ServerToClientInitMessage(init_params) for init_params in list_init_params] self.socket.init_connections(init_msgs_to_client) print('Server initialized') print(self) def get_save_dir_name(self): if (not self.use_adaptive): return 'conventional' else: (mdd_100, chl) = ((100 * self.config.MAX_DEC_DIFF), self.config.ADJ_HALF_LIFE) lrhl = (self.config.LR_HALF_LIFE if hasattr(self.config, 'LR_HALF_LIFE') else None) assert ((mdd_100 - int(mdd_100)) == 0) return 'mdd{}_chl{}_lrhl{}'.format(int(mdd_100), lrhl, chl) def calc_model_params(self, display=False): sum_param_in_use = 0 sum_all_param = 0 for (layer, layer_prefix) in zip(self.model.param_layers, self.model.param_layer_prefixes): num_bias = (0 if (layer.bias is None) else layer.bias.nelement()) layer_param_in_use = (layer.mask.sum().int().item() + num_bias) layer_all_param = (layer.mask.nelement() + num_bias) sum_param_in_use += layer_param_in_use sum_all_param += layer_all_param if display: print('\t{} remaining: {}/{} = {}'.format(layer_prefix, layer_param_in_use, layer_all_param, (layer_param_in_use / layer_all_param))) if display: print('\tTotal: {}/{} = {}'.format(sum_param_in_use, sum_all_param, (sum_param_in_use / sum_all_param))) return sum_param_in_use def adjust_model(self, display=True): print('Running control algorithm') alg_start = timer() max_dec_diff = self.control_scheduler.max_dec_diff(self.round) self.control.adjust(max_dec_diff, None) print('Algorithm completed in {}s'.format((timer() - alg_start))) if display: print('New params:') self.calc_model_params(display=True) self.check_client_to_sparse() _grad() def merge_accumulate_client_update(self, list_num_proc, list_state_dict, lr): total_num_proc = sum(list_num_proc) dict_keys = list_state_dict[0].keys() for state_dict in list_state_dict[1:]: assert (state_dict.keys() == dict_keys) if (self.use_adaptive and self.is_adj_round()): prefix = 'extra.' for state_dict in list_state_dict: del_list = [] for (key, param) in state_dict.items(): if (key[:len(prefix)] == prefix): sgrad_key = key[len(prefix):] mask_0 = (self.model.get_mask_by_name(sgrad_key) == 0.0) dense_sgrad = torch.zeros_like(mask_0, dtype=torch.float) dense_sgrad.masked_scatter_(mask_0, param) self.control.accumulate(sgrad_key, dense_sgrad) del_list.append(key) for del_key in del_list: del state_dict[del_key] server_state_dict = self.model.state_dict() for key in dict_keys: server_param = server_state_dict[key] avg_inc_val = None for (num_proc, state_dict) in zip(list_num_proc, list_state_dict): if (state_dict[key].size() != server_state_dict[key].size()): mask = self.model.get_mask_by_name(key) inc_val = (server_param.masked_scatter(mask, state_dict[key]) - server_param) else: inc_val = (state_dict[key] - server_param) if (avg_inc_val is None): avg_inc_val = ((num_proc / total_num_proc) * inc_val) else: avg_inc_val += ((num_proc / total_num_proc) * inc_val) if (self.use_adaptive and (key in dict(self.model.named_parameters()).keys())): self.control.accumulate(key, (((inc_val / lr) ** 2) * num_proc)) server_param.add_(avg_inc_val) def check_termination(self) -> bool: return self.terminate def evaluate(self): if (self.eval_thread is not None): self.eval_thread.join() t = Thread(target=eval_model_async, args=(deepcopy(self.model).evaluate, self.test_loader, self.list_loss, self.list_acc)) t.start() self.eval_thread = t elapsed_time = (timer() - self.start_time) self.list_time_stamp.append(elapsed_time) model_size = self.calc_model_params(display=False) self.list_model_size.append(model_size) len_loss = len(self.list_loss) len_acc = len(self.list_acc) assert (len_loss == len_acc) print('Evaluation at round #{}. Loss/acc (at round {}) = {}/{}. Elapsed time = {}'.format(self.round, (((len_acc - 1) * self.config.EVAL_DISP_INTERVAL) if (len_acc > 0) else 'NaN'), (self.list_loss[(- 1)] if (len_acc > 0) else 'NaN'), (self.list_acc[(- 1)] if (len_acc > 0) else 'NaN'), elapsed_time)) self.save_exp() def is_adj_round(self, rd=None) -> bool: if (rd is None): rd = self.round return (self.use_adaptive and (rd > 0) and ((rd % self.config.ADJ_INTERVAL) == 0)) def is_one_before_adj_round(self) -> bool: return self.is_adj_round((self.round + 1)) def check_client_to_sparse(self): if ((not self.client_is_sparse) and (self.model.density() <= self.config.TO_SPARSE_THR)): self.client_is_sparse = True def clean_dict_to_client(self) -> dict: clean_state_dict = copy_dict(self.model.state_dict()) if self.client_is_sparse: for (layer, prefix) in zip(self.model.param_layers, self.model.param_layer_prefixes): key = (prefix + '.bias') if (isinstance(layer, DenseLinear) and (key in clean_state_dict.keys())): clean_state_dict[key] = clean_state_dict[key].view(((- 1), 1)) return clean_state_dict _grad() def process_state_dict_to_client(self) -> dict: clean_state_dict = self.clean_dict_to_client() if (not self.client_is_sparse): return clean_state_dict if self.is_adj_round(): for (layer, prefix) in zip(self.model.prunable_layers, self.model.prunable_layer_prefixes): key_w = (prefix + '.weight') if (key_w in clean_state_dict.keys()): weight = clean_state_dict[key_w] w_mask = self.model.get_mask_by_name(key_w) sparse_weight = (weight * w_mask).view(weight.size(0), (- 1)).to_sparse() clean_state_dict[key_w] = sparse_weight else: for prefix in self.model.prunable_layer_prefixes: key_w = (prefix + '.weight') if (key_w in clean_state_dict.keys()): clean_state_dict[key_w] = clean_state_dict[key_w].masked_select(self.model.get_mask_by_name(key_w)) return clean_state_dict def save_exp(self): mkdir_save(self.list_loss, os.path.join(self.save_path, 'loss.pt')) mkdir_save(self.list_acc, os.path.join(self.save_path, 'accuracy.pt')) mkdir_save(self.list_time_stamp, os.path.join(self.save_path, 'time.pt')) mkdir_save(self.list_model_size, os.path.join(self.save_path, 'model_size.pt')) mkdir_save(self.model, os.path.join(self.save_path, 'model.pt')) def main(self): assert (not self.terminate) msgs = self.socket.recv_update_msg_from_all() list_lr = [msg.lr for msg in msgs] list_num_proc = [msg.num_processed for msg in msgs] list_state_dict = [msg.state_dict for msg in msgs] lr = list_lr[0] for client_lr in list_lr[1:]: assert (client_lr == lr) self.merge_accumulate_client_update(list_num_proc, list_state_dict, lr) if (self.use_evaluate and ((self.round % self.config.EVAL_DISP_INTERVAL) == 0)): self.evaluate() if self.is_adj_round(): self.adjust_model() terminate = self.check_termination() if (self.round >= (self.max_round - 1)): terminate = True state_dict_to_client = self.process_state_dict_to_client() client_adj = self.is_one_before_adj_round() to_sparse = self.client_is_sparse msg_to_clients = ServerToClientUpdateMessage((state_dict_to_client, client_adj, to_sparse, terminate)) self.socket.send_msg_to_all(msg_to_clients) self.round += 1 if terminate: self.socket.recv_ack_msg_from_all() self.socket.close() self.eval_thread.join() self.save_exp() self.terminate = True print('Task completed') return terminate def __repr__(self): return 'Experiment = {}'.format(self.config.EXP_NAME)
def equishwidth(elem, spec, specerr, refspec=None): if (refspec is None): refspec = numpy.zeros_like(spec) win = read(elem, apStarWavegrid=True) (startindxs, endindxs) = waveregions(elem, asIndex=True, pad=0) lams = apStarWavegrid() startlams = lams[startindxs] endlams = lams[endindxs] outval = 0.0 norm = 0.0 for (startindx, endindx, startlam, endlam) in zip(startindxs, endindxs, startlams, endlams): norm += numpy.sum((win[startindx:endindx] / (specerr[startindx:endindx] ** 2.0))) if (not numpy.all((refspec == 0.0))): outval += (((endlam - startlam) / (endindx - startindx)) * numpy.sum(((win[startindx:endindx] / (specerr[startindx:endindx] ** 2.0)) * (1.0 - (spec[startindx:endindx] / refspec[startindx:endindx]))))) else: outval += (((endlam - startlam) / (endindx - startindx)) * numpy.sum(((win[startindx:endindx] / (specerr[startindx:endindx] ** 2.0)) * (1.0 - spec[startindx:endindx])))) outval /= norm return outval
class PyrBlock(nn.Module): def __init__(self, in_channels, out_channels, stride): super(PyrBlock, self).__init__() self.conv1 = pre_conv3x3_block(in_channels=in_channels, out_channels=out_channels, stride=stride, activate=False) self.conv2 = pre_conv3x3_block(in_channels=out_channels, out_channels=out_channels) def forward(self, x): x = self.conv1(x) x = self.conv2(x) return x
class BasicBlock(nn.Module): def __init__(self, in_chs, out_chs, kernel_size=3, stride=1, dilation=(1, 1), group_size=None, bottle_ratio=1.0, downsample='avg', linear_out=False, layers: LayerFn=None, drop_block=None, drop_path_rate=0.0): super(BasicBlock, self).__init__() layers = (layers or LayerFn()) mid_chs = make_divisible((out_chs * bottle_ratio)) groups = num_groups(group_size, mid_chs) if ((in_chs != out_chs) or (stride != 1) or (dilation[0] != dilation[1])): self.shortcut = create_downsample(downsample, in_chs=in_chs, out_chs=out_chs, stride=stride, dilation=dilation[0], apply_act=False, layers=layers) else: self.shortcut = nn.Identity() self.conv1_kxk = layers.conv_norm_act(in_chs, mid_chs, kernel_size, stride=stride, dilation=dilation[0]) self.conv2_kxk = layers.conv_norm_act(mid_chs, out_chs, kernel_size, dilation=dilation[1], groups=groups, drop_block=drop_block, apply_act=False) self.attn = (nn.Identity() if (layers.attn is None) else layers.attn(out_chs)) self.drop_path = (DropPath(drop_path_rate) if (drop_path_rate > 0.0) else nn.Identity()) self.act = (nn.Identity() if linear_out else layers.act(inplace=True)) def init_weights(self, zero_init_last_bn=False): if zero_init_last_bn: nn.init.zeros_(self.conv2_kxk.bn.weight) def forward(self, x): shortcut = self.shortcut(x) x = self.conv1_kxk(x) x = self.conv2_kxk(x) x = self.attn(x) x = self.drop_path(x) x = self.act((x + shortcut)) return x
(scope='module') def simple_dtype(): ld = np.dtype('longdouble') return np.dtype({'names': ['bool_', 'uint_', 'float_', 'ldbl_'], 'formats': ['?', 'u4', 'f4', f'f{ld.itemsize}'], 'offsets': [0, 4, 8, (16 if (ld.alignment > 4) else 12)]})
class EsmModel(metaclass=DummyObject): _backends = ['torch'] def __init__(self, *args, **kwargs): requires_backends(self, ['torch'])
def matthews_corrcoef(predictions, targets) -> dict: return {'matthews_correlation': (100 * sklearn.metrics.matthews_corrcoef(targets, predictions))}
class CosineSchedule(FairseqLRScheduler): def __init__(self, args, optimizer): super().__init__(args, optimizer) if (len(args.lr) > 1): raise ValueError('Cannot use a fixed learning rate schedule with cosine. Consider --lr-scheduler=fixed instead.') warmup_end_lr = args.max_lr if (args.warmup_init_lr < 0): args.warmup_init_lr = args.lr[0] self.min_lr = args.lr[0] self.max_lr = args.max_lr assert (self.max_lr > self.min_lr), 'max_lr must be more than lr' self.t_mult = args.t_mult self.period = args.lr_period_updates if (self.period <= 0): assert (args.max_update >= 0), 'Either --max_update or --lr-period-updates must be set' self.period = (args.max_update - args.warmup_updates) if (args.warmup_updates > 0): self.lr_step = ((warmup_end_lr - args.warmup_init_lr) / args.warmup_updates) else: self.lr_step = 1 self.warmup_updates = args.warmup_updates self.lr_shrink = args.lr_shrink self.lr = args.warmup_init_lr self.optimizer.set_lr(self.lr) def add_args(parser): parser.add_argument('--warmup-updates', default=0, type=int, metavar='N', help='warmup the learning rate linearly for the first N updates') parser.add_argument('--warmup-init-lr', default=(- 1), type=float, metavar='LR', help='initial learning rate during warmup phase; default is args.lr') parser.add_argument('--max-lr', required=True, type=float, metavar='LR', help='max learning rate, must be more than args.lr') parser.add_argument('--t-mult', default=1, type=float, metavar='LR', help='factor to grow the length of each period') parser.add_argument('--lr-period-updates', default=(- 1), type=float, metavar='LR', help='initial number of updates per period') def step(self, epoch, val_loss=None): super().step(epoch, val_loss) return self.optimizer.get_lr() def step_update(self, num_updates): if (num_updates < self.args.warmup_updates): self.lr = (self.args.warmup_init_lr + (num_updates * self.lr_step)) else: curr_updates = (num_updates - self.args.warmup_updates) if (self.t_mult != 1): i = math.floor(math.log((1 - ((curr_updates / self.period) * (1 - self.t_mult))), self.t_mult)) t_i = ((self.t_mult ** i) * self.period) t_curr = (curr_updates - (((1 - (self.t_mult ** i)) / (1 - self.t_mult)) * self.period)) else: i = math.floor((curr_updates / self.period)) t_i = self.period t_curr = (curr_updates - (self.period * i)) lr_shrink = (self.lr_shrink ** i) min_lr = (self.min_lr * lr_shrink) max_lr = (self.max_lr * lr_shrink) self.lr = (min_lr + ((0.5 * (max_lr - min_lr)) * (1 + math.cos(((math.pi * t_curr) / t_i))))) self.optimizer.set_lr(self.lr) return self.lr
class FocalLoss(nn.Module): def __init__(self, gamma=0, size_average=True): super(FocalLoss, self).__init__() self.gamma = gamma self.size_average = size_average def forward(self, input, target): target = target.view((- 1), 1) logpt = F.log_softmax(input, dim=1) logpt = logpt.gather(1, target) logpt = logpt.view((- 1)) pt = logpt.exp() loss = (((- 1) * ((1 - pt) ** self.gamma)) * logpt) if self.size_average: return loss.mean() else: return loss.sum()
class CNNEncoder(BaseEncoder): def __init__(self, latent_dimensions: int, feature_size: Iterable, variational: bool=False, channels: tuple=None, kernel_sizes: tuple=None, strides: tuple=None, paddings: tuple=None, activation=nn.LeakyReLU(), dropout=0): super(CNNEncoder, self).__init__(latent_dimensions, variational=variational) default_len = (2 if (channels is None) else len(channels)) if (channels is None): channels = (1, 1) kernel_sizes = (kernel_sizes or ((5,) * default_len)) strides = (strides or ((1,) * default_len)) paddings = (paddings or ((2,) * default_len)) self.conv_layers = self._build_conv_layers(channels, kernel_sizes, strides, paddings, activation) final_channels = channels[(- 1)] final_size = feature_size[0] linear_input_size = ((final_channels * final_size) * final_size) if self.variational: self.create_variational_layers(linear_input_size) else: self.fc = nn.Sequential(nn.Dropout(p=dropout), nn.Linear(linear_input_size, latent_dimensions)) def _build_conv_layers(self, channels, kernel_sizes, strides, paddings, activation): layers = [] current_channels = 1 for idx in range(len(channels)): layers.append(nn.Sequential(nn.Conv2d(in_channels=current_channels, out_channels=channels[idx], kernel_size=kernel_sizes[idx], stride=strides[idx], padding=paddings[idx]), activation)) current_channels = channels[idx] return nn.Sequential(*layers) def forward(self, x): x = self.conv_layers(x) x = x.view(x.size(0), (- 1)) if self.variational: return self.forward_variational(x) return self.fc(x)
class ShuffleNetV2(nn.Module): def __init__(self, stages_repeats, stages_out_channels, num_classes=1000): super(ShuffleNetV2, self).__init__() if (len(stages_repeats) != 3): raise ValueError('expected stages_repeats as list of 3 positive ints') if (len(stages_out_channels) != 5): raise ValueError('expected stages_out_channels as list of 5 positive ints') self._stage_out_channels = stages_out_channels input_channels = 3 output_channels = self._stage_out_channels[0] self.conv1 = nn.Sequential(nn.Conv2d(input_channels, output_channels, 3, 2, 1, bias=False), nn.BatchNorm2d(output_channels), nn.ReLU(inplace=True)) input_channels = output_channels self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1) stage_names = ['stage{}'.format(i) for i in [2, 3, 4]] for (name, repeats, output_channels) in zip(stage_names, stages_repeats, self._stage_out_channels[1:]): seq = [InvertedResidual(input_channels, output_channels, 2)] for i in range((repeats - 1)): seq.append(InvertedResidual(output_channels, output_channels, 1)) setattr(self, name, nn.Sequential(*seq)) input_channels = output_channels output_channels = self._stage_out_channels[(- 1)] self.conv5 = nn.Sequential(nn.Conv2d(input_channels, output_channels, 1, 1, 0, bias=False), nn.BatchNorm2d(output_channels), nn.ReLU(inplace=True)) self.fc = nn.Linear(output_channels, num_classes) def forward(self, x): x = self.conv1(x) x = self.maxpool(x) x = self.stage2(x) x = self.stage3(x) x = self.stage4(x) x = self.conv5(x) x = x.mean(2) x = x.mean(2) x = self.fc(x) return x
class Sentiment(_Sentiment): def load(self, path=None): _Sentiment.load(self, path) if (not path): for (w, pos) in list(dict.items(self)): if ('JJ' in pos): if w.endswith('y'): w = (w[:(- 1)] + 'i') if w.endswith('le'): w = w[:(- 2)] (p, s, i) = pos['JJ'] self.annotate((w + 'ly'), 'RB', p, s, i)
((_HAS_VIDEO_OPT is False), "Didn't compile with ffmpeg") class TestVideo(unittest.TestCase): _SKIP ((av is None), 'PyAV unavailable') def test_read_video_tensor(self): torchvision.set_video_backend('pyav') for (test_video, config) in test_videos.items(): full_path = os.path.join(VIDEO_DIR, test_video) (tv_result, _, _) = torchvision.io.read_video(full_path, pts_unit='sec') tv_result = tv_result.permute(0, 3, 1, 2) reader = VideoReader(full_path, 'video') frames = [] for frame in reader: frames.append(frame['data']) new_api = torch.stack(frames, 0) self.assertEqual(tv_result.size(), new_api.size()) ((av is None), 'PyAV unavailable') def test_metadata(self): torchvision.set_video_backend('pyav') for (test_video, config) in test_videos.items(): full_path = os.path.join(VIDEO_DIR, test_video) reader = VideoReader(full_path, 'video') reader_md = reader.get_metadata() self.assertAlmostEqual(config.video_fps, reader_md['video']['fps'][0], delta=0.0001) self.assertAlmostEqual(config.duration, reader_md['video']['duration'][0], delta=0.5) _SKIP ((av is None), 'PyAV unavailable') def test_video_reading_fn(self): for (test_video, config) in test_videos.items(): full_path = os.path.join(VIDEO_DIR, test_video) ref_result = _decode_frames_by_av_module(full_path) reader = VideoReader(full_path, 'video') newapi_result = _template_read_video(reader) if ((newapi_result.vframes.numel() > 0) and (ref_result.vframes.numel() > 0)): mean_delta = torch.mean(torch.abs((newapi_result.vframes.float() - ref_result.vframes.float()))) self.assertAlmostEqual(mean_delta, 0, delta=8.0) self.assertEqual(newapi_result.vframes.size(), ref_result.vframes.size()) if (config.check_aframes and (newapi_result.aframes.numel() > 0) and (ref_result.aframes.numel() > 0)): is_same = torch.all(torch.eq(newapi_result.aframes, ref_result.aframes)).item() self.assertEqual(is_same, True)
def TestFlag(flag, test_val, default_val): env_var = ('GTEST_' + flag.upper()) SetEnvVar(env_var, test_val) AssertEq(test_val, GetFlag(flag)) SetEnvVar(env_var, None) AssertEq(default_val, GetFlag(flag))
def _run_tensor_parallel_optimization(num_nodes=1, num_devices_per_node=2): torch.manual_seed(42) model_context = create_model_context() parallel_optimization = TensorParallelOptimization(shard_planner='base', num_nodes=num_nodes, num_devices_per_node=num_devices_per_node, tracer_backend='meta_fx', prop_mode='meta_tracer') (status, best_config, model_context) = parallel_optimization.tune(model_context, {'tp_ranks': [0, 1]}, []) assert (('replacement_map' in best_config) and status)
class TestReproducibility(unittest.TestCase): def _test_reproducibility(self, name, extra_flags=None, delta=0.0001, resume_checkpoint='checkpoint1.pt', max_epoch=3): if (extra_flags is None): extra_flags = [] with tempfile.TemporaryDirectory(name) as data_dir: with self.assertLogs() as logs: test_binaries.create_dummy_data(data_dir) test_binaries.preprocess_translation_data(data_dir) with self.assertLogs() as logs: test_binaries.train_translation_model(data_dir, 'fconv_iwslt_de_en', (['--dropout', '0.0', '--log-format', 'json', '--log-interval', '1', '--max-epoch', str(max_epoch)] + extra_flags)) (train_log, valid_log) = map((lambda rec: json.loads(rec.msg)), logs.records[(- 4):(- 2)]) os.rename(os.path.join(data_dir, resume_checkpoint), os.path.join(data_dir, 'checkpoint_last.pt')) with self.assertLogs() as logs: test_binaries.train_translation_model(data_dir, 'fconv_iwslt_de_en', (['--dropout', '0.0', '--log-format', 'json', '--log-interval', '1', '--max-epoch', str(max_epoch)] + extra_flags)) (train_res_log, valid_res_log) = map((lambda rec: json.loads(rec.msg)), logs.records[(- 4):(- 2)]) for k in ['train_loss', 'train_ppl', 'train_num_updates', 'train_gnorm']: self.assertAlmostEqual(float(train_log[k]), float(train_res_log[k]), delta=delta) for k in ['valid_loss', 'valid_ppl', 'valid_num_updates', 'valid_best_loss']: self.assertAlmostEqual(float(valid_log[k]), float(valid_res_log[k]), delta=delta) def test_reproducibility(self): self._test_reproducibility('test_reproducibility') ((not torch.cuda.is_available()), 'test requires a GPU') def test_reproducibility_fp16(self): self._test_reproducibility('test_reproducibility_fp16', ['--fp16', '--fp16-init-scale', '4096'], delta=0.011) ((not torch.cuda.is_available()), 'test requires a GPU') def test_reproducibility_memory_efficient_fp16(self): self._test_reproducibility('test_reproducibility_memory_efficient_fp16', ['--memory-efficient-fp16', '--fp16-init-scale', '4096']) def test_mid_epoch_reproducibility(self): self._test_reproducibility('test_mid_epoch_reproducibility', ['--save-interval-updates', '3'], resume_checkpoint='checkpoint_1_3.pt', max_epoch=1)
def post_processing_function(examples, features, predictions, stage='eval'): predictions = postprocess_qa_predictions(examples=examples, features=features, predictions=predictions, version_2_with_negative=args.version_2_with_negative, n_best_size=args.n_best_size, max_answer_length=args.max_answer_length, null_score_diff_threshold=args.null_score_diff_threshold, output_dir=args.output_dir, prefix=stage) if args.version_2_with_negative: formatted_predictions = [{'id': k, 'prediction_text': v, 'no_answer_probability': 0.0} for (k, v) in predictions.items()] else: formatted_predictions = [{'id': k, 'prediction_text': v} for (k, v) in predictions.items()] references = [{'id': ex['id'], 'answers': ex[answer_column_name]} for ex in examples] return EvalPrediction(predictions=formatted_predictions, label_ids=references)
def replace_tags_board(board_san): board_san = board_san.split(' ')[0] board_san = board_san.replace('2', '11') board_san = board_san.replace('3', '111') board_san = board_san.replace('4', '1111') board_san = board_san.replace('5', '11111') board_san = board_san.replace('6', '111111') board_san = board_san.replace('7', '1111111') board_san = board_san.replace('8', '') return board_san.replace('/', '')
class Joint(xmlr.Object): TYPES = ['unknown', 'revolute', 'continuous', 'prismatic', 'floating', 'planar', 'fixed'] def __init__(self, name=None, parent=None, child=None, joint_type=None, axis=None, origin=None, limit=None, dynamics=None, safety_controller=None, calibration=None, mimic=None, hardwareInterface=None): self.name = name self.parent = parent self.child = child self.type = joint_type self.axis = axis self.origin = origin self.limit = limit self.dynamics = dynamics self.safety_controller = safety_controller self.calibration = calibration self.mimic = mimic self.hardwareInterface = hardwareInterface def check_valid(self): assert (self.type in self.TYPES), 'Invalid joint type: {}'.format(self.type) def joint_type(self): return self.type _type.setter def joint_type(self, value): self.type = value
def evaluate(model): from neural_compressor.model import Model from neural_compressor import Metric model = Model(model) input_tensor = model.input_tensor output_tensor = (model.output_tensor if (len(model.output_tensor) > 1) else model.output_tensor[0]) iteration = (- 1) if (FLAGS.benchmark and (FLAGS.mode == 'performance')): iteration = FLAGS.iters from neural_compressor import METRICS metrics = METRICS('tensorflow') metric = metrics['topk']() def eval_func(dataloader): latency_list = [] for (idx, (inputs, labels)) in enumerate(dataloader): inputs = np.array([inputs]) assert (len(input_tensor) == len(inputs)), 'inputs len must equal with input_tensor' feed_dict = dict(zip(input_tensor, inputs)) start = time.time() predictions = model.sess.run(output_tensor, feed_dict) end = time.time() metric.update(predictions, labels) latency_list.append((end - start)) if ((idx + 1) == iteration): break latency = (np.array(latency_list).mean() / FLAGS.batch_size) return latency from neural_compressor.data import DataLoader dataloader = DataLoader(framework='tensorflow', dataset=Dataset(), batch_size=FLAGS.batch_size) latency = eval_func(dataloader) if (FLAGS.benchmark and (FLAGS.mode == 'performance')): print('Batch size = {}'.format(FLAGS.batch_size)) print('Latency: {:.3f} ms'.format((latency * 1000))) print('Throughput: {:.3f} images/sec'.format((1.0 / latency))) acc = metric.result() return acc
def average_gradients(tower_grads): average_grads = [] for grad_and_vars in zip(*tower_grads): grads = [] for (g, _) in grad_and_vars: expanded_g = tf.expand_dims(g, 0) grads.append(expanded_g) grad = tf.concat(0, grads) grad = tf.reduce_mean(grad, 0) v = grad_and_vars[0][1] grad_and_var = (grad, v) average_grads.append(grad_and_var) return average_grads
_sentencepiece _tokenizers class TestMarian_EN_DE_More(MarianIntegrationTest): def test_forward(self): (src, tgt) = (['I am a small frog'], ['Ich bin ein kleiner Frosch.']) expected_ids = [38, 121, 14, 697, 38848, 0] model_inputs: dict = self.tokenizer.prepare_seq2seq_batch(src, tgt_texts=tgt, return_tensors='pt').to(torch_device) self.assertListEqual(expected_ids, model_inputs.input_ids[0].tolist()) desired_keys = {'input_ids', 'attention_mask', 'labels'} self.assertSetEqual(desired_keys, set(model_inputs.keys())) model_inputs['decoder_input_ids'] = shift_tokens_right(model_inputs.labels, self.tokenizer.pad_token_id) model_inputs['return_dict'] = True model_inputs['use_cache'] = False with torch.no_grad(): outputs = self.model(**model_inputs) max_indices = outputs.logits.argmax((- 1)) self.tokenizer.batch_decode(max_indices) def test_unk_support(self): t = self.tokenizer ids = t.prepare_seq2seq_batch(['||'], return_tensors='pt').to(torch_device).input_ids[0].tolist() expected = [t.unk_token_id, t.unk_token_id, t.eos_token_id] self.assertEqual(expected, ids) def test_pad_not_split(self): input_ids_w_pad = self.tokenizer.prepare_seq2seq_batch(['I am a small frog <pad>'], return_tensors='pt').input_ids[0].tolist() expected_w_pad = [38, 121, 14, 697, 38848, self.tokenizer.pad_token_id, 0] self.assertListEqual(expected_w_pad, input_ids_w_pad) def test_batch_generation_en_de(self): self._assert_generated_batch_equal_expected() def test_auto_config(self): config = AutoConfig.from_pretrained(self.model_name) self.assertIsInstance(config, MarianConfig)
_REGISTRY.register() def build_mnv2_backbone(cfg, input_shape): out_features = cfg.MODEL.RESNETS.OUT_FEATURES out_feature_channels = {'res2': 24, 'res3': 32, 'res4': 96, 'res5': 320} out_feature_strides = {'res2': 4, 'res3': 8, 'res4': 16, 'res5': 32} model = MobileNetV2(cfg) model._out_features = out_features model._out_feature_channels = out_feature_channels model._out_feature_strides = out_feature_strides return model
def test_double_double_syspool(vrblvl=0): initialize_double_double_syspool(3, vrblvl) dim = size_double_double_syspool(vrblvl) print('The size of the systems pool :', dim) pol1 = ['t - 1/3;'] set_double_double_system(1, pol1, vrblvl) copy_to_double_double_syspool(1) pol2 = ['t - 2/3;'] set_double_double_system(1, pol2, vrblvl) copy_to_double_double_syspool(2) pol3 = ['t - 1;'] set_double_double_system(1, pol3, vrblvl) copy_to_double_double_syspool(3) for i in range(1, (dim + 1)): clear_double_double_system(vrblvl) copy_from_double_double_syspool(i) pols = get_double_double_system(vrblvl) print('system at', i, 'in the pool :', pols) clear_double_double_syspool(vrblvl) return int((dim != 3))
def main(args): in_dir = os.path.join(args.in_dir, args.exp, args.custom_dir) out_dir = os.path.join(args.out_dir, args.exp, args.custom_dir) os.makedirs(out_dir, exist_ok=True) exp_util.clear_dir(out_dir) logger = exp_util.get_logger(os.path.join(out_dir, 'log.txt')) logger.info(args) logger.info(datetime.now()) process(args, in_dir, out_dir, logger)
def vgg16_bn_128(pretrained=False, **kwargs): model = VGG(make_layers(cfg['D_128'], batch_norm=True), **kwargs) if pretrained: model.load_state_dict(model_zoo.load_url(model_urls['vgg16_bn'])) return model
def grid_search(model_constructor, train_data, dev_data, config, acc_priors, balance_priors, epochs, label_to_ix): best_p = float('-inf') best_r = float('-inf') best_f1 = float('-inf') best_params = None best_acc_prior = None best_balance_prior = None for acc_prior in acc_priors: for balance_prior in balance_priors: model = model_constructor(acc_prior, balance_prior) (p, r, f1) = train_generative_model(model, train_data, dev_data, epochs, label_to_ix, config) if (f1 > best_f1): best_p = p best_r = r best_f1 = f1 best_params = model.state_dict() best_acc_prior = acc_prior best_balance_prior = balance_prior best_model = model_constructor(best_acc_prior, best_balance_prior) best_model.load_state_dict(best_params) return (best_model, best_p, best_r, best_f1)
def get_wavelength_from_header(hdr): if (('CRVAL1' and ('CRPIX1' in hdr.keys())) and (('CDELT1' in hdr.keys()) or ('CD1_1' in hdr.keys()))): if ('CD1_1' in hdr.keys()): cdelt = hdr['CD1_1'] else: cdelt = hdr['CDELT1'] crval = hdr['CRVAL1'] crpix = hdr['CRPIX1'] wavelength = (((np.arange(hdr['NAXIS1']) - (crpix - 1)) * cdelt) + crval) return wavelength else: raise WavelengthError('Not enough information in header to create wavelength array')
def change_label(tree, new_label, span=None, cur_label=None, in_place=True): if ((span is None) and (cur_label is None)): return change_label_by_node(tree, new_label, in_place) elif ((span is not None) and (cur_label is not None)): return change_label_by_span(tree, new_label, span, cur_label, in_place) else: raise Exception('Invalid combination of arguments for change label request')
def worker_init_fn(worker_id, num_workers, rank, seed): worker_seed = (((num_workers * rank) + worker_id) + seed) np.random.seed(worker_seed) random.seed(worker_seed) torch.manual_seed(worker_seed)
class SolutionInstance(): def __init__(self, objVal, X, Y, W, H, solNo): self.X = X self.Y = Y self.W = W self.H = H self.objVal = objVal self.solNo = solNo
class GraphIR(): vars: Dict[(str, AbsTensor)] = field(default_factory=dict) insts: List[InstIR] = field(default_factory=list) def __str__(self) -> str: ret = '' for inst in self.insts: ret += f'''{inst} ''' return ret def pretty(self) -> str: inst_remap = {inst.identifier: f'{idx}' for (idx, inst) in enumerate(self.insts)} ret = '' for inst in self.insts: pretty_args = [] for arg in inst.iexpr.args: (inst_id, ret_idx) = InstIR.var_inst_idx(arg) pretty_args.append(InstIR.retval_string(inst_remap[inst_id], ret_idx)) pretty_retvals = [InstIR.retval_string(inst_remap[inst.identifier], ret_idx) for ret_idx in range(inst.n_output())] ret += f"{', '.join(pretty_retvals)} = {inst.iexpr.op}({', '.join(pretty_args)})" ret += f''' # inst id: {inst_remap[inst.identifier]} ''' return ret def n_inst(self) -> int: return len(self.insts) def n_compute_inst(self) -> int: return sum((1 for inst in self.insts if (not isinstance(inst.iexpr.op, (Input, Constant, Placeholder))))) def n_var(self) -> int: return len(self.vars) def leaf_inst(self) -> List[InstIR]: return [inst for inst in self.insts if inst.no_users()] def leaf_var(self) -> List[str]: lvs = [] for inst in self.insts: for lv in inst.leaf_var(): if (lv not in lvs): lvs.append(lv) return lvs def input_var(self) -> List[str]: return [inst.retval() for inst in self.insts if isinstance(inst.iexpr.op, Input)] def add_inst(self, iexpr: InstExpr) -> InstIR: new_inst = InstIR(iexpr, irctx=self) otensors = iexpr.op.output_like if any([(t is None) for t in otensors]): otensors = iexpr.op.checked_type_transfer([self.vars[arg] for arg in iexpr.args]) for (ridx, abstensor) in enumerate(otensors): vname = new_inst.retval(ridx) assert (vname not in self.vars), ('Variable name is not unique: ' + vname) self.vars[vname] = abstensor min_user_idx = 0 for arg in set(iexpr.args): assert (arg in self.vars), ('Variable not defined: ' + arg) (inst_id, ret_idx) = InstIR.var_inst_idx(arg) for (idx, may_prod) in enumerate(self.insts): if (inst_id == may_prod.identifier): may_prod.users[ret_idx].add(new_inst) min_user_idx = max(min_user_idx, (idx + 1)) break self.insts.insert(min_user_idx, new_inst) return new_inst def find_inst_by_id(self, obj_id: int) -> Optional[InstIR]: for inst in self.insts: if (inst.identifier == obj_id): return inst return None def replace_alluse(self, oldvar: str, newvar: str, type_check=True) -> None: assert (oldvar in self.vars), ('oldvar not defined: ' + oldvar) assert (newvar in self.vars), ('newvar not defined: ' + newvar) if (type_check and (self.vars[oldvar] is not None) and (self.vars[newvar] is not None)): assert self.vars[oldvar].weak_compare(self.vars[newvar]), f'Type mismatch: {self.vars[oldvar]} != {self.vars[newvar]}' (old_inst_id, old_ret_idx) = InstIR.var_inst_idx(oldvar) old_inst = self.find_inst_by_id(old_inst_id) assert (old_inst is not None), ('oldvar not defined: ' + oldvar) for ouser in old_inst.users[old_ret_idx]: ouser.iexpr.args = tuple(((newvar if (a == oldvar) else a) for a in ouser.iexpr.args)) (new_inst_id, new_ret_idx) = InstIR.var_inst_idx(newvar) new_inst = self.find_inst_by_id(new_inst_id) new_inst.users[new_ret_idx] = old_inst.users[old_ret_idx] old_inst.users[old_ret_idx] = set() def replace_arg(self, inst: InstIR, arg_idx: int, newvar: str, type_check=True): assert (newvar in self.vars), ('newvar not defined: ' + newvar) assert (0 <= arg_idx < len(inst.iexpr.args)), f'Invalid argument index {arg_idx} for {inst}' oldvar = inst.iexpr.args[arg_idx] if (type_check and (self.vars[oldvar] is not None) and (self.vars[newvar] is not None)): assert (self.vars[oldvar] == self.vars[newvar]), f'Type mismatch: {self.vars[oldvar]} != {self.vars[newvar]}' inst.iexpr.args[arg_idx] = newvar if (oldvar not in inst.iexpr.args): (old_inst_id, old_ret_idx) = InstIR.var_inst_idx(oldvar) old_inst = self.find_inst_by_id(old_inst_id) old_inst.users[old_ret_idx].remove(inst) (new_inst_id, new_ret_idx) = InstIR.var_inst_idx(newvar) new_inst = self.find_inst_by_id(new_inst_id) new_inst.users[new_ret_idx].add(inst) def remove_unused(self, inst: InstIR) -> None: assert (inst in self.insts), f'Instruction not in graph: {inst}' assert inst.no_users(), f'{inst} has users {inst.users}.' for other in self.insts: if (other != inst): for users in other.users: if (inst in users): users.remove(inst) for val in inst.retvals(): del self.vars[val] self.insts.remove(inst) def assert_wellform(self): defined = set() for inst in self.insts: for arg in inst.iexpr.args: assert (arg in self.vars), f'Variable not defined: {arg}' assert (arg in defined), f'Variable not defined yet: {arg}' (usee_id, ret_idx) = InstIR.var_inst_idx(arg) usee = self.find_inst_by_id(usee_id) assert (inst in usee.users[ret_idx]), f'Use-Def chain broken: {usee} should be used by {inst}' for rv in inst.retvals(): assert (rv in self.vars), f'Return var not in self.vars: {rv}' assert (rv not in defined), f'Variable defined twice: {rv}' for (ret_idx, users) in enumerate(inst.users): val = inst.retval(ret_idx) for user in users: assert (val in user.iexpr.args), f'Use-Def chain broken: {inst} should be used by {user}' defined.update(inst.retvals()) def _topological_sort(self): defined = set() def swap(i, j): (self.insts[i], self.insts[j]) = (self.insts[j], self.insts[i]) ptr = 0 while (ptr < len(self.insts)): frontier = [] for idx in range(ptr, len(self.insts)): inst = self.insts[idx] if all(((arg in defined) for arg in inst.iexpr.args)): frontier.append(idx) if (len(frontier) == 0): CORE_LOG.error(f'''Bad IR: {self.pretty()}''') raise RuntimeError('Cyclic dependency detected.') for idx in frontier: swap(ptr, idx) defined.update(self.insts[ptr].retvals()) ptr += 1 def _udchain_repair(self): for inst in self.insts: for arg in inst.iexpr.args: (usee_id, ret_idx) = InstIR.var_inst_idx(arg) usee = self.find_inst_by_id(usee_id) usee.users[ret_idx].add(inst) for (ret_idx, users) in enumerate(inst.users): val = inst.retval(ret_idx) for user in list(users): if (val not in user.iexpr.args): users.remove(user) def wellform_repair(self): self._udchain_repair() self._topological_sort() def concretize(self, model: ModelRef) -> None: for inst in self.insts: op = concretize_op(inst.iexpr.op, model) itensors = [self.vars[vname] for vname in inst.iexpr.args] otensors = op.checked_type_transfer(itensors) op.bind_input_like(itensors) op.bind_output_like(otensors) inst.iexpr.op = op for (vname, tensor) in zip(inst.retvals(), otensors): self.vars[vname] = tensor def debug(self): print('=== IR VARIABLES ===') print(self.vars) print('=== IR INSTS ===') has_inferred = False for inst in self.insts: if (inst.iexpr.op.__class__.__name__ == 'AutoInfOpBase'): ainst = inst.iexpr.op.inst has_inferred |= (not ainst.infer_failed()) print(inst.retvals(), '<-', ainst.invoke_str(inst.iexpr.op.attrs), '<-', inst.iexpr.args) else: print(inst.retvals(), '<-', inst.iexpr.op, '<-', inst.iexpr.args) if has_inferred: print('=== IR INSTS by MASKING ATTR w/ `` ===') for inst in self.insts: if (inst.iexpr.op.__class__.__name__ == 'AutoInfOpBase'): ainst = inst.iexpr.op.inst if (not ainst.infer_failed()): print(inst.retvals(), '<-', ainst.invoke_str({k: f'{k}' for k in ainst.A}), '<-', inst.iexpr.args) print('SHAPE PROP:', [t for t in ainst.type_transfer_dbg_info.split('\n') if t]) print('REQUIRES:', ([t for t in ainst.requires_dbg_info.split('\n') if t] if (len(ainst.nnsmith_rules()) == 0) else ainst.nnsmith_rules()[0].__name__)) def to_dot(self) -> str: text = 'digraph D {\n' text += ' node [shape=Mrecord];\n' def render_node(inst: InstIR): label = '{' extra = '' if (not isinstance(inst.iexpr.op, (Input, Constant, Placeholder))): label += '{' label += '|'.join([f'<i{idx}> {arg}' for (idx, arg) in enumerate(inst.iexpr.args)]) label += '}|' elif isinstance(inst.iexpr.op, Input): extra += 'fillcolor=cadetblue1,style=filled,' elif isinstance(inst.iexpr.op, Constant): extra += 'fillcolor=lightpink,style=filled,' elif isinstance(inst.iexpr.op, Placeholder): extra += 'fillcolor=lightgray,style=filled,' label += f'{inst.iexpr.op}|'.replace('{', '\\{').replace('}', '\\}') label += '{' label += '|'.join([f'<o{idx}> {rv}' for (idx, rv) in enumerate(inst.retvals())]) label += '}}' return f''' {inst.identifier} [label="{label}",{extra}]; ''' for inst in self.insts: text += render_node(inst) for inst in self.insts: for (idx, arg) in enumerate(inst.iexpr.args): (usee_id, ret_idx) = InstIR.var_inst_idx(arg) text += f''' {usee_id}:o{ret_idx} -> {inst.identifier}:i{idx} [label="{self.vars[arg].pretty()}"]; ''' text += '}\n' return text
def get_model(outputs, width, height, scale, n_patches, patch_size, reg): x_in = Input(shape=(height, width, 3)) x_high = ImageLinearTransform()(x_in) x_high = ImagePan(horizontally=True, vertically=True)(x_high) x_low = ResizeImages((int((height * scale)), int((width * scale))))(x_high) (features, att, patches) = attention_sampling(attention, resnet, patch_size, n_patches, replace=False, attention_regularizer=multinomial_entropy(reg), receptive_field=9)([x_low, x_high]) y = Dense(outputs, activation='softmax')(features) return (Model(inputs=x_in, outputs=[y]), Model(inputs=x_in, outputs=[att, patches, x_low]))
def transferFileToHdfsPath(sourcepath, targetpath): hdfspath = targetpath targetdir = os.path.dirname(targetpath) os.system('/opt/hadoop/latest/bin/hdfs dfs -mkdir -p {}'.format(targetdir)) result = os.system('/opt/hadoop/latest/bin/hdfs dfs -copyFromLocal -f {} {}'.format(sourcepath, hdfspath)) if (result != 0): raise OSError('Cannot copyFromLocal {} {} returned {}'.format(sourcepath, hdfspath, result))
def get_model_fwk_name(model): onnx = LazyImport('onnx') tf = LazyImport('tensorflow') torch = LazyImport('torch') def _is_onnxruntime(model): if isinstance(model, str): try: graph = onnx.load(model) assert (len(graph.graph.node) != 0) except: return 'NA' else: return 'onnxruntime' elif ('onnx' in str(type(model))): return 'onnxruntime' return 'NA' def _is_tensorflow(model): try: if isinstance(model, str): graph_def = tf.compat.v1.GraphDef() with open(model, 'rb') as f: graph_def.ParseFromString(f.read()) else: graph = model.graph_def except: pass else: return 'tensorflow' return 'NA' def _is_torch(model): if isinstance(model, str): try: torch.jit.load(model) except: return 'NA' else: return 'torch' elif ('torch' in str(type(model))): return 'torch' return 'NA' def _is_neural_engine(model): if (model and isinstance(model, str) and os.path.isdir(model)): if (os.path.exists(os.path.join(model, 'conf.yaml')) and os.path.exists(os.path.join(model, 'model.bin'))): return 'neural engine' else: logger.error('Please Input yaml and bin for neural engine.') return 'NA' else: return 'NA' if isinstance(model, str): absmodel = os.path.abspath(os.path.expanduser(model)) assert (os.path.exists(absmodel) or os.path.exists((absmodel + '.pb'))), 'invalid input path, the file does not exist!' checker = [_is_onnxruntime, _is_neural_engine, _is_tensorflow, _is_torch] for handler in checker: fwk_name = handler(model) if (fwk_name != 'NA'): break assert (fwk_name != 'NA'), 'Framework is not detected correctly from model format.' return fwk_name
def dynamic_import_st(module, backend): model_class = dynamic_import(module, predefined_st.get(backend, dict())) assert issubclass(model_class, STInterface), f'{module} does not implement STInterface' return model_class
def resnet50(pretrained=False, **kwargs): model = ResNet(Bottleneck, [3, 4, 6, 3], **kwargs) if pretrained: model.load_state_dict(torch.load(os.path.join(models_dir, model_name['resnet50']))) return model
def im_detect_bbox(model, images, target_scale, target_max_size, device, rois=None): transform = T.Compose([T.Resize(target_scale, target_max_size), T.ToTensor(), T.Normalize(mean=cfg.INPUT.PIXEL_MEAN, std=cfg.INPUT.PIXEL_STD, to_bgr255=cfg.INPUT.TO_BGR255)]) t_images = [] t_rois = [] for (image, roi) in zip(images, rois): (t_img, _, t_roi) = transform(image, rois=roi) t_images.append(t_img) t_rois.append(t_roi) t_images = to_image_list(t_images, cfg.DATALOADER.SIZE_DIVISIBILITY) t_rois = [(r.to(device) if (r is not None) else None) for r in t_rois] return model(t_images.to(device), rois=t_rois)
def prepare_img(image, label, num_classes): if (args.data_set == 'Imagenet'): image = tf.image.resize(image, [224, 224]) elif (args.data_set == 'Cifar10'): image = tf.image.resize(image, [32, 32]) label = tf.squeeze(label) label = tf.one_hot(label, num_classes) return (image, label)
class BaseARD(torch.nn.Module): def penalty(self): raise NotImplementedError('Derived classes must compute their own penalty.') def relevance(self, **kwargs): raise NotImplementedError('Derived classes must implement a float mask of relevant coefficients.')
def require_set_backend(): assert (backend is not None), 'distributed backend is not set. Please call `distributed_utils.set_backend_from_args` at the start of your script'
def presnet152(pretrained=False, **kwargs): return presnet(Bottleneck, [3, 8, 36, 3], 'presnet152', pre=pretrained, **kwargs)
def inference_multi_modality_detector(model, pcd, image, ann_file): cfg = model.cfg device = next(model.parameters()).device test_pipeline = deepcopy(cfg.data.test.pipeline) test_pipeline = Compose(test_pipeline) (box_type_3d, box_mode_3d) = get_box_type(cfg.data.test.box_type_3d) data_infos = mmcv.load(ann_file) image_idx = int(re.findall('\\d+', image)[(- 1)]) for x in data_infos: if (int(x['image']['image_idx']) != image_idx): continue info = x break data = dict(pts_filename=pcd, img_prefix=osp.dirname(image), img_info=dict(filename=osp.basename(image)), box_type_3d=box_type_3d, box_mode_3d=box_mode_3d, img_fields=[], bbox3d_fields=[], pts_mask_fields=[], pts_seg_fields=[], bbox_fields=[], mask_fields=[], seg_fields=[]) if (box_mode_3d == Box3DMode.DEPTH): data.update(dict(calib=info['calib'])) data = test_pipeline(data) if (box_mode_3d == Box3DMode.LIDAR): rect = info['calib']['R0_rect'].astype(np.float32) Trv2c = info['calib']['Tr_velo_to_cam'].astype(np.float32) P2 = info['calib']['P2'].astype(np.float32) lidar2img = ((P2 rect) Trv2c) data['img_metas'][0].data['lidar2img'] = lidar2img elif (box_mode_3d == Box3DMode.DEPTH): data['calib'][0]['Rt'] = data['calib'][0]['Rt'].astype(np.float32) data['calib'][0]['K'] = data['calib'][0]['K'].astype(np.float32) data = collate([data], samples_per_gpu=1) if next(model.parameters()).is_cuda: data = scatter(data, [device.index])[0] else: data['img_metas'] = data['img_metas'][0].data data['points'] = data['points'][0].data data['img'] = data['img'][0].data if (box_mode_3d == Box3DMode.DEPTH): data['calib'][0]['Rt'] = data['calib'][0]['Rt'][0].data data['calib'][0]['K'] = data['calib'][0]['K'][0].data with torch.no_grad(): result = model(return_loss=False, rescale=True, **data) return (result, data)
def main(): args = get_args() task_name = 'Task043_BraTS2019' downloaded_data_dir = args.downloaded_data_dir target_base = join(nnUNet_raw_data, task_name) target_imagesTr = join(target_base, 'imagesTr') target_imagesVal = join(target_base, 'imagesVal') target_imagesTs = join(target_base, 'imagesTs') target_labelsTr = join(target_base, 'labelsTr') maybe_mkdir_p(target_imagesTr) maybe_mkdir_p(target_imagesVal) maybe_mkdir_p(target_imagesTs) maybe_mkdir_p(target_labelsTr) patient_names = [] for tpe in ['HGG', 'LGG']: cur = join(downloaded_data_dir, tpe) for p in subdirs(cur, join=False): patdir = join(cur, p) patient_name = ((tpe + '__') + p) patient_names.append(patient_name) t1 = join(patdir, (p + '_t1.nii.gz')) t1c = join(patdir, (p + '_t1ce.nii.gz')) t2 = join(patdir, (p + '_t2.nii.gz')) flair = join(patdir, (p + '_flair.nii.gz')) seg = join(patdir, (p + '_seg.nii.gz')) assert all([isfile(t1), isfile(t1c), isfile(t2), isfile(flair), isfile(seg)]), ('%s' % patient_name) shutil.copy(t1, join(target_imagesTr, (patient_name + '_0000.nii.gz'))) shutil.copy(t1c, join(target_imagesTr, (patient_name + '_0001.nii.gz'))) shutil.copy(t2, join(target_imagesTr, (patient_name + '_0002.nii.gz'))) shutil.copy(flair, join(target_imagesTr, (patient_name + '_0003.nii.gz'))) copy_BraTS_segmentation_and_convert_labels(seg, join(target_labelsTr, (patient_name + '.nii.gz'))) json_dict = OrderedDict() json_dict['name'] = 'BraTS2019' json_dict['description'] = 'nothing' json_dict['tensorImageSize'] = '4D' json_dict['reference'] = 'see BraTS2019' json_dict['licence'] = 'see BraTS2019 license' json_dict['release'] = '0.0' json_dict['modality'] = {'0': 'T1', '1': 'T1ce', '2': 'T2', '3': 'FLAIR'} json_dict['labels'] = {'0': 'background', '1': 'edema', '2': 'non-enhancing', '3': 'enhancing'} json_dict['numTraining'] = len(patient_names) json_dict['numTest'] = 0 json_dict['training'] = [{'image': ('./imagesTr/%s.nii.gz' % i), 'label': ('./labelsTr/%s.nii.gz' % i)} for i in patient_names] json_dict['test'] = [] save_json(json_dict, join(target_base, 'dataset.json'))
def statcast_date_range(start: date, stop: date, step: int, verbose: bool=True) -> Iterator[Tuple[(date, date)]]: low = start while (low <= stop): date_span = (low.replace(month=3, day=15), low.replace(month=11, day=15)) (season_start, season_end) = STATCAST_VALID_DATES.get(low.year, date_span) if (low < season_start): low = season_start if verbose: print('Skipping offseason dates') elif (low > season_end): (low, _) = STATCAST_VALID_DATES.get((low.year + 1), (date(month=3, day=15, year=(low.year + 1)), None)) if verbose: print('Skipping offseason dates') if (low > stop): return high = min((low + timedelta((step - 1))), stop) (yield (low, high)) low += timedelta(days=step)
class RunningSampleData(): index: int session_id: int session: Session asyncio_task: asyncio.Task def __init__(self, index, session_id, session, task): self.index = index self.session_id = session_id self.session = session self.asyncio_task = task
class EnvironmentCommand(BaseTransformersCLICommand): def register_subcommand(parser: ArgumentParser): download_parser = parser.add_parser('env') download_parser.set_defaults(func=info_command_factory) def run(self): pt_version = 'not installed' pt_cuda_available = 'NA' if is_torch_available(): import torch pt_version = torch.__version__ pt_cuda_available = torch.cuda.is_available() tf_version = 'not installed' tf_cuda_available = 'NA' if is_tf_available(): import tensorflow as tf tf_version = tf.__version__ try: tf_cuda_available = tf.test.is_gpu_available() except AttributeError: tf_cuda_available = bool(tf.config.list_physical_devices('GPU')) info = {'`transformers` version': version, 'Platform': platform.platform(), 'Python version': platform.python_version(), 'PyTorch version (GPU?)': '{} ({})'.format(pt_version, pt_cuda_available), 'Tensorflow version (GPU?)': '{} ({})'.format(tf_version, tf_cuda_available), 'Using GPU in script?': '<fill in>', 'Using distributed or parallel set-up in script?': '<fill in>'} print('\nCopy-and-paste the text below in your GitHub issue and FILL OUT the two last points.\n') print(self.format_dict(info)) return info def format_dict(d): return ('\n'.join(['- {}: {}'.format(prop, val) for (prop, val) in d.items()]) + '\n')
class Nlvr2PairedDataset(DetectFeatTxtTokDataset): def __init__(self, txt_db, img_db, use_img_type=True): assert isinstance(txt_db, TxtTokLmdb) assert isinstance(img_db, DetectFeatLmdb) self.txt_db = txt_db self.img_db = img_db (txt_lens, self.ids) = get_ids_and_lens(txt_db) txt2img = txt_db.txt2img self.lens = [((2 * tl) + sum((self.img_db.name2nbb[img] for img in txt2img[id_]))) for (tl, id_) in zip(txt_lens, self.ids)] self.use_img_type = use_img_type def __getitem__(self, i): example = super().__getitem__(i) target = example['target'] outs = [] for (i, img) in enumerate(example['img_fname']): (img_feat, img_pos_feat, num_bb) = self._get_img_feat(img) input_ids = copy.deepcopy(example['input_ids']) input_ids = (([self.txt_db.cls_] + input_ids) + [self.txt_db.sep]) attn_masks = ([1] * (len(input_ids) + num_bb)) input_ids = torch.tensor(input_ids) attn_masks = torch.tensor(attn_masks) if self.use_img_type: img_type_ids = torch.tensor(([(i + 1)] * num_bb)) else: img_type_ids = None outs.append((input_ids, img_feat, img_pos_feat, attn_masks, img_type_ids)) return (tuple(outs), target)
class Collator(): def __init__(self, image_size, url_label, text_label, image_label, name, channels): self.url_label = url_label self.text_label = text_label self.image_label = image_label self.download = (url_label is not None) self.name = name self.channels = channels self.transform = T.Compose([T.Resize(image_size), T.CenterCrop(image_size), T.ToTensor()]) def __call__(self, batch): texts = [] images = [] for item in batch: try: if self.download: image = self.fetch_single_image(item[self.url_label]) else: image = item[self.image_label] image = self.transform(image.convert(self.channels)) except: continue text = t5.t5_encode_text([item[self.text_label]], name=self.name) texts.append(torch.squeeze(text)) images.append(image) if (len(texts) == 0): return None texts = pad_sequence(texts, True) newbatch = [] for i in range(len(texts)): newbatch.append((images[i], texts[i])) return torch.utils.data.dataloader.default_collate(newbatch) def fetch_single_image(self, image_url, timeout=1): try: request = urllib.request.Request(image_url, data=None, headers={'user-agent': USER_AGENT}) with urllib.request.urlopen(request, timeout=timeout) as req: image = Image.open(io.BytesIO(req.read())).convert('RGB') except Exception: image = None return image
def collect_annotations(files, dataset, nproc=1): assert isinstance(files, list) assert isinstance(dataset, str) assert dataset assert isinstance(nproc, int) load_img_info_with_dataset = partial(load_img_info, dataset=dataset) if (nproc > 1): images = mmcv.track_parallel_progress(load_img_info_with_dataset, files, nproc=nproc) else: images = mmcv.track_progress(load_img_info_with_dataset, files) return images
def atari_learn(env, args, num_timesteps): logdir = os.path.join('data', args.exp_name) num_iterations = (float(num_timesteps) / 4.0) def stopping_criterion(env): return (get_wrapper_by_name(env, 'Monitor').get_total_steps() >= num_timesteps) optimizer_spec = OptimizerSpec(constructor=optim.Adam, kwargs=dict(lr=LEARNING_RATE)) exploration_schedule = LinearSchedule(30000, 0.01) logz.configure_output_dir(logdir) if args.dueling: dqn_learning(env=env, method=args.method, game=args.env, q_func=Dueling_DQN, optimizer_spec=optimizer_spec, exploration=exploration_schedule, stopping_criterion=stopping_criterion, replay_buffer_size=REPLAY_BUFFER_SIZE, batch_size=args.batch_size, gamma=args.gamma, learning_starts=LEARNING_STARTS, learning_freq=LEARNING_FREQ, frame_history_len=FRAME_HISTORY_LEN, target_update_freq=TARGET_UPDATE_FREQ, double=args.double, dueling=args.dueling, logdir=logdir, svrl=args.svrl, me_type=args.me_type, maskp=args.maskp, maskstep=args.maskstep, maskscheduler=args.maskscheduler) else: dqn_learning(env=env, method=args.method, game=args.env, q_func=DQN, optimizer_spec=optimizer_spec, exploration=exploration_schedule, stopping_criterion=stopping_criterion, replay_buffer_size=REPLAY_BUFFER_SIZE, batch_size=args.batch_size, gamma=args.gamma, learning_starts=LEARNING_STARTS, learning_freq=LEARNING_FREQ, frame_history_len=FRAME_HISTORY_LEN, target_update_freq=TARGET_UPDATE_FREQ, double=args.double, dueling=args.dueling, logdir=logdir, svrl=args.svrl, me_type=args.me_type, maskp=args.maskp, maskstep=args.maskstep, maskscheduler=args.maskscheduler) env.close()
class TileExtraData(BBAStruct): name_prefix: IdString tile_x: int tile_y: int sites: list[SiteInst] = field(default_factory=list) def serialise_lists(self, context: str, bba: BBAWriter): for (i, site) in enumerate(self.sites): site.serialise_lists(f'{context}_si{i}', bba) bba.label(f'{context}_sites') for (i, site) in enumerate(self.sites): site.serialise(f'{context}_si{i}', bba) def serialise(self, context: str, bba: BBAWriter): bba.u32(self.name_prefix.index) bba.u16(self.tile_x) bba.u16(self.tile_y) bba.slice(f'{context}_sites', len(self.sites))
def get_averaged_groupby(df, groupby, plot_column): return df.groupby(groupby)[plot_column].apply(np.mean)
def export_model(input_model, output_model): print('\nexport model...') model = onnx.load(input_model) model = version_converter.convert_version(model, 14) onnx.save_model(model, output_model)
class VAE(Model): def __init__(self, args): super(VAE, self).__init__(args) if (self.args.dataset_name == 'freyfaces'): h_size = 210 elif (self.args.dataset_name == 'cifar10'): h_size = 384 else: h_size = 294 self.q_z2_layers = nn.Sequential(GatedConv2d(self.args.input_size[0], 32, 7, 1, 3), GatedConv2d(32, 32, 3, 2, 1), GatedConv2d(32, 64, 5, 1, 2), GatedConv2d(64, 64, 3, 2, 1), GatedConv2d(64, 6, 3, 1, 1)) self.q_z2_mean = NonLinear(h_size, self.args.z2_size, activation=None) self.q_z2_logvar = NonLinear(h_size, self.args.z2_size, activation=nn.Hardtanh(min_val=(- 6.0), max_val=2.0)) self.q_z1_layers_x = nn.Sequential(GatedConv2d(self.args.input_size[0], 32, 3, 1, 1), GatedConv2d(32, 32, 3, 2, 1), GatedConv2d(32, 64, 3, 1, 1), GatedConv2d(64, 64, 3, 2, 1), GatedConv2d(64, 6, 3, 1, 1)) self.q_z1_layers_z2 = nn.Sequential(GatedDense(self.args.z2_size, h_size)) self.q_z1_layers_joint = nn.Sequential(GatedDense((2 * h_size), 300)) self.q_z1_mean = NonLinear(300, self.args.z1_size, activation=None) self.q_z1_logvar = NonLinear(300, self.args.z1_size, activation=nn.Hardtanh(min_val=(- 6.0), max_val=2.0)) self.p_z1_layers = nn.Sequential(GatedDense(self.args.z2_size, 300), GatedDense(300, 300)) self.p_z1_mean = NonLinear(300, self.args.z1_size, activation=None) self.p_z1_logvar = NonLinear(300, self.args.z1_size, activation=nn.Hardtanh(min_val=(- 6.0), max_val=2.0)) self.p_x_layers_z1 = nn.Sequential(GatedDense(self.args.z1_size, 300)) self.p_x_layers_z2 = nn.Sequential(GatedDense(self.args.z2_size, 300)) self.p_x_layers_joint_pre = nn.Sequential(GatedDense((2 * 300), np.prod(self.args.input_size))) act = nn.ReLU(True) self.p_x_layers_joint = nn.Sequential(GatedConv2d(self.args.input_size[0], 64, 3, 1, 1), GatedConv2d(64, 64, 3, 1, 1), GatedConv2d(64, 64, 3, 1, 1), GatedConv2d(64, 64, 3, 1, 1)) if (self.args.input_type == 'binary'): self.p_x_mean = Conv2d(64, 1, 1, 1, 0, activation=nn.Sigmoid()) elif ((self.args.input_type == 'gray') or (self.args.input_type == 'continuous')): self.p_x_mean = Conv2d(64, self.args.input_size[0], 1, 1, 0, activation=nn.Sigmoid()) self.p_x_logvar = Conv2d(64, self.args.input_size[0], 1, 1, 0, activation=nn.Hardtanh(min_val=(- 4.5), max_val=0.0)) for m in self.modules(): if isinstance(m, nn.Linear): he_init(m) if (self.args.prior == 'vampprior'): self.add_pseudoinputs() def calculate_loss(self, x, beta=1.0, average=False): (x_mean, x_logvar, z1_q, z1_q_mean, z1_q_logvar, z2_q, z2_q_mean, z2_q_logvar, z1_p_mean, z1_p_logvar) = self.forward(x) if (self.args.input_type == 'binary'): RE = log_Bernoulli(x, x_mean, dim=1) elif ((self.args.input_type == 'gray') or (self.args.input_type == 'continuous')): RE = (- log_Logistic_256(x, x_mean, x_logvar, dim=1)) else: raise Exception('Wrong input type!') log_p_z1 = log_Normal_diag(z1_q, z1_p_mean, z1_p_logvar, dim=1) log_q_z1 = log_Normal_diag(z1_q, z1_q_mean, z1_q_logvar, dim=1) log_p_z2 = self.log_p_z2(z2_q) log_q_z2 = log_Normal_diag(z2_q, z2_q_mean, z2_q_logvar, dim=1) KL = (- (((log_p_z1 + log_p_z2) - log_q_z1) - log_q_z2)) loss = ((- RE) + (beta * KL)) if average: loss = torch.mean(loss) RE = torch.mean(RE) KL = torch.mean(KL) return (loss, RE, KL) def calculate_likelihood(self, X, dir, mode='test', S=5000, MB=500): N_test = X.size(0) likelihood_test = [] if (S <= MB): R = 1 else: R = (S / MB) S = MB for j in range(N_test): if ((j % 100) == 0): print('{:.2f}%'.format(((j / (1.0 * N_test)) * 100))) x_single = X[j].unsqueeze(0) a = [] for r in range(0, int(R)): x = x_single.expand(S, x_single.size(1)).contiguous() (a_tmp, _, _) = self.calculate_loss(x) a.append((- a_tmp.cpu().data.numpy())) a = np.asarray(a) a = np.reshape(a, ((a.shape[0] * a.shape[1]), 1)) likelihood_x = logsumexp(a) likelihood_test.append((likelihood_x - np.log(len(a)))) likelihood_test = np.array(likelihood_test) plot_histogram((- likelihood_test), dir, mode) return (- np.mean(likelihood_test)) def calculate_lower_bound(self, X_full, MB=500): lower_bound = 0.0 RE_all = 0.0 KL_all = 0.0 I = int(math.ceil((X_full.size(0) / MB))) for i in range(I): x = X_full[(i * MB):((i + 1) * MB)].view((- 1), np.prod(self.args.input_size)) (loss, RE, KL) = self.calculate_loss(x, average=True) RE_all += RE.cpu().data[0] KL_all += KL.cpu().data[0] lower_bound += loss.cpu().data[0] lower_bound /= I return lower_bound def generate_x(self, N=25): if (self.args.prior == 'standard'): z2_sample_rand = Variable(torch.FloatTensor(N, self.args.z1_size).normal_()) if self.args.cuda: z2_sample_rand = z2_sample_rand.cuda() elif (self.args.prior == 'vampprior'): means = self.means(self.idle_input)[0:N].view((- 1), self.args.input_size[0], self.args.input_size[1], self.args.input_size[2]) (z2_sample_gen_mean, z2_sample_gen_logvar) = self.q_z2(means) z2_sample_rand = self.reparameterize(z2_sample_gen_mean, z2_sample_gen_logvar) (z1_sample_mean, z1_sample_logvar) = self.p_z1(z2_sample_rand) z1_sample_rand = self.reparameterize(z1_sample_mean, z1_sample_logvar) (samples_gen, _) = self.p_x(z1_sample_rand, z2_sample_rand) return samples_gen def reconstruct_x(self, x): (x_reconstructed, _, _, _, _, _, _, _, _, _) = self.forward(x) return x_reconstructed def q_z2(self, x): h = self.q_z2_layers(x) h = h.view(x.size(0), (- 1)) z2_q_mean = self.q_z2_mean(h) z2_q_logvar = self.q_z2_logvar(h) return (z2_q_mean, z2_q_logvar) def q_z1(self, x, z2): x = self.q_z1_layers_x(x) x = x.view(x.size(0), (- 1)) z2 = self.q_z1_layers_z2(z2) h = torch.cat((x, z2), 1) h = self.q_z1_layers_joint(h) z1_q_mean = self.q_z1_mean(h) z1_q_logvar = self.q_z1_logvar(h) return (z1_q_mean, z1_q_logvar) def p_z1(self, z2): z2 = self.p_z1_layers(z2) z1_p_mean = self.p_z1_mean(z2) z1_p_logvar = self.p_z1_logvar(z2) return (z1_p_mean, z1_p_logvar) def p_x(self, z1, z2): z2 = self.p_x_layers_z2(z2) z1 = self.p_x_layers_z1(z1) h = torch.cat((z1, z2), 1) h = self.p_x_layers_joint_pre(h) h = h.view((- 1), self.args.input_size[0], self.args.input_size[1], self.args.input_size[2]) h_decoder = self.p_x_layers_joint(h) x_mean = self.p_x_mean(h_decoder).view((- 1), np.prod(self.args.input_size)) if (self.args.input_type == 'binary'): x_logvar = 0.0 else: x_mean = torch.clamp(x_mean, min=(0.0 + (1.0 / 512.0)), max=(1.0 - (1.0 / 512.0))) x_logvar = self.p_x_logvar(h_decoder).view((- 1), np.prod(self.args.input_size)) return (x_mean, x_logvar) def log_p_z2(self, z2): if (self.args.prior == 'standard'): log_prior = log_Normal_standard(z2, dim=1) elif (self.args.prior == 'vampprior'): C = self.args.number_components X = self.means(self.idle_input).view((- 1), self.args.input_size[0], self.args.input_size[1], self.args.input_size[2]) (z2_p_mean, z2_p_logvar) = self.q_z2(X) z_expand = z2.unsqueeze(1) means = z2_p_mean.unsqueeze(0) logvars = z2_p_logvar.unsqueeze(0) a = (log_Normal_diag(z_expand, means, logvars, dim=2) - math.log(C)) (a_max, _) = torch.max(a, 1) log_prior = (a_max + torch.log(torch.sum(torch.exp((a - a_max.unsqueeze(1))), 1))) else: raise Exception('Wrong name of the prior!') return log_prior def forward(self, x): x = x.view((- 1), self.args.input_size[0], self.args.input_size[1], self.args.input_size[2]) (z2_q_mean, z2_q_logvar) = self.q_z2(x) z2_q = self.reparameterize(z2_q_mean, z2_q_logvar) (z1_q_mean, z1_q_logvar) = self.q_z1(x, z2_q) z1_q = self.reparameterize(z1_q_mean, z1_q_logvar) (z1_p_mean, z1_p_logvar) = self.p_z1(z2_q) (x_mean, x_logvar) = self.p_x(z1_q, z2_q) return (x_mean, x_logvar, z1_q, z1_q_mean, z1_q_logvar, z2_q, z2_q_mean, z2_q_logvar, z1_p_mean, z1_p_logvar)
def add_dataset_config(cfg): _C = cfg _C.MODEL.ROI_HEADS.NUM_OUTPUT_CLASSES = 80 _C.MODEL.ROI_HEADS.EMBEDDINGS_PATH = '' _C.MODEL.ROI_HEADS.EMBEDDINGS_PATH_COCO = '' _C.MODEL.ROI_HEADS.LINGUAL_MATRIX_THRESHOLD = 0.05 _C.MODEL.ROI_HEADS.MASK_NUM_CLASSES = 80 _C.MODEL.FREEZE_LAYERS = CN() _C.MODEL.FREEZE_LAYERS.META_ARCH = [] _C.MODEL.FREEZE_LAYERS.ROI_HEADS = [] _C.DATASETS.TYPE = '' _C.DATASETS.VISUAL_GENOME = CN() _C.DATASETS.VISUAL_GENOME.IMAGES = './datasets/vg/VG_100K/' _C.DATASETS.VISUAL_GENOME.MAPPING_DICTIONARY = './datasets/vg/orig_and_generated_files/VG-SGG-dicts-with-attri.json' _C.DATASETS.VISUAL_GENOME.IMAGE_DATA = './datasets/vg/orig_and_generated_files/image_data.json' _C.DATASETS.VISUAL_GENOME.VG_ATTRIBUTE_H5 = './datasets/vg/orig_and_generated_files/VG-SGG-with-attri.h5' _C.DATASETS.VISUAL_GENOME.TRAIN_MASKS = '' _C.DATASETS.VISUAL_GENOME.TEST_MASKS = '' _C.DATASETS.VISUAL_GENOME.VAL_MASKS = '' _C.DATASETS.VISUAL_GENOME.CLIPPED = False _C.DATASETS.ARXIVDOCS_TARGET = CN() _C.DATASETS.ARXIVDOCS_TARGET.CLIPPED = False _C.DATASETS.ARXIVDOCS_TARGET.FILTER_EMPTY_RELATIONS = True _C.DATASETS.ARXIVDOCS_TARGET.FILTER_DUPLICATE_RELATIONS = True _C.DATASETS.ARXIVDOCS_TARGET.FILTER_NON_OVERLAP = False _C.DATASETS.ARXIVDOCS_TARGET.BOX_SCALE = 1024 _C.DATASETS.ARXIVDOCS_TARGET.TRAIN_MASKS = './datasets/ADtgt_VGv2/additional_processed_anns/train/attribute_files/arxivdocs_target_layout_v2_train_scene_graph_segmentations.json' _C.DATASETS.ARXIVDOCS_TARGET.TEST_MASKS = './datasets/ADtgt_VGv2/additional_processed_anns/test/attribute_files/arxivdocs_target_layout_v2_test_scene_graph_segmentations.json' _C.DATASETS.ARXIVDOCS_TARGET.VAL_MASKS = './datasets/ADtgt_VGv2/additional_processed_anns/dev/attribute_files/arxivdocs_target_layout_v2_dev_scene_graph_segmentations.json' _C.DATASETS.ARXIVDOCS_TARGET.TRAIN_IMAGES = './datasets/ADtgt_VGv2/imgs/train' _C.DATASETS.ARXIVDOCS_TARGET.TRAIN_MAPPING_DICTIONARY = './datasets/ADtgt_VGv2/additional_processed_anns/train/attribute_files/arxivdocs_target_layout_v2_train_scene_graph_dicts_with_attri.json' _C.DATASETS.ARXIVDOCS_TARGET.TRAIN_IMAGE_DATA = './datasets/ADtgt_VGv2/anns/train/arxivdocs_target_layout_v2_train_scene_graph_image_data.json' _C.DATASETS.ARXIVDOCS_TARGET.TRAIN_ARXIVDOCS_TARGET_ATTRIBUTE_H5 = './datasets/ADtgt_VGv2/additional_processed_anns/train/attribute_files/arxivdocs_target_layout_v2_train_scene_graph_with_attri.h5' _C.DATASETS.ARXIVDOCS_TARGET.VAL_IMAGES = './datasets/ADtgt_VGv2/imgs/dev' _C.DATASETS.ARXIVDOCS_TARGET.VAL_MAPPING_DICTIONARY = './datasets/ADtgt_VGv2/additional_processed_anns/dev/attribute_files/arxivdocs_target_layout_v2_dev_scene_graph_dicts_with_attri.json' _C.DATASETS.ARXIVDOCS_TARGET.VAL_IMAGE_DATA = './datasets/ADtgt_VGv2/anns/dev/arxivdocs_target_layout_v2_dev_scene_graph_image_data.json' _C.DATASETS.ARXIVDOCS_TARGET.VAL_ARXIVDOCS_TARGET_ATTRIBUTE_H5 = './datasets/ADtgt_VGv2/additional_processed_anns/dev/attribute_files/arxivdocs_target_layout_v2_dev_scene_graph_with_attri.h5' _C.DATASETS.ARXIVDOCS_TARGET.TEST_IMAGES = './datasets/ADtgt_VGv2/imgs/test' _C.DATASETS.ARXIVDOCS_TARGET.TEST_MAPPING_DICTIONARY = './datasets/ADtgt_VGv2/additional_processed_anns/test/attribute_files/arxivdocs_target_layout_v2_test_scene_graph_dicts_with_attri.json' _C.DATASETS.ARXIVDOCS_TARGET.TEST_IMAGE_DATA = './datasets/ADtgt_VGv2/anns/test/arxivdocs_target_layout_v2_test_scene_graph_image_data.json' _C.DATASETS.ARXIVDOCS_TARGET.TEST_ARXIVDOCS_TARGET_ATTRIBUTE_H5 = './datasets/ADtgt_VGv2/additional_processed_anns/test/attribute_files/arxivdocs_target_layout_v2_test_scene_graph_with_attri.h5' _C.DATASETS.ARXIVDOCS_WEAK = CN() _C.DATASETS.ARXIVDOCS_WEAK.CLIPPED = False _C.DATASETS.ARXIVDOCS_WEAK.FILTER_EMPTY_RELATIONS = True _C.DATASETS.ARXIVDOCS_WEAK.FILTER_DUPLICATE_RELATIONS = True _C.DATASETS.ARXIVDOCS_WEAK.FILTER_NON_OVERLAP = False _C.DATASETS.ARXIVDOCS_WEAK.BOX_SCALE = 1024 _C.DATASETS.ARXIVDOCS_WEAK.TRAIN_MASKS = './datasets/ADwk_VGv2/additional_processed_anns/train/attribute_files/arxivdocs_weak_layout_v2_train_scene_graph_segmentations.json' _C.DATASETS.ARXIVDOCS_WEAK.TEST_MASKS = './datasets/ADwk_VGv2/additional_processed_anns/test/attribute_files/arxivdocs_weak_layout_v2_test_scene_graph_segmentations.json' _C.DATASETS.ARXIVDOCS_WEAK.VAL_MASKS = './datasets/ADwk_VGv2/additional_processed_anns/dev/attribute_files/arxivdocs_weak_layout_v2_dev_scene_graph_segmentations.json' _C.DATASETS.ARXIVDOCS_WEAK.TRAIN_IMAGES = './datasets/ADwk_VGv2/imgs/train' _C.DATASETS.ARXIVDOCS_WEAK.TRAIN_MAPPING_DICTIONARY = './datasets/ADwk_VGv2/additional_processed_anns/train/attribute_files/arxivdocs_weak_layout_v2_train_scene_graph_dicts_with_attri.json' _C.DATASETS.ARXIVDOCS_WEAK.TRAIN_IMAGE_DATA = './datasets/ADwk_VGv2/anns/train/arxivdocs_weak_layout_v2_train_scene_graph_image_data.json' _C.DATASETS.ARXIVDOCS_WEAK.TRAIN_ARXIVDOCS_WEAK_ATTRIBUTE_H5 = './datasets/ADwk_VGv2/additional_processed_anns/train/attribute_files/arxivdocs_weak_layout_v2_train_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA = CN() _C.DATASETS.EPERIODICA.CLIPPED = False _C.DATASETS.EPERIODICA.FILTER_EMPTY_RELATIONS = True _C.DATASETS.EPERIODICA.FILTER_DUPLICATE_RELATIONS = True _C.DATASETS.EPERIODICA.FILTER_NON_OVERLAP = False _C.DATASETS.EPERIODICA.BOX_SCALE = 1024 _C.DATASETS.EPERIODICA.TRAIN_MASKS = '' _C.DATASETS.EPERIODICA.VAL_MASKS = '' _C.DATASETS.EPERIODICA.TEST_MASKS = '' _C.DATASETS.EPERIODICA.TRAIN_MASKS = './datasets/eperiodica_v3/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_segmentations.json' _C.DATASETS.EPERIODICA.VAL_MASKS = './datasets/eperiodica_v3/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_segmentations.json' _C.DATASETS.EPERIODICA.TEST_MASKS = './datasets/eperiodica_v3/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_segmentations.json' _C.DATASETS.EPERIODICA.TRAIN_IMAGES = './datasets/eperiodica_v3/imgs/train' _C.DATASETS.EPERIODICA.TRAIN_MAPPING_DICTIONARY = './datasets/eperiodica_v3/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA.TRAIN_IMAGE_DATA = './datasets/eperiodica_v3/anns/train/eperiodica_minitrain/eperiodica_minitrain_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA.TRAIN_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica_v3/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA.VAL_IMAGES = './datasets/eperiodica_v3/imgs/val' _C.DATASETS.EPERIODICA.VAL_MAPPING_DICTIONARY = './datasets/eperiodica_v3/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA.VAL_IMAGE_DATA = './datasets/eperiodica_v3/anns/val/eperiodica_minival/eperiodica_minival_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA.VAL_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica_v3/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA.TEST_IMAGES = './datasets/eperiodica_v3/imgs/test' _C.DATASETS.EPERIODICA.TEST_MAPPING_DICTIONARY = './datasets/eperiodica_v3/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA.TEST_IMAGE_DATA = './datasets/eperiodica_v3/anns/test/eperiodica_minitest/eperiodica_minitest_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA.TEST_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica_v3/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA2 = CN() _C.DATASETS.EPERIODICA2.CLIPPED = False _C.DATASETS.EPERIODICA2.FILTER_EMPTY_RELATIONS = True _C.DATASETS.EPERIODICA2.FILTER_DUPLICATE_RELATIONS = True _C.DATASETS.EPERIODICA2.FILTER_NON_OVERLAP = False _C.DATASETS.EPERIODICA2.BOX_SCALE = 1024 _C.DATASETS.EPERIODICA2.TRAIN_MASKS = '' _C.DATASETS.EPERIODICA2.TEST_MASKS = '' _C.DATASETS.EPERIODICA2.VAL_MASKS = '' _C.DATASETS.EPERIODICA2.TRAIN_IMAGES = './datasets/eperiodica2/imgs/train' _C.DATASETS.EPERIODICA2.TRAIN_MAPPING_DICTIONARY = './datasets/eperiodica2/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA2.TRAIN_IMAGE_DATA = './datasets/eperiodica2/anns/train/eperiodica_minitrain/eperiodica_minitrain_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA2.TRAIN_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica2/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA2.VAL_IMAGES = './datasets/eperiodica2/imgs/val' _C.DATASETS.EPERIODICA2.VAL_MAPPING_DICTIONARY = './datasets/eperiodica2/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA2.VAL_IMAGE_DATA = './datasets/eperiodica2/anns/val/eperiodica_minival/eperiodica_minival_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA2.VAL_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica2/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA2.TEST_IMAGES = './datasets/eperiodica2/imgs/test' _C.DATASETS.EPERIODICA2.TEST_MAPPING_DICTIONARY = './datasets/eperiodica2/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA2.TEST_IMAGE_DATA = './datasets/eperiodica2/anns/test/eperiodica_minitest/eperiodica_minitest_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA2.TEST_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica2/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA3 = CN() _C.DATASETS.EPERIODICA3.CLIPPED = False _C.DATASETS.EPERIODICA3.FILTER_EMPTY_RELATIONS = True _C.DATASETS.EPERIODICA3.FILTER_DUPLICATE_RELATIONS = True _C.DATASETS.EPERIODICA3.FILTER_NON_OVERLAP = False _C.DATASETS.EPERIODICA3.BOX_SCALE = 1024 _C.DATASETS.EPERIODICA3.TRAIN_MASKS = './datasets/eperiodica3/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_segmentations.json' _C.DATASETS.EPERIODICA3.VAL_MASKS = './datasets/eperiodica3/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_segmentations.json' _C.DATASETS.EPERIODICA3.TEST_MASKS = './datasets/eperiodica3/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_segmentations.json' _C.DATASETS.EPERIODICA3.TRAIN_IMAGES = './datasets/eperiodica3/imgs/train' _C.DATASETS.EPERIODICA3.TRAIN_MAPPING_DICTIONARY = './datasets/eperiodica3/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA3.TRAIN_IMAGE_DATA = './datasets/eperiodica3/anns/train/eperiodica_minitrain/eperiodica_minitrain_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA3.TRAIN_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica3/additional_processed_anns/train/eperiodica_minitrain/attribute_files/eperiodica_minitrain_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA3.VAL_IMAGES = './datasets/eperiodica3/imgs/val' _C.DATASETS.EPERIODICA3.VAL_MAPPING_DICTIONARY = './datasets/eperiodica3/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA3.VAL_IMAGE_DATA = './datasets/eperiodica3/anns/val/eperiodica_minival/eperiodica_minival_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA3.VAL_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica3/additional_processed_anns/val/eperiodica_minival/attribute_files/eperiodica_minival_VG_scene_graph_with_attri.h5' _C.DATASETS.EPERIODICA3.TEST_IMAGES = './datasets/eperiodica3/imgs/test' _C.DATASETS.EPERIODICA3.TEST_MAPPING_DICTIONARY = './datasets/eperiodica3/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_dicts_with_attri.json' _C.DATASETS.EPERIODICA3.TEST_IMAGE_DATA = './datasets/eperiodica3/anns/test/eperiodica_minitest/eperiodica_minitest_VG_scene_graph_image_data.json' _C.DATASETS.EPERIODICA3.TEST_EPERIODICA_TARGET_ATTRIBUTE_H5 = './datasets/eperiodica3/additional_processed_anns/test/eperiodica_minitest/attribute_files/eperiodica_minitest_VG_scene_graph_with_attri.h5' _C.DATASETS.MSCOCO = CN() _C.DATASETS.MSCOCO.ANNOTATIONS = '' _C.DATASETS.MSCOCO.DATAROOT = '' _C.DATASETS.VISUAL_GENOME.FILTER_EMPTY_RELATIONS = True _C.DATASETS.VISUAL_GENOME.FILTER_DUPLICATE_RELATIONS = True _C.DATASETS.VISUAL_GENOME.FILTER_NON_OVERLAP = True _C.DATASETS.VISUAL_GENOME.NUMBER_OF_VALIDATION_IMAGES = 5000 _C.DATASETS.VISUAL_GENOME.BOX_SCALE = 1024 _C.DATASETS.SEG_DATA_DIVISOR = 1 _C.DATASETS.TRANSFER = ('coco_train_2014',) _C.DATASETS.MASK_TRAIN = ('coco_train_2017',) _C.DATASETS.MASK_TEST = ('coco_val_2017',)
class AttnSkipUpBlock2DTests(UNetBlockTesterMixin, unittest.TestCase): block_class = AttnSkipUpBlock2D block_type = 'up' def dummy_input(self): return super().get_dummy_input(include_res_hidden_states_tuple=True) def test_output(self): expected_slice = [0.0361, 0.0617, 0.2787, (- 0.035), 0.0342, 0.3421, (- 0.0843), 0.0913, 0.3015] super().test_output(expected_slice)
class PreActResNet(nn.Module): def __init__(self, num_classes: int=10, depth: int=18, width: int=0, activation_fn: nn.Module=nn.ReLU, mean: Union[(Tuple[(float, ...)], float)]=CIFAR10_MEAN, std: Union[(Tuple[(float, ...)], float)]=CIFAR10_STD, padding: int=0, num_input_channels: int=3): super().__init__() if (width != 0): raise ValueError('Unsupported `width`.') self.mean = torch.tensor(mean).view(num_input_channels, 1, 1) self.std = torch.tensor(std).view(num_input_channels, 1, 1) self.mean_cuda = None self.std_cuda = None self.padding = padding self.conv_2d = nn.Conv2d(num_input_channels, 64, kernel_size=3, stride=1, padding=1, bias=False) if (depth == 18): num_blocks = (2, 2, 2, 2) elif (depth == 34): num_blocks = (3, 4, 6, 3) else: raise ValueError('Unsupported `depth`.') self.layer_0 = self._make_layer(64, 64, num_blocks[0], 1, activation_fn) self.layer_1 = self._make_layer(64, 128, num_blocks[1], 2, activation_fn) self.layer_2 = self._make_layer(128, 256, num_blocks[2], 2, activation_fn) self.layer_3 = self._make_layer(256, 512, num_blocks[3], 2, activation_fn) self.batchnorm = nn.BatchNorm2d(512, momentum=0.01) self.relu = activation_fn() self.logits = nn.Linear(512, num_classes) def _make_layer(self, in_planes, out_planes, num_blocks, stride, activation_fn): layers = [] for (i, stride) in enumerate(([stride] + ([1] * (num_blocks - 1)))): layers.append(_PreActBlock((((i == 0) and in_planes) or out_planes), out_planes, stride, activation_fn)) return nn.Sequential(*layers) def forward(self, x): if (self.padding > 0): x = F.pad(x, ((self.padding,) * 4)) if x.is_cuda: if (self.mean_cuda is None): self.mean_cuda = self.mean.cuda() self.std_cuda = self.std.cuda() out = ((x - self.mean_cuda) / self.std_cuda) else: out = ((x - self.mean) / self.std) out = self.conv_2d(out) out = self.layer_0(out) out = self.layer_1(out) out = self.layer_2(out) out = self.layer_3(out) out = self.relu(self.batchnorm(out)) out = F.avg_pool2d(out, 4) out = out.view(out.size(0), (- 1)) return self.logits(out)
def to_bigdl_optim_method(optimizer): optimizer = optimizer.lower() if (optimizer == 'adagrad'): return Adagrad(learningrate=0.01) elif (optimizer == 'sgd'): return SGD(learningrate=0.01) elif (optimizer == 'adam'): return Adam() elif (optimizer == 'rmsprop'): return RMSprop(learningrate=0.001, decayrate=0.9) elif (optimizer == 'adadelta'): return Adadelta(decayrate=0.95, epsilon=1e-08) elif (optimizer == 'adamax'): return Adamax(epsilon=1e-08) else: invalidInputError(False, ('Unsupported optimizer: %s' % optimizer))
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) sampler = AsyncCpuSampler(EnvCls=gym_make, env_kwargs=config['env'], CollectorCls=DbCpuResetCollector, eval_env_kwargs=config['env'], **config['sampler']) algo = TD3(optim_kwargs=config['optim'], **config['algo']) agent = Td3Agent(**config['agent']) runner = AsyncRlEval(algo=algo, agent=agent, sampler=sampler, affinity=affinity, **config['runner']) name = ('async_td3_' + config['env']['id']) with logger_context(log_dir, run_ID, name, config): runner.train()
def mkdir_p(path): try: os.makedirs(os.path.abspath(path)) except OSError as exc: if ((exc.errno == errno.EEXIST) and os.path.isdir(path)): pass else: raise
def check_all_inits(): failures = [] for (root, _, files) in os.walk(PATH_TO_TRANSFORMERS): if ('__init__.py' in files): fname = os.path.join(root, '__init__.py') objects = parse_init(fname) if (objects is not None): errors = analyze_results(*objects) if (len(errors) > 0): errors[0] = f'''Problem in {fname}, both halves do not define the same objects. {errors[0]}''' failures.append('\n'.join(errors)) if (len(failures) > 0): raise ValueError('\n\n'.join(failures))
_module() class WIDERFaceDataset(XMLDataset): CLASSES = ('face',) PALETTE = [(0, 255, 0)] def __init__(self, **kwargs): super(WIDERFaceDataset, self).__init__(**kwargs) def load_annotations(self, ann_file): data_infos = [] img_ids = mmcv.list_from_file(ann_file) for img_id in img_ids: filename = f'{img_id}.jpg' xml_path = osp.join(self.img_prefix, 'Annotations', f'{img_id}.xml') tree = ET.parse(xml_path) root = tree.getroot() size = root.find('size') width = int(size.find('width').text) height = int(size.find('height').text) folder = root.find('folder').text data_infos.append(dict(id=img_id, filename=osp.join(folder, filename), width=width, height=height)) return data_infos
class CategoricalMLPRegressor(StochasticRegressor): def __init__(self, input_shape, output_dim, name='CategoricalMLPRegressor', hidden_sizes=(32, 32), hidden_nonlinearity=tf.nn.tanh, hidden_w_init=tf.initializers.glorot_uniform(seed=deterministic.get_tf_seed_stream()), hidden_b_init=tf.zeros_initializer(), output_nonlinearity=tf.nn.softmax, output_w_init=tf.initializers.glorot_uniform(seed=deterministic.get_tf_seed_stream()), output_b_init=tf.zeros_initializer(), optimizer=None, optimizer_args=None, tr_optimizer=None, tr_optimizer_args=None, use_trust_region=True, max_kl_step=0.01, normalize_inputs=True, layer_normalization=False): super().__init__(input_shape, output_dim, name) self._use_trust_region = use_trust_region self._max_kl_step = max_kl_step self._normalize_inputs = normalize_inputs with tf.compat.v1.variable_scope(self._name, reuse=False) as vs: self._variable_scope = vs if (optimizer_args is None): optimizer_args = dict() if (tr_optimizer_args is None): tr_optimizer_args = dict() if (optimizer is None): self._optimizer = make_optimizer(LbfgsOptimizer, **optimizer_args) else: self._optimizer = make_optimizer(optimizer, **optimizer_args) if (tr_optimizer is None): self._tr_optimizer = make_optimizer(ConjugateGradientOptimizer, **tr_optimizer_args) else: self._tr_optimizer = make_optimizer(tr_optimizer, **tr_optimizer_args) self._first_optimized = False self.model = CategoricalMLPRegressorModel(input_shape, output_dim, hidden_sizes=hidden_sizes, hidden_nonlinearity=hidden_nonlinearity, hidden_w_init=hidden_w_init, hidden_b_init=hidden_b_init, output_nonlinearity=output_nonlinearity, output_w_init=output_w_init, output_b_init=output_b_init, layer_normalization=layer_normalization) self._old_model = self.model.clone(name='model_for_old_dist') self._network = None self._old_network = None self._initialize() def _initialize(self): input_var = tf.compat.v1.placeholder(tf.float32, shape=((None,) + self._input_shape)) self._old_network = self._old_model.build(input_var) with tf.compat.v1.variable_scope(self._variable_scope): self._network = self.model.build(input_var) self._old_model.parameters = self.model.parameters ys_var = tf.compat.v1.placeholder(dtype=tf.float32, name='ys', shape=(None, self._output_dim)) y_hat = self._network.y_hat dist = self._network.dist old_dist = self._old_network.dist mean_kl = tf.reduce_mean(old_dist.kl_divergence(dist)) loss = (- tf.reduce_mean(dist.log_prob(ys_var))) predicted = tf.one_hot(tf.argmax(y_hat, axis=1), depth=self._output_dim) self._f_predict = tensor_utils.compile_function([input_var], predicted) self._optimizer.update_opt(loss=loss, target=self, inputs=[input_var, ys_var]) self._tr_optimizer.update_opt(loss=loss, target=self, inputs=[input_var, ys_var], leq_constraint=(mean_kl, self._max_kl_step)) def fit(self, xs, ys): if self._normalize_inputs: self._network.x_mean.load(np.mean(xs, axis=0, keepdims=True)) self._network.x_std.load(np.std(xs, axis=0, keepdims=True)) self._old_network.x_mean.load(np.mean(xs, axis=0, keepdims=True)) self._old_network.x_std.load(np.std(xs, axis=0, keepdims=True)) inputs = [xs, ys] if self._use_trust_region: optimizer = self._tr_optimizer else: optimizer = self._optimizer loss_before = optimizer.loss(inputs) tabular.record('{}/LossBefore'.format(self._name), loss_before) optimizer.optimize(inputs) loss_after = optimizer.loss(inputs) tabular.record('{}/LossAfter'.format(self._name), loss_after) tabular.record('{}/dLoss'.format(self._name), (loss_before - loss_after)) self._first_optimized = True self._old_model.parameters = self.model.parameters def predict(self, xs): return self._f_predict(xs) def recurrent(self): return False def vectorized(self): return True def distribution(self): return self._network.dist def __getstate__(self): new_dict = super().__getstate__() del new_dict['_f_predict'] del new_dict['_network'] del new_dict['_old_network'] return new_dict def __setstate__(self, state): super().__setstate__(state) self._initialize()
def find_all_best_thresh(main_eval, preds, exact_raw, f1_raw, na_probs, qid_to_has_ans): (best_exact, exact_thresh) = find_best_thresh(preds, exact_raw, na_probs, qid_to_has_ans) (best_f1, f1_thresh) = find_best_thresh(preds, f1_raw, na_probs, qid_to_has_ans) main_eval['best_exact'] = best_exact main_eval['best_exact_thresh'] = exact_thresh main_eval['best_f1'] = best_f1 main_eval['best_f1_thresh'] = f1_thresh
class EMAML(): def __init__(self, dim_input, dim_output, dim_hidden=32, num_layers=4, num_particles=2, max_test_step=5): self.dim_input = dim_input self.dim_output = dim_output self.dim_hidden = dim_hidden self.num_layers = num_layers self.num_particles = num_particles self.in_lr = tf.placeholder_with_default(input=FLAGS.in_lr, name='in_lr', shape=[]) self.out_lr = tf.placeholder_with_default(input=FLAGS.out_lr, name='out_lr', shape=[]) self.max_test_step = max_test_step self.bnn = BNN(dim_input=self.dim_input, dim_output=self.dim_output, dim_hidden=self.dim_hidden, num_layers=self.num_layers, is_bnn=False) self.construct_network_weights = self.bnn.construct_network_weights self.forward_network = self.bnn.forward_network self.train_x = tf.placeholder(dtype=tf.float32, name='train_x') self.train_y = tf.placeholder(dtype=tf.float32, name='train_y') self.valid_x = tf.placeholder(dtype=tf.float32, name='valid_x') self.valid_y = tf.placeholder(dtype=tf.float32, name='valid_y') self.W_network_particles = None def construct_model(self, is_training=True): print('start model construction') with tf.variable_scope('model', reuse=None) as training_scope: if (is_training or (self.W_network_particles is None)): self.W_network_particles = [self.construct_network_weights(scope='network{}'.format(p_idx)) for p_idx in range(self.num_particles)] else: training_scope.reuse_variables() if is_training: max_update_step = FLAGS.in_step else: max_update_step = max(FLAGS.in_step, self.max_test_step) def fast_learn_one_task(inputs): [train_x, valid_x, train_y, valid_y] = inputs meta_loss = [] WW_update = [OrderedDict(zip(W_dic.keys(), W_dic.values())) for W_dic in self.W_network_particles] step_train_loss = ([None] * (max_update_step + 1)) step_valid_loss = ([None] * (max_update_step + 1)) step_train_pred = ([None] * (max_update_step + 1)) step_valid_pred = ([None] * (max_update_step + 1)) for s_idx in range((max_update_step + 1)): train_z_list = [] valid_z_list = [] train_mse_list = [] valid_mse_list = [] for p_idx in range(FLAGS.num_particles): train_z_list.append(self.forward_network(x=train_x, W_dict=WW_update[p_idx])) valid_z_list.append(self.forward_network(x=valid_x, W_dict=WW_update[p_idx])) train_mse_list.append(self.bnn.mse_data(predict_y=train_z_list[(- 1)], target_y=train_y)) valid_mse_list.append(self.bnn.mse_data(predict_y=valid_z_list[(- 1)], target_y=valid_y)) if (s_idx < max_update_step): particle_loss = tf.reduce_mean(train_mse_list[(- 1)]) dWp = tf.gradients(ys=particle_loss, xs=list(WW_update[p_idx].values())) if FLAGS.stop_grad: dWp = [tf.stop_gradient(grad) for grad in dWp] dWp = OrderedDict(zip(WW_update[p_idx].keys(), dWp)) param_names = [] param_vals = [] for key in list(WW_update[p_idx].keys()): if (FLAGS.in_grad_clip > 0): grad = tf.clip_by_value(dWp[key], (- FLAGS.in_grad_clip), FLAGS.in_grad_clip) else: grad = dWp[key] param_names.append(key) param_vals.append((WW_update[p_idx][key] - (self.in_lr * grad))) WW_update[p_idx] = OrderedDict(zip(param_names, param_vals)) else: meta_loss.append(tf.reduce_mean(valid_mse_list[(- 1)])) step_train_loss[s_idx] = tf.reduce_mean([tf.reduce_mean(train_mse) for train_mse in train_mse_list]) step_valid_loss[s_idx] = tf.reduce_mean([tf.reduce_mean(valid_mse) for valid_mse in valid_mse_list]) step_train_pred[s_idx] = tf.concat([tf.expand_dims(train_z, 0) for train_z in train_z_list], axis=0) step_valid_pred[s_idx] = tf.concat([tf.expand_dims(valid_z, 0) for valid_z in valid_z_list], axis=0) meta_loss = tf.reduce_sum(meta_loss) return [step_train_loss, step_valid_loss, step_train_pred, step_valid_pred, meta_loss] out_dtype = [([tf.float32] * (max_update_step + 1)), ([tf.float32] * (max_update_step + 1)), ([tf.float32] * (max_update_step + 1)), ([tf.float32] * (max_update_step + 1)), tf.float32] result = tf.map_fn(fast_learn_one_task, elems=[self.train_x, self.valid_x, self.train_y, self.valid_y], dtype=out_dtype, parallel_iterations=FLAGS.num_tasks) full_step_train_loss = result[0] full_step_valid_loss = result[1] full_step_train_pred = result[2] full_step_valid_pred = result[3] full_meta_loss = result[4] if is_training: self.total_train_loss = [tf.reduce_mean(full_step_train_loss[j]) for j in range((FLAGS.in_step + 1))] self.total_valid_loss = [tf.reduce_mean(full_step_valid_loss[j]) for j in range((FLAGS.in_step + 1))] self.total_meta_loss = tf.reduce_mean(full_meta_loss) self.total_train_z_list = full_step_train_pred self.total_valid_z_list = full_step_valid_pred update_params_list = [] update_params_name = [] for p in range(FLAGS.num_particles): for name in self.W_network_particles[0].keys(): update_params_name.append([p, name]) update_params_list.append(self.W_network_particles[p][name]) optimizer = tf.train.AdamOptimizer(learning_rate=self.out_lr) gv_list = optimizer.compute_gradients(loss=self.total_meta_loss, var_list=update_params_list) if (FLAGS.out_grad_clip > 0): gv_list = [(tf.clip_by_value(grad, (- FLAGS.out_grad_clip), FLAGS.out_grad_clip), var) for (grad, var) in gv_list] self.metatrain_op = optimizer.apply_gradients(gv_list) else: self.eval_train_loss = [tf.reduce_mean(full_step_train_loss[j]) for j in range((max_update_step + 1))] self.eval_valid_loss = [tf.reduce_mean(full_step_valid_loss[j]) for j in range((max_update_step + 1))] self.eval_train_z_list = full_step_train_pred self.eval_valid_z_list = full_step_valid_pred print('end of model construction')
def is_trained(cfg: NamespaceMap, stage: str, is_force_retrain: bool=False) -> bool: pretrained_path = Path(cfg.paths.pretrained.save) filename = BEST_CHECKPOINT.format(stage=stage) if (is_force_retrain and (pretrained_path / filename).is_file()): results_path = Path(cfg.paths.results) ckpt_path = Path(cfg.checkpoint.kwargs.dirpath) log_path = Path(cfg.paths.logs) logger.info(f'Forcing the retraining of {stage}, even though {(pretrained_path / filename)} exists. Deleting {pretrained_path} and {results_path} and {ckpt_path} and {log_path}.') remove_rf(pretrained_path, not_exist_ok=True) remove_rf(results_path, not_exist_ok=True) remove_rf(ckpt_path, not_exist_ok=True) remove_rf(log_path, not_exist_ok=True) pretrained_path.mkdir(parents=True) results_path.mkdir(parents=True) ckpt_path.mkdir(parents=True) log_path.mkdir(parents=True) return False else: return (pretrained_path / filename).is_file()
class ORTModel(): def __init__(self, model: Union[(str, os.PathLike)], compute_metrics: Optional[Callable[([EvalPrediction], Dict)]]=None, label_names: Optional[List[str]]=None): self.compute_metrics = compute_metrics self.label_names = (['labels'] if (label_names is None) else label_names) self.session = onnxruntime.InferenceSession(model.SerializeToString(), providers=onnxruntime.get_available_providers()) self.onnx_input_names = {input_key.name: idx for (idx, input_key) in enumerate(self.session.get_inputs())} def evaluation_loop(self, dataset: Dataset): logger.info('***** Running evaluation *****') all_preds = None all_labels = None onnx_inputs = {} for (step, inputs) in enumerate(dataset): has_labels = all(((inputs.get(k) is not None) for k in self.label_names)) if has_labels: labels = tuple((np.array([inputs.get(name)]) for name in self.label_names)) if (len(labels) == 1): labels = labels[0] else: labels = None "\n LayoutLMV2 inputs (with order):\n {\n 'input_ids': {0: 'batch_size', 1: 'sequence_length'}, \n 'bbox': {0: 'batch_size', 1: 'sequence_length'}, \n 'image': {0: 'batch_size', 1: 'num_channels'}, # dtype is np.int64 not float\n 'attention_mask': {0: 'batch_size', 1: 'sequence_length'}, \n }\n " for key in self.onnx_input_names: if (key in inputs): onnx_inputs[key] = np.array([inputs[key]]) elif (key == 'image'): onnx_inputs[key] = np.array([inputs['images']], dtype=np.int64) preds = self.session.run(None, onnx_inputs) if (len(preds) == 1): preds = preds[0] all_preds = (preds if (all_preds is None) else nested_concat(all_preds, preds, padding_index=(- 100))) all_labels = (labels if (all_labels is None) else nested_concat(all_labels, labels, padding_index=(- 100))) if ((self.compute_metrics is not None) and (all_preds is not None) and (all_labels is not None)): metrics = self.compute_metrics(EvalPrediction(predictions=all_preds, label_ids=all_labels)) else: metrics = {} return EvalLoopOutput(predictions=all_preds, label_ids=all_labels, metrics=metrics, num_samples=len(dataset))
class ResLayer(Sequential): def __init__(self, block, inplanes, planes, num_blocks, stride=1, dilation=1, avg_down=False, conv_cfg=None, norm_cfg=dict(type='BN'), multi_grid=None, contract_dilation=False, **kwargs): self.block = block downsample = None if ((stride != 1) or (inplanes != (planes * block.expansion))): downsample = [] conv_stride = stride if avg_down: conv_stride = 1 downsample.append(nn.AvgPool2d(kernel_size=stride, stride=stride, ceil_mode=True, count_include_pad=False)) downsample.extend([build_conv_layer(conv_cfg, inplanes, (planes * block.expansion), kernel_size=1, stride=conv_stride, bias=False), build_norm_layer(norm_cfg, (planes * block.expansion))[1]]) downsample = nn.Sequential(*downsample) layers = [] if (multi_grid is None): if ((dilation > 1) and contract_dilation): first_dilation = (dilation // 2) else: first_dilation = dilation else: first_dilation = multi_grid[0] layers.append(block(inplanes=inplanes, planes=planes, stride=stride, dilation=first_dilation, downsample=downsample, conv_cfg=conv_cfg, norm_cfg=norm_cfg, **kwargs)) inplanes = (planes * block.expansion) for i in range(1, num_blocks): layers.append(block(inplanes=inplanes, planes=planes, stride=1, dilation=(dilation if (multi_grid is None) else multi_grid[i]), conv_cfg=conv_cfg, norm_cfg=norm_cfg, **kwargs)) super(ResLayer, self).__init__(*layers)
def save_blip_diffusion_model(model, args): qformer = get_qformer(model) qformer.eval() text_encoder = ContextCLIPTextModel.from_pretrained('runwayml/stable-diffusion-v1-5', subfolder='text_encoder') vae = AutoencoderKL.from_pretrained('runwayml/stable-diffusion-v1-5', subfolder='vae') unet = UNet2DConditionModel.from_pretrained('runwayml/stable-diffusion-v1-5', subfolder='unet') vae.eval() text_encoder.eval() scheduler = PNDMScheduler(beta_start=0.00085, beta_end=0.012, beta_schedule='scaled_linear', set_alpha_to_one=False, skip_prk_steps=True) tokenizer = CLIPTokenizer.from_pretrained('runwayml/stable-diffusion-v1-5', subfolder='tokenizer') image_processor = BlipImageProcessor() blip_diffusion = BlipDiffusionPipeline(tokenizer=tokenizer, text_encoder=text_encoder, vae=vae, unet=unet, scheduler=scheduler, qformer=qformer, image_processor=image_processor) blip_diffusion.save_pretrained(args.checkpoint_path)
class Net(nn.Module): def __init__(self): super().__init__() self.conv1 = nn.Conv2d(3, 6, 5) self.pool = nn.MaxPool2d(2, 2) self.conv2 = nn.Conv2d(6, 16, 5) self.fc1 = nn.Linear(((16 * 5) * 5), 120) self.fc2 = nn.Linear(120, 84) self.fc3 = nn.Linear(84, 10) def forward(self, x): x = self.pool(F.relu(self.conv1(x))) x = self.pool(F.relu(self.conv2(x))) x = torch.flatten(x, 1) x = F.relu(self.fc1(x)) x = F.relu(self.fc2(x)) x = self.fc3(x) return x
def zscore_from_pval(pval, one_minus_pval=None, distrib='norm'): if (distrib == 'norm'): zscore = norm.isf(pval) if (one_minus_pval is not None): ind = (pval > 0.5) zscore[ind] = norm.ppf(one_minus_pval[ind]) zscore = _replace_infinity(zscore, replace_val=40, method='plus-one') return zscore
class Cell(nn.Module): def __init__(self, genotype, C_prev_prev, C_prev, C, reduction, reduction_prev, SE=False): super(Cell, self).__init__() print(C_prev_prev, C_prev, C) self.se_layer = None if reduction_prev: self.preprocess0 = FactorizedReduce(C_prev_prev, C) else: self.preprocess0 = ReLUConvBN(C_prev_prev, C, 1, 1, 0) self.preprocess1 = ReLUConvBN(C_prev, C, 1, 1, 0) if reduction: (op_names, indices) = zip(*genotype.reduce) concat = genotype.reduce_concat else: (op_names, indices) = zip(*genotype.normal) concat = genotype.normal_concat self._compile(C, op_names, indices, concat, reduction) if SE: self.se_layer = SELayer(channel=(self.multiplier * C)) def _compile(self, C, op_names, indices, concat, reduction): assert (len(op_names) == len(indices)) self._steps = (len(op_names) // 2) self._concat = concat self.multiplier = len(concat) self._ops = nn.ModuleList() for (name, index) in zip(op_names, indices): stride = (2 if (reduction and (index < 2)) else 1) op = OPS[name](C, stride, True) self._ops += [op] self._indices = indices def forward(self, s0, s1, drop_prob): s0 = self.preprocess0(s0) s1 = self.preprocess1(s1) states = [s0, s1] for i in range(self._steps): h1 = states[self._indices[(2 * i)]] h2 = states[self._indices[((2 * i) + 1)]] op1 = self._ops[(2 * i)] op2 = self._ops[((2 * i) + 1)] h1 = op1(h1) h2 = op2(h2) if (self.training and (drop_prob > 0.0)): if (not isinstance(op1, Identity)): h1 = drop_path(h1, drop_prob) if (not isinstance(op2, Identity)): h2 = drop_path(h2, drop_prob) s = (h1 + h2) states += [s] if (self.se_layer is None): return torch.cat([states[i] for i in self._concat], dim=1) else: return self.se_layer(torch.cat([states[i] for i in self._concat], dim=1))
def dump_topset(topset: Dict[(str, OpConfig)], path: PathLike): OmegaConf.save({'topset': topset}, path)
_builder('msvd_caption') class MSVDCapBuilder(BaseDatasetBuilder): train_dataset_cls = VideoCaptionDataset eval_dataset_cls = VideoCaptionEvalDataset DATASET_CONFIG_DICT = {'default': 'configs/datasets/msvd/defaults_cap.yaml'}
class Triple(): def __init__(self, s, p, o): self.s = s self.o = o self.p = p
def last_zero_init(m): if isinstance(m, nn.Sequential): constant_init(m[(- 1)], val=0) m[(- 1)].inited = True else: constant_init(m, val=0) m.inited = True
def rlagru_resnet101_eca(rla_channel=32, k_size=[5, 5, 5, 7]): print('Constructing rlagru_resnet101_eca......') model = RLAgru_ResNet(RLA_Bottleneck, [3, 4, 23, 3], rla_channel=rla_channel, ECA=k_size) return model
def read_image_RGB(img_path): got_img = False if (not osp.exists(img_path)): raise IOError('{} does not exist'.format(img_path)) while (not got_img): try: img = Image.open(img_path).convert('RGB') got_img = True except IOError: print("IOError incurred when reading '{}'. Will redo. Don't worry. Just chill.".format(img_path)) pass return img
def main(args): if args.kitti_to_yolo_labels: from utils.utils import kitti_labels_to_yolo kitti_labels_to_yolo(args.kitti_to_yolo_labels) exit() cudnn.benchmark = True start_time = datetime.now() log.info(' NEW RUN ') log.info(f"Running: {' '.join(sys.argv)}") log.info('Full args list:') for arg in vars(args): log.info(f'{arg}: {getattr(args, arg)}') log.info('') results = ResultsManager(('' if (args.dataset == 'kitti') else 'Error')) init_settings(args) if args.usr: set_paths(args) for run in range(args.num_runs): net = init_net(args) for args.severity_idx in range(args.num_severities): if ('dua' in args.methods): methods.dua(args, net) if results.has_results(): timestamp_str = time.strftime('%b-%d-%Y_%H%M', time.localtime()) results.save_to_file(file_name=f'{timestamp_str}_raw_results.pkl') results.print_summary_latex() if (args.num_runs > 1): results.reset_results() log.info(f"{('>' * 50)} FINISHED RUN #{(run + 1)} {('<' * 50)}") runtime = (datetime.now() - start_time) log.info(f'Runtime so far: {timedelta_to_str(runtime)}') torch.cuda.empty_cache() del net if (args.num_runs > 1): results.print_multiple_runs_results() runtime = (datetime.now() - start_time) log.info(f'Execution finished in {timedelta_to_str(runtime)}')
def make_gsm_loss_evaluator(cfg): loss_evaluators = dict() loss_weights = dict() if ('l1_loss' in cfg.model.losses): l1_loss_evaluator = make_sll_loss_evaluator(cfg) loss_evaluators['l1_loss'] = l1_loss_evaluator loss_weights['l1_loss'] = cfg.model.losses.l1_loss.weight if ('gerf_loss' in cfg.model.losses): gerf_loss_evaluator = make_gerf_loss_evaluator(cfg) loss_evaluators['gerf_loss'] = gerf_loss_evaluator loss_weights['gerf_loss'] = cfg.model.losses.gerf_loss.weight if ('focal_loss' in cfg.model.losses): focal_loss_evaluator = make_focal_loss_evaluator(cfg) loss_evaluators['focal_loss'] = focal_loss_evaluator loss_weights['focal_loss'] = cfg.model.losses.focal_loss.weight if ('relative_loss' in cfg.model.losses): relative_loss_evaluators = make_relative_loss_evaluator(cfg) loss_evaluators['relative_loss'] = relative_loss_evaluators loss_weights['relative_loss'] = cfg.model.losses.relative_loss.weight return CombinedLossEvaluators(cfg, loss_evaluators, loss_weights)
class TrainingInterface(): def __init__(self, device, model, parallel, log_path_mng, data_loaders, summary_writers, opt_scheduler, param_scheduler, n_epoch, **kwargs): self.model = model self.model.device = device if parallel: self.model = nn.DataParallel(self.model) self.model.to(device) self.path_mng = log_path_mng self.summary_writers = summary_writers self.data_loaders = data_loaders self.opt_scheduler = opt_scheduler self.param_scheduler = param_scheduler self.device = device self.n_epoch = n_epoch self.epoch = 0 self.train_step = 0 self.val_step = 0 self.parallel = parallel for (key, val) in kwargs.items(): setattr(self, key, val) def name(self): if self.parallel: return self.model.module.name else: return self.model.name def log_path(self): return self.path_mng.log_path def model_path(self): return self.path_mng.model_path def writer_path(self): return self.path_mng.writer_path def writer_names(self): return self.summary_writers.writer_names def _init_loss_dic(self): loss_dic = {} for key in self.writer_names: loss_dic[key] = 0.0 return loss_dic def _accumulate_loss_dic(self, loss_dic, loss_items): assert (len(self.writer_names) == len(loss_items)) for (key, val) in zip(self.writer_names, loss_items): loss_dic[key] += val.item() return loss_dic def _write_loss_to_dic(self, loss_items): loss_dic = {} assert (len(self.writer_names) == len(loss_items)) for (key, val) in zip(self.writer_names, loss_items): loss_dic[key] = val.item() return loss_dic def _batch_to_inputs(self, batch): raise NotImplementedError def train(self, **kwargs): self.model.train() self.param_scheduler.train() epoch_loss_dic = self._init_loss_dic() for (i, batch) in enumerate(self.data_loaders.train_loader): inputs = self._batch_to_inputs(batch) self.opt_scheduler.optimizer_zero_grad() input_params = self.param_scheduler.step() outputs = self.model('train', *inputs, **input_params) outputs = self._sum_parallel_loss(outputs) loss = outputs[0] loss.backward() torch.nn.utils.clip_grad_norm_(self.model.parameters(), self.opt_scheduler.clip) self.opt_scheduler.step() self._accumulate_loss_dic(epoch_loss_dic, outputs) batch_loss_dic = self._write_loss_to_dic(outputs) self.summary_writers.write_task('train', batch_loss_dic, self.train_step) self.train_step += 1 return epoch_loss_dic def _sum_parallel_loss(self, loss): if self.parallel: if isinstance(loss, tuple): return tuple([x.mean() for x in loss]) else: return loss.mean() else: return loss def eval(self): self.model.eval() self.param_scheduler.eval() epoch_loss_dic = self._init_loss_dic() for (i, batch) in enumerate(self.data_loaders.val_loader): inputs = self._batch_to_inputs(batch) input_params = self.param_scheduler.step() with torch.no_grad(): outputs = self.model('train', *inputs, **input_params) outputs = self._sum_parallel_loss(outputs) self._accumulate_loss_dic(epoch_loss_dic, outputs) batch_loss_dic = self._write_loss_to_dic(outputs) self.summary_writers.write_task('val', batch_loss_dic, self.val_step) self.val_step += 1 return epoch_loss_dic def save_model(self, fn): if self.parallel: torch.save(self.model.module.state_dict(), fn) else: torch.save(self.model.state_dict(), fn) def epoch_report(self, start_time, end_time, train_loss, valid_loss): (epoch_mins, epoch_secs) = epoch_time(start_time, end_time) print(f'Epoch: {(self.epoch + 1):02} | Time: {epoch_mins}m {epoch_secs}s', flush=True) print(f' Train Loss: {train_loss:.3f}', flush=True) print(f' Valid. Loss: {valid_loss:.3f}', flush=True) def run(self, start_epoch=0, start_train_step=0, start_val_step=0): self.epoch = start_epoch self.train_step = start_train_step self.val_step = start_val_step best_valid_loss = float('inf') for i in range(self.n_epoch): start_time = time.time() train_loss = self.train()['loss'] val_loss = self.eval()['loss'] end_time = time.time() self.save_model(self.path_mng.epoch_model_path(self.name)) if (val_loss < best_valid_loss): best_valid_loss = val_loss self.save_model(self.path_mng.valid_model_path(self.name)) self.epoch_report(start_time, end_time, train_loss, val_loss) self.epoch += 1 self.save_model(self.path_mng.final_model_path(self.name)) print('Model saved.')
class DenseProjection(nn.Module): def __init__(self, in_channels, nr, scale, up=True, bottleneck=True): super(DenseProjection, self).__init__() if bottleneck: self.bottleneck = nn.Sequential(*[nn.Conv2d(in_channels, nr, 1), nn.PReLU(nr)]) inter_channels = nr else: self.bottleneck = None inter_channels = in_channels self.conv_1 = nn.Sequential(*[projection_conv(inter_channels, nr, scale, up), nn.PReLU(nr)]) self.conv_2 = nn.Sequential(*[projection_conv(nr, inter_channels, scale, (not up)), nn.PReLU(inter_channels)]) self.conv_3 = nn.Sequential(*[projection_conv(inter_channels, nr, scale, up), nn.PReLU(nr)]) def forward(self, x): if (self.bottleneck is not None): x = self.bottleneck(x) a_0 = self.conv_1(x) b_0 = self.conv_2(a_0) e = b_0.sub(x) a_1 = self.conv_3(e) out = a_0.add(a_1) return out
def main(): parser = argparse.ArgumentParser(description='Supplement dataset') parser.add_argument('--data', type=str, required=True, help='Block data file to use (e.g. inputs/data/time_skylake.data') parser.add_argument('--embedding', type=str, required=True, help='Token embedding file to use (e.g. inputs/embeddings/code_delim.emb)') parser.add_argument('--table-name', type=str, required=True, help='Table to write augmentations to (will be freshly created)') parser.add_argument('--execute-sql', action='store_true', default=False) parser.add_argument('--store-sql', action='store_true', default=False) parser.add_argument('--optimize-sql', action='store_true', default=False) subparsers = parser.add_subparsers(dest='command') perms_parser = subparsers.add_parser('permutations') perms_parser.add_argument('--desired-n-perms', default='all') perms_parser.add_argument('--max-block-size', type=int, default=None, help='Maximum block size to attempt to generate permutations for. Default none') perms_parser.add_argument('--min-perms-per-block', type=int, default=None, help='Minimum number of permutations to include when generating permutations (otherwise throw out block)') perms_parser.add_argument('--max-perms-per-block', type=int, default=None, help='Maximum numnber of permutations to include when generating permuations.') ports_parser = subparsers.add_parser('ports') ports_parser.add_argument('--dup-template', type=str, default=_DEFAULT_DUP_TEMPLATE) ports_parser.add_argument('--max-dups', type=int, default=10, help='Max number of times to duplicate a given instruction') args = parser.parse_args() data = read_dataset(args.data, args.embedding) if (args.command == 'permutations'): if (args.desired_n_perms == 'all'): desired_n_perms = None elif (args.desired_n_perms == 'equal'): desired_n_perms = len(data.data) else: desired_n_perms = int(args.desired_n_perms) augs = gen_permutations(data, desired_n_perms=desired_n_perms, max_block_size=args.max_block_size, min_perms_per_block=args.min_perms_per_block, max_perms_per_block=args.max_perms_per_block) else: augs = gen_duplicated_instructions(data, args.max_dups) sql_commands = gen_sql_commands_of_augs(augs, args.table_name) if args.optimize_sql: sql_commands.insert(0, 'SET autocommit=0;') sql_commands.insert(1, 'SET unique_checks=0;') sql_commands.insert(2, 'SET foreign_key_checks=0;') sql_commands.append('COMMIT;') sql_commands.append('SET unique_checks=1;') sql_commands.append('SET foreign_key_checks=1;') sql_commands.append('SET autocommit=1;') if args.store_sql: with open(os.path.join(_DATA_DIR, 'table_{}.sql'.format(time_str())), 'w') as f: print('\n'.join(sql_commands), file=f) if args.execute_sql: execute_sql(sql_commands)
class BaseNetwork(nn.Module): def __init__(self): super(BaseNetwork, self).__init__() def modify_commandline_options(parser, is_train): return parser def print_network(self): if isinstance(self, list): self = self[0] num_params = 0 for param in self.parameters(): num_params += param.numel() print(('Network [%s] was created. Total number of parameters: %.1f million. To see the architecture, do print(network).' % (type(self).__name__, (num_params / 1000000)))) def init_weights(self, init_type='normal', gain=0.02): def init_func(m): classname = m.__class__.__name__ if (classname.find('BatchNorm2d') != (- 1)): if (hasattr(m, 'weight') and (m.weight is not None)): init.normal_(m.weight.data, 1.0, gain) if (hasattr(m, 'bias') and (m.bias is not None)): init.constant_(m.bias.data, 0.0) elif (hasattr(m, 'weight') and ((classname.find('Conv') != (- 1)) or (classname.find('Linear') != (- 1)))): if (init_type == 'normal'): init.normal_(m.weight.data, 0.0, gain) elif (init_type == 'xavier'): init.xavier_normal_(m.weight.data, gain=gain) elif (init_type == 'xavier_uniform'): init.xavier_uniform_(m.weight.data, gain=1.0) elif (init_type == 'kaiming'): init.kaiming_normal_(m.weight.data, a=0, mode='fan_in') elif (init_type == 'orthogonal'): init.orthogonal_(m.weight.data, gain=gain) elif (init_type == 'none'): m.reset_parameters() else: raise NotImplementedError(('initialization method [%s] is not implemented' % init_type)) if (hasattr(m, 'bias') and (m.bias is not None)): init.constant_(m.bias.data, 0.0) self.apply(init_func) for m in self.children(): if hasattr(m, 'init_weights'): m.init_weights(init_type, gain)
def get_final_path(closed, goal_node): (reversed_x, reversed_y, reversed_yaw) = (list(reversed(goal_node.x_list)), list(reversed(goal_node.y_list)), list(reversed(goal_node.yaw_list))) direction = list(reversed(goal_node.directions)) nid = goal_node.parent_index final_cost = goal_node.cost while nid: n = closed[nid] reversed_x.extend(list(reversed(n.x_list))) reversed_y.extend(list(reversed(n.y_list))) reversed_yaw.extend(list(reversed(n.yaw_list))) direction.extend(list(reversed(n.directions))) nid = n.parent_index reversed_x = list(reversed(reversed_x)) reversed_y = list(reversed(reversed_y)) reversed_yaw = list(reversed(reversed_yaw)) direction = list(reversed(direction)) direction[0] = direction[1] path = Path(reversed_x, reversed_y, reversed_yaw, direction, final_cost) return path
class FasterRCNNResnetV1FeatureExtractor(faster_rcnn_meta_arch.FasterRCNNFeatureExtractor): def __init__(self, architecture, resnet_model, is_training, first_stage_features_stride, reuse_weights=None, weight_decay=0.0): if ((first_stage_features_stride != 8) and (first_stage_features_stride != 16)): raise ValueError('`first_stage_features_stride` must be 8 or 16.') self._architecture = architecture self._resnet_model = resnet_model super(FasterRCNNResnetV1FeatureExtractor, self).__init__(is_training, first_stage_features_stride, reuse_weights, weight_decay) def preprocess(self, resized_inputs): channel_means = [123.68, 116.779, 103.939] return (resized_inputs - [[channel_means]]) def _extract_proposal_features(self, preprocessed_inputs, scope): if (len(preprocessed_inputs.get_shape().as_list()) != 4): raise ValueError(('`preprocessed_inputs` must be 4 dimensional, got a tensor of shape %s' % preprocessed_inputs.get_shape())) shape_assert = tf.Assert(tf.logical_and(tf.greater_equal(tf.shape(preprocessed_inputs)[1], 33), tf.greater_equal(tf.shape(preprocessed_inputs)[2], 33)), ['image size must at least be 33 in both height and width.']) with tf.control_dependencies([shape_assert]): with slim.arg_scope(resnet_utils.resnet_arg_scope(batch_norm_epsilon=1e-05, batch_norm_scale=True, weight_decay=self._weight_decay)): with tf.variable_scope(self._architecture, reuse=self._reuse_weights) as var_scope: (_, activations) = self._resnet_model(preprocessed_inputs, num_classes=None, is_training=False, global_pool=False, output_stride=self._first_stage_features_stride, spatial_squeeze=False, scope=var_scope) handle = (scope + ('/%s/block3' % self._architecture)) return activations[handle] def _extract_box_classifier_features(self, proposal_feature_maps, scope): with tf.variable_scope(self._architecture, reuse=self._reuse_weights): with slim.arg_scope(resnet_utils.resnet_arg_scope(batch_norm_epsilon=1e-05, batch_norm_scale=True, weight_decay=self._weight_decay)): with slim.arg_scope([slim.batch_norm], is_training=False): blocks = [resnet_utils.Block('block4', resnet_v1.bottleneck, ([{'depth': 2048, 'depth_bottleneck': 512, 'stride': 1}] * 3))] proposal_classifier_features = resnet_utils.stack_blocks_dense(proposal_feature_maps, blocks) return proposal_classifier_features