code
stringlengths
101
5.91M
def velocity_after_collision(n: np.ndarray, velocity: np.ndarray, m: float, j: float) -> np.ndarray: return (velocity + ((j * n) / m))
def paths_to_tensors(paths, max_path_length, baseline_predictions, discount): baselines = [] returns = [] for (idx, path) in enumerate(paths): path['baselines'] = baseline_predictions[idx] baselines.append(path['baselines']) path['returns'] = tensor_utils.discount_cumsum(path['rewards'], discount) returns.append(path['returns']) obs = [path['observations'] for path in paths] obs = tensor_utils.pad_tensor_n(obs, max_path_length) actions = [path['actions'] for path in paths] actions = tensor_utils.pad_tensor_n(actions, max_path_length) rewards = [path['rewards'] for path in paths] rewards = tensor_utils.pad_tensor_n(rewards, max_path_length) agent_infos = [path['agent_infos'] for path in paths] agent_infos = tensor_utils.stack_tensor_dict_list([tensor_utils.pad_tensor_dict(p, max_path_length) for p in agent_infos]) env_infos = [path['env_infos'] for path in paths] env_infos = tensor_utils.stack_tensor_dict_list([tensor_utils.pad_tensor_dict(p, max_path_length) for p in env_infos]) valids = [np.ones_like(path['returns']) for path in paths] valids = tensor_utils.pad_tensor_n(valids, max_path_length) samples_data = dict(observations=obs, actions=actions, rewards=rewards, agent_infos=agent_infos, env_infos=env_infos, valids=valids) return samples_data
def benchmark_shortest_path_image(configuration_space, source): def shortest_path_image(): graph = shortest_paths.GridGraph(configuration_space) graph.shortest_path_image(source) benchmark(shortest_path_image)
def FindEndOfExpressionInLine(line, startpos, depth, startchar, endchar): for i in xrange(startpos, len(line)): if (line[i] == startchar): depth += 1 elif (line[i] == endchar): depth -= 1 if (depth == 0): return ((i + 1), 0) return ((- 1), depth)
class FlaxBigBirdForMultipleChoice(metaclass=DummyObject): _backends = ['flax'] def __init__(self, *args, **kwargs): requires_backends(self, ['flax'])
class AssignScoreWithK(Function): def forward(ctx, scores, points, centers, knn_idx, aggregate): agg = {'sum': 0, 'avg': 1, 'max': 2} (B, N, M, O) = points.size() K = scores.size(2) output = torch.zeros([B, O, N], dtype=points.dtype, device=points.device) output = output.contiguous() src.gpu.assign_score_withk_forward_cuda(B, N, M, K, O, agg[aggregate], points.contiguous(), centers.contiguous(), scores.contiguous(), knn_idx.contiguous(), output) ctx.save_for_backward(output, points, centers, scores, knn_idx) ctx.agg = agg[aggregate] return output def backward(ctx, grad_out): (output, points, centers, scores, knn_idx) = ctx.saved_tensors agg = ctx.agg (B, N, M, O) = points.size() K = scores.size(2) grad_points = torch.zeros_like(points, dtype=points.dtype, device=points.device).contiguous() grad_centers = torch.zeros_like(centers, dtype=points.dtype, device=points.device).contiguous() grad_scores = torch.zeros_like(scores, dtype=scores.dtype, device=scores.device).contiguous() src.gpu.assign_score_withk_backward_cuda(B, N, M, K, O, agg, grad_out.contiguous(), points.contiguous(), centers.contiguous(), scores.contiguous(), knn_idx.contiguous(), grad_points, grad_centers, grad_scores) return (grad_scores, grad_points, grad_centers, None, None, None)
def list_non_bool_dtypes(): return [o3c.float32, o3c.float64, o3c.int8, o3c.int16, o3c.int32, o3c.int64, o3c.uint8, o3c.uint16, o3c.uint32, o3c.uint64]
def get_hbond_donor_indice(m): smarts = ['[!#6;!H0]'] indice = [] for s in smarts: s = Chem.MolFromSmarts(s) indice += [i[0] for i in m.GetSubstructMatches(s)] indice = np.array(indice) return indice
class FeatureFusionModule(nn.Module): def __init__(self, in_chan, out_chan, *args, **kwargs): super(FeatureFusionModule, self).__init__() self.convblk = ConvBNReLU(in_chan, out_chan, ks=1, stride=1, padding=0) self.conv1 = nn.Conv2d(out_chan, (out_chan // 4), kernel_size=1, stride=1, padding=0, bias=False) self.conv2 = nn.Conv2d((out_chan // 4), out_chan, kernel_size=1, stride=1, padding=0, bias=False) self.relu = nn.ReLU(inplace=True) self.sigmoid = nn.Sigmoid() self.init_weight() def forward(self, fsp, fcp): fcat = torch.cat([fsp, fcp], dim=1) feat = self.convblk(fcat) atten = F.avg_pool2d(feat, feat.size()[2:]) atten = self.conv1(atten) atten = self.relu(atten) atten = self.conv2(atten) atten = self.sigmoid(atten) feat_atten = torch.mul(feat, atten) feat_out = (feat_atten + feat) return feat_out def init_weight(self): for ly in self.children(): if isinstance(ly, nn.Conv2d): nn.init.kaiming_normal_(ly.weight, a=1) if (not (ly.bias is None)): nn.init.constant_(ly.bias, 0) def get_params(self): (wd_params, nowd_params) = ([], []) for (name, module) in self.named_modules(): if (isinstance(module, nn.Linear) or isinstance(module, nn.Conv2d)): wd_params.append(module.weight) if (not (module.bias is None)): nowd_params.append(module.bias) elif isinstance(module, nn.BatchNorm2d): nowd_params += list(module.parameters()) return (wd_params, nowd_params)
class MetaFeature(AbstractMetaFeature): def __init__(self): super(MetaFeature, self).__init__() self.type_ = 'METAFEATURE'
def makeplot_bar(experiments, planner, ttl): if (plt_cfg['qt'] == 1): N = len(experiments) x = np.linspace(1, N, N, endpoint=True) plt.figure(ttl) plt.xlabel('Test run') if ('dur' in ttl): plt.ylabel('Execution Time in [s]') else: plt.ylabel('Path Length in [m]') plt.bar(x, experiments, color=planner_stl[planner]) plt.axhline(y=np.mean(experiments), color=planner_stl[planner], linewidth=2) else: data_dur = {} data_len = {} for key in experiments: if ('dur' in key): data_dur[key] = experiments[key] elif ('len' in key): data_len[key] = experiments[key] (fig, ax) = plt.subplots() ax.set_title(ttl) bar_plot(ax, data_dur, total_width=0.8, single_width=0.9) (fig, ax) = plt.subplots() ax.set_title(ttl) bar_plot(ax, data_len, total_width=0.8, single_width=0.9) plt.savefig((('quanti/' + ttl) + '.png'), bbox_inches='tight')
class AutoregressiveRationalQuadraticSplineBijection(RationalQuadraticSplineBijection): def __init__(self, num_input_channels, num_hidden_layers, num_hidden_channels, num_bins, tail_bound, activation, dropout_probability): super().__init__(num_input_channels=num_input_channels, flow=MaskedPiecewiseRationalQuadraticAutoregressiveTransform(features=num_input_channels, hidden_features=num_hidden_channels, context_features=None, num_bins=num_bins, tails='linear', tail_bound=tail_bound, num_blocks=num_hidden_layers, use_residual_blocks=True, random_mask=False, activation=activation(), dropout_probability=dropout_probability, use_batch_norm=False))
def rearrange_tf_to_pt(value, depthwise=False): if (value.ndim == 4): if depthwise: return einops.rearrange(value, 'h w c_in c_out -> c_in c_out h w') else: return einops.rearrange(value, 'h w c_in c_out -> c_out c_in h w') elif (value.ndim == 2): return einops.rearrange(value, 'c_in c_out -> c_out c_in') elif (value.ndim == 1): return value
class HPText(): dataset = os.path.join(os.path.dirname(os.path.abspath(__file__)), 'datasets/data/LJSpeech-1.1') (num_train, num_valid) = (13000, 13099) punctuation = list('\'",.:?!') graphemes = ((['<pad>', '<unk>'] + list('abcdefghijklmnopqrstuvwxyz ')) + punctuation) use_phonemes = True
def test_construct_arguments_without_duplicates_passes(): s = Signature(bariza) s.construct_arguments([1, 2], {'c': 5}, {}) s = Signature(complex_function_name) s.construct_arguments([1], {'b': 4}, {}) s = Signature(FunCTIonWithCAPItals) s.construct_arguments([], {'a': 6, 'b': 6, 'c': 6}, {})
def all_metrics(preds, labels): acc = simple_accuracy(preds, labels) f1 = f1_score(y_true=labels, y_pred=preds) pre = precision_score(y_true=labels, y_pred=preds) rec = recall_score(y_true=labels, y_pred=preds) return {'acc': acc, 'precision': pre, 'recall': rec, 'f1': f1}
def apiurl(args: argparse.Namespace, subpath: str, query_args: typing.Dict=None) -> str: if (query_args is None): query_args = {} if (args.api_key is not None): query_args['api_key'] = args.api_key if query_args: return (((args.url + subpath) + '?') + '&'.join(('{x}={y}'.format(x=x, y=quote_plus((y if isinstance(y, str) else json.dumps(y)))) for (x, y) in query_args.items()))) else: return (args.url + subpath)
class DRN_A(nn.Module): def __init__(self, block, layers, BatchNorm=None): self.inplanes = 64 super(DRN_A, self).__init__() self.out_dim = (512 * block.expansion) self.conv1 = nn.Conv2d(3, 64, kernel_size=7, stride=2, padding=3, bias=False) self.bn1 = BatchNorm(64) self.relu = nn.ReLU(inplace=True) self.maxpool = nn.MaxPool2d(kernel_size=3, stride=2, padding=1) self.layer1 = self._make_layer(block, 64, layers[0], BatchNorm=BatchNorm) self.layer2 = self._make_layer(block, 128, layers[1], stride=2, BatchNorm=BatchNorm) self.layer3 = self._make_layer(block, 256, layers[2], stride=1, dilation=2, BatchNorm=BatchNorm) self.layer4 = self._make_layer(block, 512, layers[3], stride=1, dilation=4, BatchNorm=BatchNorm) self._init_weight() def _init_weight(self): for m in self.modules(): if isinstance(m, nn.Conv2d): n = ((m.kernel_size[0] * m.kernel_size[1]) * m.out_channels) m.weight.data.normal_(0, math.sqrt((2.0 / n))) elif isinstance(m, SynchronizedBatchNorm2d): m.weight.data.fill_(1) m.bias.data.zero_() elif isinstance(m, nn.BatchNorm2d): m.weight.data.fill_(1) m.bias.data.zero_() def _make_layer(self, block, planes, blocks, stride=1, dilation=1, BatchNorm=None): downsample = None if ((stride != 1) or (self.inplanes != (planes * block.expansion))): downsample = nn.Sequential(nn.Conv2d(self.inplanes, (planes * block.expansion), kernel_size=1, stride=stride, bias=False), BatchNorm((planes * block.expansion))) layers = [] layers.append(block(self.inplanes, planes, stride, downsample, BatchNorm=BatchNorm)) self.inplanes = (planes * block.expansion) for i in range(1, blocks): layers.append(block(self.inplanes, planes, dilation=(dilation, dilation), BatchNorm=BatchNorm)) return nn.Sequential(*layers) def forward(self, x): x = self.conv1(x) x = self.bn1(x) x = self.relu(x) x = self.maxpool(x) x = self.layer1(x) x = self.layer2(x) x = self.layer3(x) x = self.layer4(x) return x
def convert(item): (qid, example) = item q = {'image_id': example['imageId'], 'question_id': qid, 'question': example['question']} if ('answer' in example): a = {'image_id': example['imageId'], 'question_id': qid, 'answers': [{'answer': example['answer']}]} else: a = None return (q, a)
class Xception_dilation(nn.Module): def __init__(self, input_channel=None, num_classes=None): super(Xception_dilation, self).__init__() self.num_classes = num_classes self.conv1 = nn.Conv2d(input_channel, 32, 3, 2, 0, bias=False) self.bn1 = nn.BatchNorm2d(32) self.relu1 = nn.ReLU(inplace=True) self.conv2 = nn.Conv2d(32, 64, 3, bias=False) self.bn2 = nn.BatchNorm2d(64) self.relu2 = nn.ReLU(inplace=True) self.block1 = Block(64, 128, 2, 2, start_with_relu=False, grow_first=True) self.block2 = Block(128, 256, 2, 2, start_with_relu=True, grow_first=True) self.block3 = Block(256, 728, 2, 2, start_with_relu=True, grow_first=True) self.block4 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block5 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block6 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block7 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block8 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block9 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block10 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.block11 = Block(728, 728, 3, 1, start_with_relu=True, grow_first=True) self.residual = nn.Sequential(nn.Conv2d(728, 1024, 1, 1, dilation=2, bias=False), nn.BatchNorm2d(1024)) self.SepConv1 = nn.Sequential(nn.ReLU(inplace=False), SeparableConv2d(728, 728, 3, stride=1, padding=1, bias=False), nn.BatchNorm2d(728)) self.SepConv2 = nn.Sequential(nn.ReLU(inplace=False), SeparableConv2d(728, 1024, 3, stride=1, padding=1, bias=False), nn.BatchNorm2d(1024)) self.SepConv3 = nn.Sequential(SeparableConv2d(1024, 1536, 3, dilation=2, stride=1, padding=2, bias=False), nn.BatchNorm2d(1536), nn.ReLU(inplace=False)) self.SepConv4 = nn.Sequential(SeparableConv2d(1536, 2048, 3, dilation=2, stride=1, padding=2, bias=False), nn.BatchNorm2d(2048), nn.ReLU(inplace=False)) self.avgpool = nn.AdaptiveAvgPool2d(1) self.cls = nn.Linear(2048, num_classes) def get_layers(self): return self.layers def forward(self, x): self.layers = [] x = self.conv1(x) x = self.bn1(x) x = self.relu1(x) x = self.conv2(x) x = self.bn2(x) x = self.relu2(x) x = self.block1(x) x = self.block2(x) x = self.block3(x) x = self.block4(x) x = self.block5(x) x = self.block6(x) x = self.block7(x) x = self.block8(x) x = self.block9(x) x = self.block10(x) x = self.block11(x) res = self.residual(x) x = self.SepConv1(x) x = self.SepConv2(x) x += res x = self.SepConv3(x) x = self.SepConv4(x) self.layers.append(x) x = self.avgpool(x) x = x.view(x.size(0), (- 1)) x = self.cls(x) self.layers.append(x) return x
class Pinwheel(VAE): def __init__(self, params): assert (params.latent_dim == 2) super(Pinwheel, self).__init__(NormalMixture, dist.Normal, dist.Normal, Enc(params.latent_dim, params.num_hidden_layers, params.hidden_dim), Dec(params.latent_dim, params.num_hidden_layers, params.hidden_dim), params) self.modelName = 'pinwheel' def init_pz(self, o): pz_mu = nn.Parameter(torch.tensor([[0.0, (- 1.0)], [0.0, 1.0], [1.0, 0.0], [(- 1.0), 0.0]]), requires_grad=False) pz_logvar = nn.Parameter(torch.zeros(4, o.latent_dim), requires_grad=o.learn_prior_variance) return (pz_mu, pz_logvar) def getDataLoaders(batch_size, shuffle=True, device='cuda'): kwargs = ({'num_workers': 1, 'pin_memory': True, 'drop_last': True} if (device == 'cuda') else {'drop_last': True}) variant = 'small' train_loader = DataLoader(PinwheelDataset('../data', train=True, variant=variant), batch_size=batch_size, shuffle=shuffle, **kwargs) test_loader = DataLoader(PinwheelDataset('../data', train=False, variant=variant), batch_size=batch_size, shuffle=shuffle, **kwargs) return (train_loader, test_loader) def generate(self, runPath, epoch): (N, K) = (1000, 100) (mean, means, samples) = super(Pinwheel, self).generate(N, K) scatter_plot(mean.data.cpu(), '{}/gen_mean_{:03d}.png'.format(runPath, epoch)) scatter_plot(means.data.cpu(), '{}/gen_means_{:03d}.png'.format(runPath, epoch)) scatter_plot(samples.data.cpu(), '{}/gen_samples_{:03d}.png'.format(runPath, epoch)) def reconstruct(self, data, runPath, epoch): n = min(400, data.size(0)) recon = super(Pinwheel, self).reconstruct(data[:n]) scatter_plot([data[:n].data.cpu(), recon.data.cpu()], '{}/recon_{:03d}.png'.format(runPath, epoch)) def posterior_plot(self, qz_x_mean, qz_x_std, runPath, epoch): posterior_plot_pinwheel(qz_x_mean, qz_x_std, '{}/posterior_{:03d}.png'.format(runPath, epoch))
def np_F1(y_true, y_pred): score = [] for (yy_true, yy_pred) in zip(y_true, y_pred): this = f1_score((yy_true > 0.5).astype('int').ravel(), (yy_pred > 0.5).astype('int').ravel()) that = f1_score((yy_true > 0.5).astype('int').ravel(), ((1 - yy_pred) > 0.5).astype('int').ravel()) score.append(max(this, that)) return np.mean(score).astype('float32')
def indice_maxpool(features, indice_pairs, indice_pair_num, num_activate_out): if (features.dtype == torch.float32): return sparse_conv_ext.indice_maxpool_fp32(features, indice_pairs, indice_pair_num, num_activate_out) elif (features.dtype == torch.half): return sparse_conv_ext.indice_maxpool_half(features, indice_pairs, indice_pair_num, num_activate_out) else: raise NotImplementedError
def int2bitstr(integer): four_bytes = struct.pack('>I', integer) return ''.join((f'{byte:08b}' for byte in four_bytes))
class MLPComponentTest(BaseRegressionComponentTest): __test__ = True res = dict() res['default_boston'] = 0. res['default_boston_places'] = 4 res['boston_n_calls'] = 8 res['boston_iterative_n_iter'] = 161 res['default_boston_iterative'] = res['default_boston'] res['default_boston_iterative_places'] = 1 res['default_boston_sparse'] = (- 0.) res['default_boston_sparse_places'] = 6 res['default_boston_iterative_sparse'] = res['default_boston_sparse'] res['default_boston_iterative_sparse_places'] = 6 res['default_diabetes'] = 0. res['diabetes_n_calls'] = 9 res['diabetes_iterative_n_iter'] = 106 res['default_diabetes_iterative'] = res['default_diabetes'] res['default_diabetes_sparse'] = 0. res['default_diabetes_iterative_sparse'] = res['default_diabetes_sparse'] sk_mod = sklearn.neural_network.MLPRegressor module = MLPRegressor step_hyperparameter = {'name': 'n_iter_', 'value': module.get_max_iter()}
class DukeMTMC(Dataset): url = ' md5 = '2f93496f9b516d1ee5ef51c1d5e7d601' def __init__(self, root, split_id=0, num_val=100, download=True): super(DukeMTMC, self).__init__(root, split_id=split_id) if download: self.download() if (not self._check_integrity()): raise RuntimeError(('Dataset not found or corrupted. ' + 'You can use download=True to download it.')) self.load(num_val) def download(self): if self._check_integrity(): print('Files already downloaded and verified') return import re import hashlib import shutil from glob import glob from zipfile import ZipFile raw_dir = osp.join(self.root, 'raw') mkdir_if_missing(raw_dir) fpath = osp.join(raw_dir, 'DukeMTMC-reID.zip') if (osp.isfile(fpath) and (hashlib.md5(open(fpath, 'rb').read()).hexdigest() == self.md5)): print(('Using downloaded file: ' + fpath)) else: raise RuntimeError('Please download the dataset manually from {} to {}'.format(self.url, fpath)) exdir = osp.join(raw_dir, 'DukeMTMC-reID') if (not osp.isdir(exdir)): print('Extracting zip file') with ZipFile(fpath) as z: z.extractall(path=raw_dir) images_dir = osp.join(self.root, 'images') mkdir_if_missing(images_dir) identities = [] all_pids = {} def register(subdir, pattern=re.compile('([-\\d]+)_c(\\d)')): fpaths = sorted(glob(osp.join(exdir, subdir, '*.jpg'))) pids = set() for fpath in fpaths: fname = osp.basename(fpath) (pid, cam) = map(int, pattern.search(fname).groups()) assert (1 <= cam <= 8) cam -= 1 if (pid not in all_pids): all_pids[pid] = len(all_pids) pid = all_pids[pid] pids.add(pid) if (pid >= len(identities)): assert (pid == len(identities)) identities.append([[] for _ in range(8)]) fname = '{:08d}_{:02d}_{:04d}.jpg'.format(pid, cam, len(identities[pid][cam])) identities[pid][cam].append(fname) shutil.copy(fpath, osp.join(images_dir, fname)) return pids trainval_pids = register('bounding_box_train') gallery_pids = register('bounding_box_test') query_pids = register('query') assert (query_pids <= gallery_pids) assert trainval_pids.isdisjoint(gallery_pids) meta = {'name': 'DukeMTMC', 'shot': 'multiple', 'num_cameras': 8, 'identities': identities} write_json(meta, osp.join(self.root, 'meta.json')) splits = [{'trainval': sorted(list(trainval_pids)), 'query': sorted(list(query_pids)), 'gallery': sorted(list(gallery_pids))}] write_json(splits, osp.join(self.root, 'splits.json'))
def generate_cross_cols(self, df: pd.DataFrame, crossed_cols): df_cc = df.copy() crossed_colnames = [] for cols in crossed_cols: for c in cols: df_cc[c] = df_cc[c].astype('str') colname = '_'.join(cols) df_cc[colname] = df_cc[list(cols)].apply((lambda x: '-'.join(x)), axis=1) crossed_colnames.append(colname) return df_cc[crossed_colnames]
def model_opts(parser): group = parser.add_argument_group('Model-Embeddings') group.add_argument('-src_word_vec_size', type=int, default=500, help='Word embedding size for src.') group.add_argument('-tgt_word_vec_size', type=int, default=500, help='Word embedding size for tgt.') group.add_argument('-word_vec_size', type=int, default=(- 1), help='Word embedding size for src and tgt.') group.add_argument('-share_decoder_embeddings', action='store_true', help='Use a shared weight matrix for the input and\n\t\t\t\t\t output word embeddings in the decoder.') group.add_argument('-share_embeddings', action='store_true', help='Share the word embeddings between encoder\n\t\t\t\t\t and decoder. Need to use shared dictionary for this\n\t\t\t\t\t option.') group.add_argument('-position_encoding', action='store_true', help='Use a sin to mark relative words positions.\n\t\t\t\t\t Necessary for non-RNN style models.\n\t\t\t\t\t ') group.add_argument('-context_type', type=str, default=None, help='Where to consider context: [HAN_enc|HAN_dec|HAN_dec_source|HAN_dec_context|HAN_join]') group.add_argument('-context_size', type=int, default=3, help='Number of previous sentences.') group = parser.add_argument_group('Model-Embedding Features') group.add_argument('-feat_merge', type=str, default='concat', choices=['concat', 'sum', 'mlp'], help='Merge action for incorporating features embeddings.\n\t\t\t\t\t Options [concat|sum|mlp].') group.add_argument('-feat_vec_size', type=int, default=(- 1), help='If specified, feature embedding sizes\n\t\t\t\t\t will be set to this. Otherwise, feat_vec_exponent\n\t\t\t\t\t will be used.') group.add_argument('-feat_vec_exponent', type=float, default=0.7, help='If -feat_merge_size is not set, feature\n\t\t\t\t\t embedding sizes will be set to N^feat_vec_exponent\n\t\t\t\t\t where N is the number of values the feature takes.') group = parser.add_argument_group('Model- Encoder-Decoder') group.add_argument('-model_type', default='text', help='Type of source model to use. Allows\n\t\t\t\t\t the system to incorporate non-text inputs.\n\t\t\t\t\t Options are [text|img|audio].') group.add_argument('-encoder_type', type=str, default='rnn', choices=['rnn', 'brnn', 'mean', 'transformer', 'cnn'], help='Type of encoder layer to use. Non-RNN layers\n\t\t\t\t\t are experimental. Options are\n\t\t\t\t\t [rnn|brnn|mean|transformer|cnn].') group.add_argument('-decoder_type', type=str, default='rnn', choices=['rnn', 'transformer', 'cnn'], help='Type of decoder layer to use. Non-RNN layers\n\t\t\t\t\t are experimental. Options are\n\t\t\t\t\t [rnn|transformer|cnn].') group.add_argument('-layers', type=int, default=(- 1), help='Number of layers in enc/dec.') group.add_argument('-enc_layers', type=int, default=2, help='Number of layers in the encoder') group.add_argument('-dec_layers', type=int, default=2, help='Number of layers in the decoder') group.add_argument('-rnn_size', type=int, default=500, help='Size of rnn hidden states') group.add_argument('-cnn_kernel_width', type=int, default=3, help='Size of windows in the cnn, the kernel_size is\n\t\t\t\t\t (cnn_kernel_width, 1) in conv layer') group.add_argument('-input_feed', type=int, default=1, help='Feed the context vector at each time step as\n\t\t\t\t\t additional input (via concatenation with the word\n\t\t\t\t\t embeddings) to the decoder.') group.add_argument('-bridge', action='store_true', help='Have an additional layer between the last encoder\n\t\t\t\t\t state and the first decoder state') group.add_argument('-rnn_type', type=str, default='LSTM', choices=['LSTM', 'GRU', 'SRU'], action=CheckSRU, help='The gate type to use in the RNNs') group.add_argument('-brnn', action=DeprecateAction, help='Deprecated, use `encoder_type`.') group.add_argument('-brnn_merge', default='concat', choices=['concat', 'sum'], help='Merge action for the bidir hidden states') group.add_argument('-context_gate', type=str, default=None, choices=['source', 'target', 'both'], help='Type of context gate to use.\n\t\t\t\t\t Do not select for no context gate.') group = parser.add_argument_group('Model- Attention') group.add_argument('-global_attention', type=str, default='general', choices=['dot', 'general', 'mlp'], help='The attention type to use:\n\t\t\t\t\t dotprod or general (Luong) or MLP (Bahdanau)') group.add_argument('-copy_attn', action='store_true', help='Train copy attention layer.') group.add_argument('-copy_attn_force', action='store_true', help='When available, train to copy.') group.add_argument('-reuse_copy_attn', action='store_true', help='Reuse standard attention for copy') group.add_argument('-copy_loss_by_seqlength', action='store_true', help='Divide copy loss by length of sequence') group.add_argument('-coverage_attn', action='store_true', help='Train a coverage attention layer.') group.add_argument('-lambda_coverage', type=float, default=1, help='Lambda value for coverage.')
def update_context(job_args: JobArgs): for (node_type, node_args) in job_args.node_args.items(): if (node_type == NodeType.WORKER): _dlrover_context.auto_worker_enabled = node_args.auto_scale elif (node_type == NodeType.PS): _dlrover_context.auto_ps_enabled = node_args.auto_scale _dlrover_context.relaunch_always = job_args.relaunch_always _dlrover_context.set_params_from_brain() _dlrover_context.print_config()
class L2BallProj(L2Ball): def __init__(self, X, epsilon, k): DualObject.__init__(self) self.epsilon = epsilon n = X[0].numel() self.nu_x = [X] self.nu = [X.new(1, k, *X.size()[1:]).normal_()] def apply(self, dual_layer): self.nu_x.append(dual_layer(*self.nu_x)) self.nu.append(dual_layer(*self.nu)) def bounds(self, network=None): if (network is None): nu = self.nu[(- 1)] nu_x = self.nu_x[(- 1)] else: nu = network(self.nu[0]) nu_x = network(self.nu_x[0]) k = nu.size(1) l2 = (nu.norm(2, 1) / (k ** 0.5)) return ((nu_x - (self.epsilon * l2)), (nu_x + (self.epsilon * l2)))
class SmoothQuantCalibrationLLM(SmoothQuantCalibration): def __init__(self, model_path, dataloader, iterations, op_types, percentile, temp_path, weight_name_mapping): self.func = None self.graph_def = None self.frozen_func = None self._saved_model = None self.model = model_path self.dataloader = dataloader self.iterations = iterations self.op_types = op_types self.percentile = percentile self.temp_path = temp_path self.weight_name_mapping = weight_name_mapping self.print_node_list = [] self._sq_input_node_names = [] self._sq_target_node_names = {} self._sq_output_tensor_dict = {} self._sq_weight_tensor_dict = {} def _parse_calibration_logs(self, tmp_dump_file): valid_data = [] with open(tmp_dump_file) as file: for i in file.readlines(): if i.startswith(';'): valid_data.append(i.strip()) for activation in valid_data: activation = activation.split(' ') data = [] activation_name = '' per_channel = [] for (idx, s) in enumerate(activation): if (idx == 0): per_channel.append(float(s.rsplit(':')[(- 1)].strip('['))) activation_name = s.rsplit(':')[0][1:(- 9)] elif (s.find('][') != (- 1)): pairs = [float(i) for i in s.split('][')] per_channel.append(pairs[0]) data.append(per_channel) per_channel = [pairs[1]] elif (s.find(']]') != (- 1)): per_channel.append(float(s.strip(']'))) data.append(per_channel) else: per_channel.append(float(s)) if (activation_name not in self._sq_output_tensor_dict): self._sq_output_tensor_dict[activation_name] = [np.array(data)] else: self._sq_output_tensor_dict[activation_name].append(np.array(data)) def _insert_print_for_activation(self, graph_def): cur_graph = GraphAnalyzer() cur_graph.graph = graph_def graph_info = cur_graph.parse_graph() for cur_list in self.print_node_list: pre_node_name = cur_list[0] post_node_name = cur_list[(- 1)] insert_node_pairs = [] top_node = graph_info[pre_node_name].node if (top_node.op == 'ConcatV2'): for i in range(top_node.attr['N'].i): insert_node_pairs.append([top_node.input[i], post_node_name]) elif (top_node.op in ('BatchMatMul', 'BatchMatMulV2')): insert_node_pairs.append([top_node.input[0], post_node_name]) if (graph_info[top_node.input[1]].node.op != 'Const'): insert_node_pairs.append([top_node.input[1], post_node_name]) elif (top_node.op in ('Conv2DBackpropInput', 'Conv3DBackpropInputV2')): insert_node_pairs.append([top_node.input[2], post_node_name]) else: refresh_pre_node_name = graph_info[pre_node_name].node.input[0] refresh_pre_node = graph_info[Helper.node_name_from_input(refresh_pre_node_name)].node if ((refresh_pre_node.op == 'Pad') and (top_node.op in ('Conv2D', 'Conv3D'))): insert_node_pairs.append([refresh_pre_node_name, post_node_name]) refresh_pre_node_name = refresh_pre_node.input[0] insert_node_pairs.append([refresh_pre_node_name, post_node_name]) output_names = [] for node_pair_names in insert_node_pairs: for (index, each_node_name) in enumerate(node_pair_names): name_with_sig = each_node_name node_name_prefix = name_with_sig.replace(':', '__port__').replace('^', '__hat__') print_node = Helper.create_node('Print', (node_name_prefix + '_print__{}'.format(index)), [(each_node_name + ':0'), (each_node_name + ':0')]) if (index == 0): msg = ';{}__print__:'.format(each_node_name) if ('swish_f32' in graph_info[pre_node_name].node.name): src_dt = attr_value_pb2.AttrValue(type=dtypes.float32.as_datatype_enum) else: src_dt = graph_info[pre_node_name].node.attr['T'] else: break print_node.attr['T'].CopyFrom(src_dt) print_node.attr['message'].s = msg.encode() print_node.attr['first_n'].i = (- 1) print_node.attr['summarize'].i = attr_u = [dtypes.as_dtype(src_dt.type).as_datatype_enum] print_node.attr['U'].list.CopyFrom(attr_value_pb2.AttrValue.ListValue(type=attr_u)) post_node_names = graph_info[Helper.node_name_from_input(each_node_name)].outputs if post_node_names: for post_node_name in post_node_names: post_node = graph_info[post_node_name].node if (each_node_name not in post_node.input): continue if ((post_node.op == 'FusedBatchNormV3') and ('_print_identity' not in graph_info[Helper.node_name_from_input(post_node.name)].node.input[0])): identity_node = Helper.create_node('Identity', (post_node.name + '_print_identity'), [graph_info[Helper.node_name_from_input(post_node.name)].node.input[0]]) identity_node.attr['T'].CopyFrom(src_dt) cur_graph.add_node(identity_node, graph_info[Helper.node_name_from_input(post_node.name)].node.input[0], [post_node.name]) identity_node.input.append(('^' + print_node.name)) else: post_node.input.append(('^' + print_node.name)) cur_graph.add_node(print_node, each_node_name, []) else: identity_node1 = Helper.create_node('Identity', (print_node.name + '_identity'), [print_node.name]) identity_node1.attr['T'].CopyFrom(src_dt) cur_graph.add_node(print_node, each_node_name, [identity_node1.name]) cur_graph.add_node(identity_node1, print_node.name, []) output_names.append(identity_node1.name) return cur_graph.dump_graph() def evaluate(self, model): input_tensor_names = model.input_tensor_names auto_trackable = model.model infer = auto_trackable.signatures['serving_default'] for (idx, (inputs, _)) in enumerate(self.dataloader): feed_dict = {} if (len(input_tensor_names) == 1): feed_dict[input_tensor_names[0]] = inputs else: assert (len(input_tensor_names) == len(inputs)), 'inputs len must equal with input_tensor' for (i, input_tensor_name) in enumerate(input_tensor_names): feed_dict[input_tensor_name] = inputs[i] _ = infer(**feed_dict) if (idx >= self.iterations): break def _inference(self, sampling_graph_def): logger.info('Start sampling on calibration dataset for Smooth Quantization.') reconstruct_saved_model(sampling_graph_def, self.func, self.frozen_func, self._saved_model, self.temp_path) model = Model(self.temp_path, modelType='llm_saved_model') self.evaluate(model) def _inference_for_calibration(self, model): sampling_graph_def = self._insert_print_for_activation(model) tmp_dump_file = tempfile.mkstemp(suffix='.log')[1] with CaptureOutputToFile(tmp_dump_file): self._inference(sampling_graph_def) self._parse_calibration_logs(tmp_dump_file) del sampling_graph_def def _get_weight_tensors(self): model = load.load(self.model, [tag_constants.SERVING]) for weight_tensor in model.variables: parsed_name = self.weight_name_mapping(weight_tensor.name) if (parsed_name in self._sq_target_node_names): self._sq_weight_tensor_dict[parsed_name] = weight_tensor.numpy() assert (len(self._sq_weight_tensor_dict) == len(self._sq_target_node_names)), 'Failed to get weights for some nodes, please check variables' def _generate_calibration_data(self, input_node_names, output_node_names): sorted_graph = QuantizeGraphHelper().get_sorted_graph(self.graph_def, input_node_names, output_node_names) for node in sorted_graph.node: if (node.op not in self.op_types): continue if ('while' in node.input[0]): continue self._sq_input_node_names.append(node.input[0]) self.print_node_list.append([node.name]) self._sq_target_node_names[node.input[1]] = node.name self._get_weight_tensors() sampling_graph_def = copy.deepcopy(self.graph_def) self._inference_for_calibration(sampling_graph_def) def __call__(self, input_node_names, output_node_names): (self.graph_def, self._saved_model, self.func, self.frozen_func, _, _) = parse_saved_model(self.model) self._generate_calibration_data(input_node_names, output_node_names) max_vals_per_channel = {} for (activation_name, output_tensor) in self._sq_output_tensor_dict.items(): max_val_per_channel = self._get_maxval_per_channel(output_tensor, percentile=self.percentile) max_vals_per_channel[activation_name] = max_val_per_channel return (max_vals_per_channel, self._sq_target_node_names, self._sq_weight_tensor_dict, self.graph_def)
class ImageExtents(): def __init__(self, minX, minY, minZ, maxX, maxY, maxZ): self.minX: float = minX self.minY: float = minY self.minZ: float = minZ self.maxX: float = maxX self.maxY: float = maxY self.maxZ: float = maxZ def get_c_image_extents(self): c_image_extents = bpConverterTypesC_ImageExtent(self.minX, self.minY, self.minZ, self.maxX, self.maxY, self.maxZ) return bpConverterTypesC_ImageExtentPtr(c_image_extents)
def register_box(name: str, box_type: Type=None, force: bool=False) -> Union[(Type, Callable)]: if (not isinstance(force, bool)): raise TypeError(f'force must be a boolean, but got {type(force)}') if (box_type is not None): _register_box(name=name, box_type=box_type, force=force) return box_type def _register(cls): _register_box(name=name, box_type=cls, force=force) return cls return _register
def disc(samples1, samples2, kernel, is_parallel=True, *args, **kwargs): d = 0 if (not is_parallel): for s1 in samples1: for s2 in samples2: d += kernel(s1, s2, *args, **kwargs) else: with concurrent.futures.ThreadPoolExecutor() as executor: for dist in executor.map(kernel_parallel_worker, [(s1, samples2, partial(kernel, *args, **kwargs)) for s1 in samples1]): d += dist d /= (len(samples1) * len(samples2)) return d
class _CLAM_Base(nn.Module): sizes = {'small': [1024, 512, 256], 'big': [1024, 512, 384], 'multiscale': [2048, 512, 256]} def __init__(self, size: Union[(str, List[int])]='small', dropout: bool=False, k_sample: int=8, n_classes: int=2, instance_loss_fn: Optional[Callable]=None, subtyping: bool=False, gate: bool=True, multi_head_attention: bool=False) -> None: super().__init__() if (instance_loss_fn is None): instance_loss_fn = nn.CrossEntropyLoss() self.size = (self.sizes[size] if isinstance(size, str) else size) fc = [nn.Linear(self.size[0], self.size[1]), nn.ReLU()] if dropout: fc.append(nn.Dropout(0.25)) att_fn = (Attn_Net_Gated if gate else Attn_Net) n_att = (1 if (not multi_head_attention) else n_classes) fc.append(att_fn(L=self.size[1], D=self.size[2], dropout=dropout, n_classes=n_att)) self.attention_net = nn.Sequential(*fc) self.instance_classifiers = nn.ModuleList([nn.Linear(self.size[1], 2) for _ in range(n_classes)]) self.k_sample = k_sample self.instance_loss_fn = instance_loss_fn self.n_classes = n_classes self.subtyping = subtyping def relocate(self): device = get_device() self.attention_net = self.attention_net.to(device) self.classifiers = self.classifiers.to(device) self.instance_classifiers = self.instance_classifiers.to(device) def create_positive_targets(length, device): return torch.full((length,), 1, device=device).long() def create_negative_targets(length, device): return torch.full((length,), 0, device=device).long() def _inst_eval(self, A, h, classifier, index=None): device = h.device if (len(A.shape) == 1): A = A.view(1, (- 1)) try: top_p_ids = torch.topk(A, self.k_sample)[1][(- 1)] except RuntimeError as e: raise RuntimeError(f'Error selecting top_k from sample with shape {A.shape}. Verify that all slides have at least {self.k_sample} tiles (min_tiles={self.k_sample}). Error: {e}') top_p = torch.index_select(h, dim=0, index=top_p_ids) top_n_ids = torch.topk((- A), self.k_sample, dim=1)[1][(- 1)] top_n = torch.index_select(h, dim=0, index=top_n_ids) p_targets = self.create_positive_targets(self.k_sample, device) n_targets = self.create_negative_targets(self.k_sample, device) all_targets = torch.cat([p_targets, n_targets], dim=0) all_instances = torch.cat([top_p, top_n], dim=0) logits = classifier(all_instances) all_preds = torch.topk(logits, 1, dim=1)[1].squeeze(1) instance_loss = self.instance_loss_fn(logits, all_targets) return (instance_loss, all_preds, all_targets) def _inst_eval_out(self, A, h, classifier, index=None): device = h.device if (len(A.shape) == 1): A = A.view(1, (- 1)) try: top_p_ids = torch.topk(A, self.k_sample)[1][(- 1)] except RuntimeError as e: raise RuntimeError(f'Error selecting top_k from sample with shape {A.shape}. Verify that all slides have at least {self.k_sample} tiles (min_tiles={self.k_sample}). Error: {e}') top_p = torch.index_select(h, dim=0, index=top_p_ids) p_targets = self.create_negative_targets(self.k_sample, device) logits = classifier(top_p) p_preds = torch.topk(logits, 1, dim=1)[1].squeeze(1) instance_loss = self.instance_loss_fn(logits, p_targets) return (instance_loss, p_preds, p_targets) def _logits_from_m(self, M): return self.classifiers(M) def _process_inputs(self, h, label=None, instance_eval=False): if (isinstance(h, (list, tuple)) and (len(h) == 2)): (h, label) = h elif (isinstance(h, (list, tuple)) and (len(h) == 3)): (h, label, instance_eval) = h if (h.ndim == 3): h = h.squeeze() if (h.shape[1] != self.size[0]): raise RuntimeError(f'Input feature size ({h.shape[1]}) does not match size of model first linear layer ({self.size[0]}). ') return (h, label, instance_eval) def instance_loss(self, A, h, label): total_inst_loss = 0.0 all_preds = [] all_targets = [] if ((label.ndim < 2) or (label.shape[1] != self.n_classes)): inst_labels = F.one_hot(label, num_classes=self.n_classes).squeeze() else: inst_labels = label[0] for i in range(len(self.instance_classifiers)): inst_label = inst_labels[i].item() classifier = self.instance_classifiers[i] if (inst_label == 1): (instance_loss, preds, targets) = self._inst_eval(A, h, classifier, i) all_preds.extend(preds.cpu().numpy()) all_targets.extend(targets.cpu().numpy()) elif self.subtyping: (instance_loss, preds, targets) = self._inst_eval_out(A, h, classifier, i) all_preds.extend(preds.cpu().numpy()) all_targets.extend(targets.cpu().numpy()) else: continue total_inst_loss += instance_loss if self.subtyping: total_inst_loss /= len(self.instance_classifiers) return (total_inst_loss, np.array(all_preds), np.array(all_targets)) def forward(self, h, label=None, instance_eval=False, return_attention=False): (h, label, instance_eval) = self._process_inputs(h, label, instance_eval) (A, h) = self.attention_net(h) A = A_raw = torch.transpose(A, 1, 0) A = F.softmax(A, dim=1) if instance_eval: (inst_loss, inst_targets, inst_preds) = self.instance_loss(A, h, label) inst_loss_dict = {'instance_loss': inst_loss, 'inst_labels': inst_targets, 'inst_preds': inst_preds} else: inst_loss_dict = {} M = torch.mm(A, h) logits = self._logits_from_m(M) if return_attention: return (logits, A_raw, inst_loss_dict) else: return (logits, inst_loss_dict) def calculate_attention(self, h): (h, *_) = self._process_inputs(h, None, None) (A, h) = self.attention_net(h) return torch.transpose(A, 1, 0) def get_last_layer_activations(self, h): (h, *_) = self._process_inputs(h, None, None) (A, h) = self.attention_net(h) A = F.softmax(A, dim=1) A = torch.transpose(A, 1, 0) M = torch.mm(A, h) return (M, A)
class _MSDataLoaderIter(_DataLoaderIter): def __init__(self, loader): self.dataset = loader.dataset self.collate_fn = loader.collate_fn self.batch_sampler = loader.batch_sampler self.num_workers = loader.num_workers self.pin_memory = (loader.pin_memory and torch.cuda.is_available()) self.timeout = loader.timeout self.done_event = threading.Event() self.sample_iter = iter(self.batch_sampler) if (self.num_workers > 0): self.worker_init_fn = loader.worker_init_fn self.index_queues = [multiprocessing.Queue() for _ in range(self.num_workers)] self.worker_queue_idx = 0 self.worker_result_queue = multiprocessing.SimpleQueue() self.batches_outstanding = 0 self.worker_pids_set = False self.shutdown = False self.send_idx = 0 self.rcvd_idx = 0 self.reorder_dict = {} base_seed = torch.LongTensor(1).random_()[0] self.workers = [multiprocessing.Process(target=_ms_loop, args=(self.dataset, self.index_queues[i], self.worker_result_queue, self.collate_fn, (base_seed + i), self.worker_init_fn, i)) for i in range(self.num_workers)] if (self.pin_memory or (self.timeout > 0)): self.data_queue = queue.Queue() if self.pin_memory: maybe_device_id = torch.cuda.current_device() else: maybe_device_id = None self.worker_manager_thread = threading.Thread(target=_worker_manager_loop, args=(self.worker_result_queue, self.data_queue, self.done_event, self.pin_memory, maybe_device_id)) self.worker_manager_thread.daemon = True self.worker_manager_thread.start() else: self.data_queue = self.worker_result_queue for w in self.workers: w.daemon = True w.start() _update_worker_pids(id(self), tuple((w.pid for w in self.workers))) _set_SIGCHLD_handler() self.worker_pids_set = True for _ in range((2 * self.num_workers)): self._put_indices()
def train(train_loader, model, criterion, optimizer, scheduler, epoch): batch_time = AverageMeter() losses = AverageMeter() top1 = AverageMeter() model.train() end = time.time() for (i, (input, target)) in enumerate(train_loader): output = model(input) loss = criterion(output, target) prec1 = accuracy(output.data, target, topk=(1,))[0] losses.update(loss.data.item(), input.size(0)) top1.update(prec1.item(), input.size(0)) optimizer.zero_grad() loss.backward() optimizer.step() scheduler.step() batch_time.update((time.time() - end)) end = time.time() if ((i % args.print_freq) == 0): print('Epoch: [{0}][{1}/{2}]\tTime {batch_time.val:.3f} ({batch_time.avg:.3f})\tLoss {loss.val:.4f} ({loss.avg:.4f})\ {top1.val:.3f} ({top1.avg:.3f})\tLR {scheduler._last_lr[0]:.6f}'.format(epoch, i, len(train_loader), batch_time=batch_time, loss=losses, top1=top1, scheduler=scheduler)) if args.tensorboard: log_value('train_loss', losses.avg, epoch) log_value('train_acc', top1.avg, epoch) log_value('learning_rate', scheduler._last_lr[0], epoch)
class RayScenario(Scenario, ABC): def __init__(self, name: str, ray_cluster_cpus: Union[(int, float)], ray_cluster_gpus: Union[(int, float)], ray_object_store_memory_cap_gigabytes: Union[(int, float)], ray_should_log_result_filter: Callable[([ResultDict], bool)]): super().__init__(name=name) self.ray_cluster_cpus = ray_cluster_cpus self.ray_cluster_gpus = ray_cluster_gpus self.ray_object_store_memory_cap_gigabytes = ray_object_store_memory_cap_gigabytes self.ray_should_log_result_filter = ray_should_log_result_filter
class BasicBlock(nn.Module): def __init__(self, in_chan, out_chan, stride=1): super(BasicBlock, self).__init__() self.conv1 = conv3x3(in_chan, out_chan, stride) self.bn1 = BatchNorm2d(out_chan) self.conv2 = conv3x3(out_chan, out_chan) self.bn2 = BatchNorm2d(out_chan) self.relu = nn.ReLU(inplace=True) self.downsample = None if ((in_chan != out_chan) or (stride != 1)): self.downsample = nn.Sequential(nn.Conv2d(in_chan, out_chan, kernel_size=1, stride=stride, bias=False), BatchNorm2d(out_chan)) def forward(self, x): residual = self.conv1(x) residual = self.bn1(residual) residual = self.relu(residual) residual = self.conv2(residual) residual = self.bn2(residual) shortcut = x if (self.downsample is not None): shortcut = self.downsample(x) out = (shortcut + residual) out = self.relu(out) return out
(version='2.0') def check_model(model): has_integerop = False has_qlinearop = False for node in model.graph.node: if node.op_type.endswith('Integer'): has_integerop = True elif node.op_type.startswith('QLinear'): has_qlinearop = True elif (node.op_type in ['QAttention', 'QGemm', 'QEmbedLayerNormalization']): has_qlinearop = True elif (node.op_type in ['Gather']): input_data = find_by_name(node.input[0], model.graph.initializer) if ((input_data is not None) and (numpy_helper.to_array(input_data).dtype in ['int8', 'uint8'])): has_qlinearop = True if has_integerop: logger.info('This model has Integer ops, these ops will be skipped.') if has_qlinearop: return True else: logger.info('This model has no QLinear ops, save the original model.') return False
_operation def real(a: torch.Tensor): if is_real(a): raise ValueError('Last dimension must have length 2.') return a[(..., 0)]
def find_missing(xs, ys, rs): rotations = [0, 3.14, 1.57, (- 1.57), 0.78, (- 0.78), 2.35, (- 2.35)] expectedPoints = [] missingPointsX = [] missingPointsY = [] for (xd, yd) in zip(*required_points()): expectedPoints.append((xd, yd)) isMissing = [True for _ in rotations] for (x, y, r) in zip(xs, ys, rs): dist = math.sqrt((((x - xd) ** 2) + ((y - yd) ** 2))) if (dist <= 0.1): for (rdi, rd) in enumerate(rotations): angleDiff = abs(angle_difference(r, rd)) if (angleDiff <= 0.26): isMissing[rdi] = False if any(isMissing): missingPointsX.append(xd) missingPointsY.append(yd) return (missingPointsX, missingPointsY)
def initialize_hyperparameters(PATHS: dict, load_target: str, config_name: str='default', n_envs: int=1): if (load_target is None): hyperparams = load_hyperparameters_json(PATHS=PATHS, from_scratch=True, config_name=config_name) hyperparams['agent_name'] = PATHS['model'].split('/')[(- 1)] else: hyperparams = load_hyperparameters_json(PATHS=PATHS) check_batch_size(n_envs, hyperparams['batch_size'], hyperparams['m_batch_size']) hyperparams['n_steps'] = int((hyperparams['batch_size'] / n_envs)) write_hyperparameters_json(hyperparams, PATHS) print_hyperparameters(hyperparams) return hyperparams
_criterion('legacy_masked_lm_loss') class LegacyMaskedLmLoss(FairseqCriterion): def __init__(self, task, masked_lm_only, nsp_loss_weight): super().__init__(task) self.masked_lm_only = masked_lm_only self.nsp_loss_weight = nsp_loss_weight def add_args(parser): parser.add_argument('--masked-lm-only', default=False, action='store_true', help='compute MLM loss only') parser.add_argument('--nsp-loss-weight', default=1.0, type=float, help='weight for next sentence prediction loss (default 1)') def forward(self, model, sample, reduce=True): (lm_logits, output_metadata) = model(**sample['net_input']) lm_logits = lm_logits.view((- 1), lm_logits.size((- 1))) lm_targets = sample['lm_target'].view((- 1)) lm_loss = compute_cross_entropy_loss(lm_logits, lm_targets, self.padding_idx) ntokens = utils.strip_pad(lm_targets, self.padding_idx).numel() loss = (lm_loss / ntokens) nsentences = sample['nsentences'] sentence_loss = None if (not self.masked_lm_only): sentence_logits = output_metadata['sentence_logits'] sentence_targets = sample['sentence_target'].view((- 1)) nsentences = sentence_targets.size(0) if (sentence_logits is not None): sentence_loss = compute_cross_entropy_loss(sentence_logits, sentence_targets) loss += (self.nsp_loss_weight * (sentence_loss / nsentences)) sample_size = 1 logging_output = {'loss': (utils.item(loss.data) if reduce else loss.data), 'lm_loss': (utils.item(lm_loss.data) if reduce else lm_loss.data), 'sentence_loss': ((utils.item(sentence_loss.data) if reduce else sentence_loss.data) if (sentence_loss is not None) else 0.0), 'ntokens': ntokens, 'nsentences': nsentences, 'sample_size': sample_size} return (loss, sample_size, logging_output) def aggregate_logging_outputs(logging_outputs): lm_loss_sum = sum((log.get('lm_loss', 0) for log in logging_outputs)) sentence_loss_sum = sum((log.get('sentence_loss', 0) for log in logging_outputs)) ntokens = sum((log.get('ntokens', 0) for log in logging_outputs)) nsentences = sum((log.get('nsentences', 0) for log in logging_outputs)) sample_size = sum((log.get('sample_size', 0) for log in logging_outputs)) agg_loss = sum((log.get('loss', 0) for log in logging_outputs)) agg_output = {'loss': (((agg_loss / sample_size) / math.log(2)) if (sample_size > 0) else 0.0), 'lm_loss': (((lm_loss_sum / ntokens) / math.log(2)) if (ntokens > 0) else 0.0), 'sentence_loss': (((sentence_loss_sum / nsentences) / math.log(2)) if (nsentences > 0) else 0.0), 'nll_loss': (((lm_loss_sum / ntokens) / math.log(2)) if (ntokens > 0) else 0.0), 'ntokens': ntokens, 'nsentences': nsentences, 'sample_size': sample_size} return agg_output def logging_outputs_can_be_summed() -> bool: return True
class DeviceType(object): CPU = 'CPU' GPU = 'GPU' HEXAGON = 'HEXAGON' HTA = 'HTA' APU = 'APU' HTP = 'HTP' QUANTIZE = 'QUANTIZE'
class DeResNetBlockGroupNorm(nn.Module): def __init__(self, inplanes, planes, num_groups, stride=1, output_padding=0, activation='relu'): super(DeResNetBlockGroupNorm, self).__init__() assert (activation in ['relu', 'elu', 'leaky_relu']) self.deconv1 = deconv3x3(inplanes, planes, stride, output_padding) self.gn1 = nn.GroupNorm(num_groups, planes) if (activation == 'relu'): self.activation = nn.ReLU(inplace=True) elif (activation == 'elu'): self.activation = nn.ELU(inplace=True) else: self.activation = nn.LeakyReLU(inplace=True, negative_slope=0.1) self.deconv2 = deconv3x3(planes, planes) self.gn2 = nn.GroupNorm(num_groups, planes) downsample = None if ((stride != 1) or (inplanes != planes)): downsample = nn.Sequential(nn.ConvTranspose2d(inplanes, planes, kernel_size=1, stride=stride, output_padding=output_padding, bias=False), nn.GroupNorm(num_groups, planes)) self.downsample = downsample self.reset_parameters() def reset_parameters(self): nn.init.constant_(self.gn1.weight, 1.0) nn.init.constant_(self.gn1.bias, 0.0) nn.init.constant_(self.gn2.weight, 1.0) nn.init.constant_(self.gn2.bias, 0.0) if (self.downsample is not None): assert isinstance(self.downsample[1], nn.GroupNorm) nn.init.constant_(self.downsample[1].weight, 1.0) nn.init.constant_(self.downsample[1].bias, 0.0) def init(self, x, init_scale=1.0): with torch.no_grad(): return self(x) def forward(self, x): residual = x out = self.deconv1(x) out = self.gn1(out) out = self.activation(out) out = self.deconv2(out) out = self.gn2(out) if (self.downsample is not None): residual = self.downsample(x) out += residual out = self.activation(out) return out
def distributed_init(cfg: FairseqConfig): if isinstance(cfg, Namespace): from fairseq.dataclass.utils import convert_namespace_to_omegaconf cfg = convert_namespace_to_omegaconf(cfg) if (not cfg.common.tpu): if (torch.distributed.is_available() and torch.distributed.is_initialized()): warnings.warn('Distributed is already initialized, cannot initialize twice!') else: logger.info('distributed init (rank {}): {}'.format(cfg.distributed_training.distributed_rank, cfg.distributed_training.distributed_init_method)) dist.init_process_group(backend=cfg.distributed_training.distributed_backend, init_method=cfg.distributed_training.distributed_init_method, world_size=cfg.distributed_training.distributed_world_size, rank=cfg.distributed_training.distributed_rank) logger.info('initialized host {} as rank {}'.format(socket.gethostname(), cfg.distributed_training.distributed_rank)) if torch.cuda.is_available(): dist.all_reduce(torch.zeros(1).cuda()) cfg.distributed_training.distributed_rank = torch.distributed.get_rank() else: assert (xm.xrt_world_size() == cfg.distributed_training.distributed_world_size) global _USE_XLA _USE_XLA = True cfg.distributed_training.device_id = xm.get_local_ordinal() cfg.distributed_training.distributed_rank = xm.get_ordinal() xm.rendezvous('distributed_init') if is_master(cfg.distributed_training): logging.getLogger().setLevel(logging.INFO) else: logging.getLogger().setLevel(logging.WARNING) if (cfg.common.model_parallel_size > 1): try: from fairseq.model_parallel.megatron.mpu import initialize_model_parallel, model_parallel_cuda_manual_seed except ImportError: raise ImportError('\n\nPlease install the megatron submodule:\n\n git submodule update --init fairseq/model_parallel/megatron') global _USE_MEGATRON _USE_MEGATRON = True initialize_model_parallel(cfg.common.model_parallel_size) model_parallel_cuda_manual_seed(cfg.common.seed) model_part_number = get_model_parallel_rank() cfg.checkpoint.checkpoint_suffix += '-model_part-{0}'.format(model_part_number) if (hasattr(cfg, 'model') and (getattr(cfg.model, 'base_layers', 0) > 0)): cfg.checkpoint.checkpoint_suffix = f'-rank-{cfg.distributed_training.distributed_rank}' return cfg.distributed_training.distributed_rank
def evaluate(gold, guess, ks, rank_keys): pp = pprint.PrettyPrinter(indent=4) gold_dataset = kilt_utils.load_data(gold) guess_dataset = kilt_utils.load_data(guess) (gold_dataset, guess_dataset) = eval_downstream.validate_input(gold_dataset, guess_dataset) guess_dataset = filter_answers(guess_dataset) result = compute(gold_dataset, guess_dataset, ks, rank_keys) pp.pprint(result) return result
def compute(inp, outp, settings, force): sr = settings['samplerate'] _lazy_y = None def load(): nonlocal _lazy_y if (_lazy_y is None): (_lazy_y, _sr) = librosa.load(inp, sr=sr) assert (_sr == sr), _sr return _lazy_y exists = os.path.exists(outp) size = 0 if exists: size = os.stat(outp).st_size valid = (exists and (size > 0)) if ((not valid) or force): start_time = time.time() y = load() loaded_time = time.time() f = features.compute_mels(y, settings) computed_time = time.time() numpy.savez(outp, f) saved_time = time.time() if settings['augmentations']: paths = [outp.replace('.npz', '.aug{}.npz'.format(aug)) for aug in range(12)] exists = [os.path.exists(p) for p in paths] if ((not all(exists)) or force): y = load() augmented = augmentations(y, sr).values() assert (settings['augmentations'] == 12) assert (len(augmented) == settings['augmentations']), len(augmented) for (aug, (augdata, path)) in enumerate(zip(augmented, paths)): f = features.compute_mels(augdata, settings) numpy.savez(path, f) return outp
class VegaHTML(object): def __init__(self, renderer): self.specification = dict(width=renderer.figwidth, height=renderer.figheight, data=renderer.data, scales=renderer.scales, axes=renderer.axes, marks=renderer.marks) def html(self): id = random.randint(0, (2 ** 16)) html = ('<div id="vis%d"></div>' % id) html += '<script>\n' html += (VEGA_TEMPLATE % (json.dumps(self.specification), id)) html += '</script>\n' return html def _repr_html_(self): return self.html()
def build_fake_yaml(): fake_yaml = '\n model:\n name: fake_yaml\n framework: tensorflow\n inputs: x\n outputs: op_to_store\n device: cpu\n quantization:\n calibration:\n sampling_size: 10\n evaluation:\n accuracy:\n metric:\n topk: 1\n tuning:\n strategy:\n name: mse\n exit_policy:\n max_trials: 1\n accuracy_criterion:\n relative: 0.01\n workspace:\n path: saved\n ' y = yaml.load(fake_yaml, Loader=yaml.SafeLoader) with open('fake_yaml.yaml', 'w', encoding='utf-8') as f: yaml.dump(y, f) f.close()
def preprocess_iwslt17(root: str, src: str, tgt: str, bpe_size: Optional[int], need_chars: bool, bbpe_size: Optional[int], need_bytes: bool): in_root = op.join(root, f'{src}-{tgt}') for lang in [src, tgt]: _convert_train(op.join(in_root, f'train.tags.{src}-{tgt}.{lang}'), op.join(root, f'train.{lang}')) _convert_xml(op.join(in_root, f'IWSLT17.TED.dev2010.{src}-{tgt}.{lang}.xml'), op.join(root, f'valid.{lang}')) _convert_xml(op.join(in_root, f'IWSLT17.TED.tst2015.{src}-{tgt}.{lang}.xml'), op.join(root, f'test.{lang}')) for lang in [src, tgt]: for split in SPLITS: pretokenize(op.join(root, f'{split}.{lang}'), op.join(root, f'{split}.moses.{lang}'), src, tgt) if (bpe_size is not None): concated_train_path = op.join(root, 'train.all') _concat_files([op.join(root, 'train.moses.fr'), op.join(root, 'train.moses.en')], concated_train_path) bpe_model_prefix = op.join(root, f'spm_bpe{bpe_size}') _get_bpe(concated_train_path, bpe_model_prefix, bpe_size) os.remove(concated_train_path) for lang in [src, tgt]: for split in SPLITS: _apply_bpe((bpe_model_prefix + '.model'), op.join(root, f'{split}.moses.{lang}'), op.join(root, f'{split}.moses.bpe{bpe_size}.{lang}')) if need_bytes: for lang in [src, tgt]: for split in SPLITS: _get_bytes(op.join(root, f'{split}.moses.{lang}'), op.join(root, f'{split}.moses.bytes.{lang}')) if need_chars: for lang in [src, tgt]: for split in SPLITS: _get_chars(op.join(root, f'{split}.moses.{lang}'), op.join(root, f'{split}.moses.chars.{lang}')) if (bbpe_size is not None): bchar_path = op.join(root, 'train.bchar') _convert_to_bchar(op.join(root, 'train.moses'), src, tgt, bchar_path) bbpe_model_prefix = op.join(root, f'spm_bbpe{bbpe_size}') _get_bpe(bchar_path, bbpe_model_prefix, bbpe_size) os.remove(bchar_path) for lang in [src, tgt]: for split in SPLITS: _apply_bbpe((bbpe_model_prefix + '.model'), op.join(root, f'{split}.moses.{lang}'), op.join(root, f'{split}.moses.bbpe{bbpe_size}.{lang}'))
class AudioFinetuningConfig(AudioPretrainingConfig): eval_wer: bool = field(default=False, metadata={'help': 'compute WER for Seq2Seq models'}) eval_wer_config: GenerationConfig = field(default_factory=(lambda : GenerationConfig()), metadata={'help': 'beam search config for evaluating wer during training'}) eval_wer_tokenizer: Any = field(default=None, metadata={'help': 'tokenizer config for evaluating wer during training'}) eval_wer_post_process: str = field(default='letter', metadata={'help': 'remove BPE tokens before scoring (can be sentencepiece, letter, and more)'}) eval_bleu: bool = field(default=False, metadata={'help': 'evaluation with BLEU scores'}) eval_bleu_detok: Optional[str] = field(default=None, metadata={'help': "detokenize before computing BLEU (e.g., 'moses'); required if using --eval-bleu; use 'space' to disable detokenization; see fairseq.data.encoders for other options"}) eval_bleu_detok_args: str = field(default='{}', metadata={'help': 'args for building the tokenizer, if needed'}) eval_tokenized_bleu: bool = field(default=False, metadata={'help': 'compute tokenized BLEU instead of sacrebleu'}) eval_bleu_remove_bpe: Optional[str] = field(default=None, metadata={'help': 'remove BPE before computing BLEU'}) eval_bleu_args: str = field(default='{}', metadata={'help': 'generation args for BLUE scoring, e.g., \'{"beam": 4, "lenpen": 0.6}\''}) eval_bleu_print_samples: bool = field(default=False, metadata={'help': 'print sample generations during validation'}) autoregressive: bool = field(default=False, metadata={'help': "required for autoregressive decoders (like seq2seq models); adds 'prev_output_tokens' to input and appends eos to target"}) rebuild_batches: bool = True
def get_document_ids(source_docs, indexes): indexes = sorted([(key, value[0], value[1]) for (key, value) in indexes.items()], key=(lambda x: x[0])) doc_ids = [] for (i, partial_start, partial_end) in indexes: try: doc_ids.append((source_docs[i], partial_start, partial_end)) except IndexError: pass return doc_ids
def RI(sentence, alpha_ri, n_aug=9): sentence = get_only_chars(sentence) words = sentence.split(' ') num_words = len(words) augmented_sentences = [] n_ri = max(1, int((alpha_ri * num_words))) for _ in range(n_aug): a_words = random_addition(words, n_ri) augmented_sentences.append(' '.join(a_words)) augmented_sentences = [get_only_chars(sentence) for sentence in augmented_sentences] shuffle(augmented_sentences) augmented_sentences.append(sentence) return augmented_sentences
def ResNet18Compressed(channels): return ResNetCompressed(BasicBlockCompressed, [2, 2, 2, 2], channels)
_registry(operator_type='BinaryAdd') class BinaryAdd(Operator): def __init__(self): super().__init__()
def _to_py_obj(x): if (x.lower() in ['true', 'yes', 'on']): return True if (x.lower() in ['false', 'no', 'off']): return False try: obj = eval(x) if (type(obj).__name__ in ['int', 'float', 'tuple', 'list', 'dict', 'NoneType']): x = obj except: pass return x
class Speedometer(object): def __init__(self, batch_size, frequent=50, batches_per_epoch=None, epochs=None): self.batch_size = batch_size self.frequent = frequent self.batches_per_epoch = batches_per_epoch self.epochs = epochs self.epoch = (- 1) self.init = False self.tic = 0 self.last_count = 0 self.data_in_time = 0.0 self.data_transfer_time = 0.0 self.forward_time = 0.0 self.backward_time = 0.0 self.optimizer_time = 0.0 self.metric_time = 0.0 def __call__(self, param): count = param.nbatch if (self.last_count > count): self.init = False self.last_count = count self.data_in_time += param.data_in_time self.data_transfer_time += param.data_transfer_time self.forward_time += param.forward_time self.backward_time += param.backward_time self.optimizer_time += param.optimizer_time self.metric_time += param.metric_time if self.init: if ((count % self.frequent) == 0): speed = ((self.frequent * self.batch_size) / (time.time() - self.tic)) data_in_time = (self.data_in_time / self.frequent) data_transfer_time = (self.data_transfer_time / self.frequent) forward_time = (self.forward_time / self.frequent) backward_time = (self.backward_time / self.frequent) optimizer_time = (self.optimizer_time / self.frequent) metric_time = (self.metric_time / self.frequent) eta = (((((((self.epochs - self.epoch) - 1) * self.batches_per_epoch) + self.batches_per_epoch) - param.nbatch) * self.batch_size) / speed) eta = int((eta / 60.0)) eta_m = (eta % 60) eta_h = (int(((eta - eta_m) / 60)) % 24) eta_d = int((((eta - eta_m) - (eta_h * 60)) / (24 * 60))) s = '' if (param.eval_metric is not None): prefix = ('Epoch[%d] Batch [%d]\t' % (param.epoch, count)) (name, value) = param.eval_metric.get() s = (prefix + ('Speed: %.2f samples/s ETA: %d d %2d h %2d m\tData: %.3f Tran: %.3f F: %.3f B: %.3f O: %.3f M: %.3f\tTrain-' % (speed, eta_d, eta_h, eta_m, data_in_time, data_transfer_time, forward_time, backward_time, optimizer_time, metric_time))) for (n, v) in zip(name, value): s += ('%s=%f,\t' % (n, v)) else: prefix = ('Epoch[%d] Batch [%d]\t' % (param.epoch, count)) s = (prefix + ('Speed: %.2f ETA: %d d %2d h %2d m samples/s\tData: %.3f Tran: %.3f F: %.3f B: %.3f O: %.3f M: %.3f' % (speed, eta_d, eta_h, eta_m, data_in_time, data_transfer_time, forward_time, backward_time, optimizer_time, metric_time))) if (param.rank is not None): s = (('Rank[%3d]' % param.rank) + s) logging.info(s) print(s) self.tic = time.time() self.data_in_time = 0.0 self.data_transfer_time = 0.0 self.forward_time = 0.0 self.backward_time = 0.0 self.optimizer_time = 0.0 self.metric_time = 0.0 else: self.init = True self.epoch += 1 if (param.eval_metric is not None): (name, value) = param.eval_metric.get() s = ('Epoch[%d] Batch [%d]\tSpeed: - samples/sec ETA: - d - h - m\tTrain-' % (param.epoch, 0)) for (n, v) in zip(name, value): s += ('%s=%f,\t' % (n, v)) else: s = ('Epoch[%d] Batch [%d]\tSpeed: - samples/sec ETA: - d - h - m' % (param.epoch, 0)) if (param.rank is not None): s = (('Rank[%3d]' % param.rank) + s) logging.info(s) print(s) self.tic = time.time()
def sample(nodeInfor, edgeInfor): nodeId = [nodeInfor[i][0] for i in range(len(nodeInfor))] longitude = [nodeInfor[i][1] for i in range(len(nodeInfor))] latitude = [nodeInfor[i][2] for i in range(len(nodeInfor))] n = len(nodeId) A1 = np.array(([([0] * n)] * n)) Graph1 = nx.Graph(A1) column = [str(nodeId[i]) for i in range(n)] mapping = {0: str(nodeId[0])} for i in range(0, (len(column) - 1)): mapping.setdefault((i + 1), column[(i + 1)]) Graph1 = nx.relabel_nodes(Graph1, mapping) edgeSet1 = list() for i in range(len(edgeInfor)): edgeRow = edgeInfor[i] edgeSet1.append((str(edgeRow[0]), str(edgeRow[1]))) edgeSet = list(set(edgeSet1)) Graph1.add_edges_from(edgeSet) deleteNumber = int((len(Graph1.edges) * 0.2)) T = nx.minimum_spanning_tree(Graph1) potentialDelete = list((set(Graph1.edges) - set(T.edges))) realDelete1 = random.sample(potentialDelete, deleteNumber) realDelete2 = random.sample(potentialDelete, deleteNumber) print('len(realDelete1)', len(realDelete1), 'len(realDelete2)', len(realDelete2)) edgeInforNew = list() for i in range(len(edgeInfor)): edgeRow = edgeInfor[i] item = list() if (((str(edgeRow[0]), str(edgeRow[1])) in realDelete1) or ((str(edgeRow[1]), str(edgeRow[0])) in realDelete1)): item = [edgeRow[0], edgeRow[1], edgeRow[2], 0] else: item = [edgeRow[0], edgeRow[1], edgeRow[2], 1] if (((str(edgeRow[0]), str(edgeRow[1])) in realDelete2) or ((str(edgeRow[1]), str(edgeRow[0])) in realDelete2)): item.append(0) else: item.append(1) edgeInforNew.append(item) returnEdgeInforNew = list() for i in range(len(edgeInforNew)): returnEdgeInforNew.append((edgeInforNew[i][0], edgeInforNew[i][1], edgeInforNew[i][2], edgeInforNew[i][3], edgeInforNew[i][4])) return returnEdgeInforNew
def _relaunch(): log.warn('Relaunching...') if (sys.argv[:3] == ['-c', 'install', '--single-version-externally-managed']): sys.argv[0] = 'setup.py' args = ([sys.executable] + sys.argv) sys.exit(subprocess.call(args))
def find_program(basename): names = [basename] if (os.name == 'nt'): extensions = ('.exe', '.bat', '.cmd', '.dll') if (not basename.endswith(extensions)): names = ([(basename + ext) for ext in extensions] + [basename]) for name in names: path = is_program_installed(name) if path: return path
class Dataset(object): def __init__(self): self._instances = [] def add_instance(self, propertyDict): self._instances.append(propertyDict) def get_phraselist(self): return [instance['phrase'] for instance in self._instances] def get_imagelist(self): return [instance['image'] for instance in self._instances] def get_boxlist(self): return [instance['box'] for instance in self._instances] def get_instances(self): return self._instances def get_count(self): return len(self._instances) count = property(get_count) size = property(get_count) instances = property(get_instances) phrases = property(get_phraselist) images = property(get_imagelist) boxes = property(get_boxlist)
def main(args): from benchmark.models.musichubert_hf.extract_bert_features import main as extract_hubert_features_main from benchmark.models.music2vec.extract_music2vec_features import main as extract_data2vec_features_main from benchmark.models.data2vec.extract_data2vec_features import main as extract_data2vec_audio_features_main from benchmark.models.handcrafted.extract_handcrafted_features import main as extract_handcrafted_features_main from benchmark.models.jukebox.extract_jukemir_features import main as extract_jukemir_features_main from benchmark.models.musicnn.extract_musicnn_features import main as extract_musicnn_features_main from benchmark.models.clmr.extract_clmr_features import main as extract_clmr_features_main from benchmark.models.mule.extract_mule_features import main as extract_mule_features_main from benchmark.models.hubert.extract_hubert_features import main as extract_speech_hubert_features_main config = load_config(args.config, namespace=True) if ((args.override is not None) and (args.override.lower() != 'none')): override_config(args.override, config) config = merge_args_to_config(args, config) print_config(config) representation_name = config.dataset.pre_extract.feature_extractor.pretrain.name extract_main = bench.NAME_TO_EXTRACT_FEATURES_MAIN[representation_name] extract_main = eval(extract_main) extract_main(config)
class SegmentationModuleBase(nn.Module): def __init__(self): super(SegmentationModuleBase, self).__init__() def pixel_acc(self, pred, label): (_, preds) = torch.max(pred, dim=1) valid = (label >= 0).long() acc_sum = torch.sum((valid * (preds == label).long())) pixel_sum = torch.sum(valid) acc = (acc_sum.float() / (pixel_sum.float() + 1e-10)) return acc
class MultipleChoiceQuestion(NamedTuple): stem: str choices: List[Choice] id_: str = None answerKey: str = None def from_jsonl(line: str) -> 'MultipleChoiceQuestion': blob = json.loads(line) question = blob['question'] return MultipleChoiceQuestion(id_=blob['id'], stem=question['stem'], choices=[Choice(c['label'], c['text']) for c in question['choices']], answerKey=blob['answerKey']) def from_jsonl_ours(line: str, idx: int) -> 'MultipleChoiceQuestion': blob = json.loads(line) return MultipleChoiceQuestion(id_=idx, stem=blob['question'], choices=[Choice(num2char(idx), blob['options'][num2char(idx)]) for idx in range(len(blob['options']))], answerKey=blob['answer_idx'])
class FlaxRobertaModel(metaclass=DummyObject): _backends = ['flax'] def __init__(self, *args, **kwargs): requires_backends(self, ['flax'])
class MaskFromDensePoseSampler(): def __call__(self, instances: Instances) -> BitMasks: return ToMaskConverter.convert(instances.pred_densepose, instances.pred_boxes, instances.image_size)
def _calc_dynamic_intervals(start_interval, dynamic_interval_list): assert mmcv.is_list_of(dynamic_interval_list, tuple) dynamic_milestones = [0] dynamic_milestones.extend([dynamic_interval[0] for dynamic_interval in dynamic_interval_list]) dynamic_intervals = [start_interval] dynamic_intervals.extend([dynamic_interval[1] for dynamic_interval in dynamic_interval_list]) return (dynamic_milestones, dynamic_intervals)
def read_txt_file(path): with open(path, 'r') as f: lists = f.read().splitlines() lists = [s.strip().split(' ') for s in lists] return lists
def get_preprocessing(name, is_training=False): preprocessing_fn_map = {'resnet_v1_152': resnet_preprocessing, 'mobilenet_v1': mobilenet_preprocessing} if (name not in preprocessing_fn_map): raise ValueError(('Preprocessing name [%s] was not recognized' % name)) def preprocessing_fn(image, output_height, output_width): return preprocessing_fn_map[name].preprocess_image(image, output_height, output_width, is_training=is_training) return preprocessing_fn
def _data_augmentation(image, label, bbox, data_augmentation_args): print('Use data_augmentation_args: ', data_augmentation_args) if data_augmentation_args['crop_bbox']: sample_distorted_bounding_box = tf.image.sample_distorted_bounding_box(tf.shape(image), bounding_boxes=bbox, min_object_covered=0.1, aspect_ratio_range=[0.75, 1.33], area_range=[0.05, 1.0], max_attempts=100, use_image_if_no_bounding_boxes=True) (bbox_begin, bbox_size, _) = sample_distorted_bounding_box (offset_y, offset_x, _) = tf.unstack(bbox_begin) (target_height, target_width, _) = tf.unstack(bbox_size) image = tf.image.crop_to_bounding_box(image, offset_y, offset_x, target_height, target_width) image = tf.expand_dims(image, 0) image = tf.image.resize_bilinear(image, [height, width], align_corners=False) image = tf.squeeze(image) if data_augmentation_args['resize']: new_size = tf.random_uniform([], minval=256, maxval=(512 + 1), dtype=tf.int32) image = tf.expand_dims(image, 0) image = tf.image.resize_bilinear(image, [new_size, new_size], align_corners=False) image = tf.squeeze(image) image = tf.image.random_crop(image, [height, width, depth]) if data_augmentation_args['padding']: reshaped_image = tf.image.resize_image_with_crop_or_pad(image, 256, 256) image = tf.random_crop(reshaped_image, [height, width, depth]) image = tf.cast(image, tf.float32) if (not (data_augmentation_args['crop_bbox'] or data_augmentation_args['resize'] or data_augmentation_args['padding'])): image = tf.expand_dims(image, 0) image = tf.image.resize_bilinear(image, [256, 256]) image = tf.squeeze(image) image = tf.image.resize_image_with_crop_or_pad(image, height, width) if data_augmentation_args['mirroring']: image = tf.image.random_flip_left_right(image) if data_augmentation_args['bright']: image = tf.image.random_brightness(image, max_delta=32.0) image = tf.image.random_saturation(image, lower=0.5, upper=1.5) mean = data_augmentation_args['mean'] std = data_augmentation_args['std'] if isinstance(mean, list): mean = tf.reshape(mean, [1, 1, 3]) std = tf.reshape(std, [1, 1, 3]) image = ((image - mean) / std) else: image = ((image - mean) / std) image.set_shape([height, width, depth]) return (image, label, bbox)
class LogScheduler(LRScheduler): def __init__(self, optimizer, start_lr=0.03, end_lr=0.0005, epochs=50, last_epoch=(- 1), **kwargs): self.start_lr = start_lr self.end_lr = end_lr self.epochs = epochs self.lr_spaces = np.logspace(math.log10(start_lr), math.log10(end_lr), epochs) super(LogScheduler, self).__init__(optimizer, last_epoch)
class SepConvOp(nn.Module): def __init__(self, C_in, C_out, kernel_size, act_op, affine=True): super(SepConvOp, self).__init__() padding = PADDING_OPS[kernel_size] kernel_size = KERNEL_SIZE_OPS[kernel_size] activation = ACTIVATION_OPS[act_op] if (not activation): self.op = nn.Sequential(nn.Conv2d(C_in, C_in, kernel_size=kernel_size, padding=padding, groups=C_in, bias=False), nn.Conv2d(C_in, C_out, kernel_size=1, padding=0, bias=False)) else: self.op = nn.Sequential(nn.Conv2d(C_in, C_in, kernel_size=kernel_size, padding=padding, groups=C_in, bias=False), nn.Conv2d(C_in, C_out, kernel_size=1, padding=0, bias=False), activation) def forward(self, x): return self.op(x)
class ArgMaxParameter(_message.Message): __metaclass__ = _reflection.GeneratedProtocolMessageType DESCRIPTOR = _ARGMAXPARAMETER
def discretized_mix_logistic_topk(means, logscales, logit_probs, range, bin_size, lower, upper, topk=1) -> Tuple[(torch.Tensor, torch.LongTensor)]: eps = 1e-12 means = means.unsqueeze(1) logscales = logscales.unsqueeze(1) logit_probs = logit_probs.unsqueeze(1) x = torch.arange((- range), (range + 1), 1.0, device=means.device).unsqueeze(0).unsqueeze(2) x = x.div(range) centered_x = (x - means) if isinstance(logscales, float): inv_stdv = np.exp((- logscales)) else: inv_stdv = torch.exp((- logscales)) min_in = (inv_stdv * (centered_x - bin_size)) plus_in = (inv_stdv * (centered_x + bin_size)) x_in = (inv_stdv * centered_x) cdf_min = torch.sigmoid(min_in) cdf_plus = torch.sigmoid(plus_in) cdf_delta = (cdf_plus - cdf_min) log_cdf_mid = torch.log((cdf_delta + eps)) log_cdf_approx = (((x_in - logscales) - (2.0 * F.softplus(x_in))) + np.log((2 * bin_size))) log_cdf_low = (plus_in - F.softplus(plus_in)) log_cdf_up = (- F.softplus(min_in)) log_cdf = torch.where(cdf_delta.gt(1e-05), log_cdf_mid, log_cdf_approx) log_cdf = torch.where(x.ge(lower), log_cdf, log_cdf_low) log_cdf = torch.where(x.le(upper), log_cdf, log_cdf_up) log_probs = torch.logsumexp((log_cdf + logit_probs), dim=2) (log_probs, idx) = log_probs.topk(topk, dim=1) return (log_probs, (idx - range))
def controled_step(short_memory, long_memory, selected_instruction, current_paras, request: gr.Request): if (current_paras == ''): return ('', '', '', '', '', '') global _CACHE cookie = request.headers['cookie'] cookie = cookie.split('; _gat_gtag')[0] cache = _CACHE[cookie] if ('writer' not in cache): start_input_to_human = cache['start_input_to_human'] start_input_to_human['output_instruction'] = selected_instruction init_paragraphs = cache['init_paragraphs'] human = Human(input=start_input_to_human, memory=None, embedder=embedder, model=llm_model, tokenizer=llm_tokenizer) human.step() start_short_memory = init_paragraphs['Summary'] writer_start_input = human.output writer = AIWriter(input=writer_start_input, short_memory=start_short_memory, long_memory=[init_paragraphs['Paragraph 1'], init_paragraphs['Paragraph 2'], init_paragraphs['Paragraph 3']], memory_index=None, embedder=embedder, model=llm_model, tokenizer=llm_tokenizer) cache['writer'] = writer cache['human'] = human writer.step() else: human = cache['human'] writer = cache['writer'] output = writer.output output['output_memory'] = short_memory output['output_instruction'] = selected_instruction human.input = output human.step() writer.input = human.output writer.step() return (writer.output['output_memory'], parse_instructions(writer.long_memory), ((current_paras + '\n\n') + writer.output['input_paragraph']), *writer.output['output_instruction'])
class MishActivation(nn.Module): def __init__(self): super().__init__() if (version.parse(version.parse(torch.__version__).base_version) < version.parse('1.9')): self.act = self._mish_python else: self.act = nn.functional.mish def _mish_python(self, input: Tensor) -> Tensor: return (input * torch.tanh(nn.functional.softplus(input))) def forward(self, input: Tensor) -> Tensor: return self.act(input)
def rmdir(path: str) -> None: if path.startswith('s3'): invalidOperationError(False, 'not implement') elif path.startswith('hdfs://'): cmd = 'hdfs dfs -rm -r {}'.format(path) result = subprocess.getstatusoutput(cmd) if (result[0] != 0): invalidOperationError(False, result[1]) else: if path.startswith('file://'): path = path[len('file://'):] os.rmdir(path)
class TFEsmModel(metaclass=DummyObject): _backends = ['tf'] def __init__(self, *args, **kwargs): requires_backends(self, ['tf'])
def move(src_path: str, tgt_path, suffix: str, split_name: str, url): os.mkdir(os.path.join(tgt_path, split_name)) os.mkdir(os.path.join(tgt_path, split_name, 'doc')) os.mkdir(os.path.join(tgt_path, split_name, 'abs')) with open(url, 'r', encoding='utf-8') as fd: lines = fd.read().splitlines() url_names = get_url_hashes(lines) print('len of urls {}'.format(len(url_names))) f_abs = [((u + '.abs.') + suffix) for u in url_names] f_doc = [((u + '.doc.') + suffix) for u in url_names] l = len(f_abs) _l = len(f_doc) assert (l == _l) cnt = multiprocessing.cpu_count() pool = multiprocessing.Pool(processes=cnt) pool.starmap(_mv_from_to, zip(([src_path] * l), f_abs, ([os.path.join(tgt_path, split_name, 'abs')] * l))) pool.close() pool.join() pool = multiprocessing.Pool(processes=cnt) pool.starmap(_mv_from_to, zip(([src_path] * l), f_doc, ([os.path.join(tgt_path, split_name, 'doc')] * l))) pool.close() pool.join()
class ResnetBlock(nn.Module): def __init__(self, dim, dilation=1, use_spectral_norm=False): super(ResnetBlock, self).__init__() self.conv_block = nn.Sequential(nn.ReflectionPad2d(dilation), spectral_norm(nn.Conv2d(in_channels=dim, out_channels=dim, kernel_size=3, padding=0, dilation=dilation, bias=(not use_spectral_norm)), use_spectral_norm), nn.InstanceNorm2d(dim, track_running_stats=False), nn.ReLU(True), nn.ReflectionPad2d(1), spectral_norm(nn.Conv2d(in_channels=dim, out_channels=dim, kernel_size=3, padding=0, dilation=1, bias=(not use_spectral_norm)), use_spectral_norm), nn.InstanceNorm2d(dim, track_running_stats=False)) def forward(self, x): out = (x + self.conv_block(x)) return out
def exportfile(newAudio, time1, time2, filename, i): newAudio2 = newAudio[time1:time2] g = os.listdir() if ((((filename[0:(- 4)] + '_') + str(i)) + '.wav') in g): filename2 = ((str(uuid.uuid4()) + '_segment') + '.wav') print(('making %s' % filename2)) newAudio2.export(filename2, format='wav') else: filename2 = (str(uuid.uuid4()) + '.wav') print(('making %s' % filename2)) newAudio2.export(filename2, format='wav') return filename2
def label_file_from_coordinates(nifti_image, coord_list): imsh = list(np.array(nifti_image.dataobj).shape) label_array = np.zeros(tuple(imsh)) for j in range(len(coord_list)): label_array[(coord_list[j][0], coord_list[j][1], coord_list[j][2])] = 1 nib_pred = nib.Nifti1Image(dataobj=label_array, affine=nifti_image.header.get_best_affine()) return nib_pred
class BeitImageProcessor(BaseImageProcessor): model_input_names = ['pixel_values'] def __init__(self, do_resize: bool=True, size: Dict[(str, int)]=None, resample: PILImageResampling=PILImageResampling.BICUBIC, do_center_crop: bool=True, crop_size: Dict[(str, int)]=None, rescale_factor: Union[(int, float)]=(1 / 255), do_rescale: bool=True, do_normalize: bool=True, image_mean: Optional[Union[(float, List[float])]]=None, image_std: Optional[Union[(float, List[float])]]=None, do_reduce_labels: bool=False, **kwargs) -> None: if ('reduce_labels' in kwargs): warnings.warn('The `reduce_labels` parameter is deprecated and will be removed in a future version. Please use `do_reduce_labels` instead.', FutureWarning) do_reduce_labels = kwargs.pop('reduce_labels') super().__init__(**kwargs) size = (size if (size is not None) else {'height': 256, 'width': 256}) size = get_size_dict(size) crop_size = (crop_size if (crop_size is not None) else {'height': 224, 'width': 224}) crop_size = get_size_dict(crop_size, param_name='crop_size') self.do_resize = do_resize self.size = size self.resample = resample self.do_center_crop = do_center_crop self.crop_size = crop_size self.do_rescale = do_rescale self.rescale_factor = rescale_factor self.do_normalize = do_normalize self.image_mean = (image_mean if (image_mean is not None) else IMAGENET_STANDARD_MEAN) self.image_std = (image_std if (image_std is not None) else IMAGENET_STANDARD_STD) self.do_reduce_labels = do_reduce_labels def reduce_labels(self) -> bool: warnings.warn('The `reduce_labels` property is deprecated and will be removed in v4.27. Please use `do_reduce_labels` instead.', FutureWarning) return self.do_reduce_labels def from_dict(cls, image_processor_dict: Dict[(str, Any)], **kwargs): image_processor_dict = image_processor_dict.copy() if ('reduce_labels' in kwargs): image_processor_dict['reduce_labels'] = kwargs.pop('reduce_labels') return super().from_dict(image_processor_dict, **kwargs) def resize(self, image: np.ndarray, size: Dict[(str, int)], resample: PILImageResampling=PILImageResampling.BICUBIC, data_format: Optional[Union[(str, ChannelDimension)]]=None, **kwargs) -> np.ndarray: size = get_size_dict(size, default_to_square=True, param_name='size') if (('height' not in size) or ('width' not in size)): raise ValueError(f'The `size` argument must contain `height` and `width` keys. Got {size.keys()}') return resize(image, size=(size['height'], size['width']), resample=resample, data_format=data_format, **kwargs) def center_crop(self, image: np.ndarray, size: Dict[(str, int)], data_format: Optional[Union[(str, ChannelDimension)]]=None, **kwargs) -> np.ndarray: size = get_size_dict(size, default_to_square=True, param_name='size') return center_crop(image, size=(size['height'], size['width']), data_format=data_format, **kwargs) def rescale(self, image: np.ndarray, scale: Union[(int, float)], data_format: Optional[Union[(str, ChannelDimension)]]=None, **kwargs): return rescale(image, scale=scale, data_format=data_format, **kwargs) def normalize(self, image: np.ndarray, mean: Union[(float, List[float])], std: Union[(float, List[float])], data_format: Optional[Union[(str, ChannelDimension)]]=None, **kwargs) -> np.ndarray: return normalize(image, mean=mean, std=std, data_format=data_format, **kwargs) def reduce_label(self, label: ImageInput) -> np.ndarray: label = to_numpy_array(label) label[(label == 0)] = 255 label = (label - 1) label[(label == 254)] = 255 return label def _preprocess(self, image: ImageInput, do_reduce_labels: bool=None, do_resize: bool=None, size: Dict[(str, int)]=None, resample: PILImageResampling=None, do_center_crop: bool=None, crop_size: Dict[(str, int)]=None, do_rescale: bool=None, rescale_factor: float=None, do_normalize: bool=None, image_mean: Optional[Union[(float, List[float])]]=None, image_std: Optional[Union[(float, List[float])]]=None): if do_reduce_labels: image = self.reduce_label(image) if do_resize: image = self.resize(image=image, size=size, resample=resample) if do_center_crop: image = self.center_crop(image=image, size=crop_size) if do_rescale: image = self.rescale(image=image, scale=rescale_factor) if do_normalize: image = self.normalize(image=image, mean=image_mean, std=image_std) return image def _preprocess_image(self, image: ImageInput, do_resize: bool=None, size: Dict[(str, int)]=None, resample: PILImageResampling=None, do_center_crop: bool=None, crop_size: Dict[(str, int)]=None, do_rescale: bool=None, rescale_factor: float=None, do_normalize: bool=None, image_mean: Optional[Union[(float, List[float])]]=None, image_std: Optional[Union[(float, List[float])]]=None, data_format: Optional[Union[(str, ChannelDimension)]]=None) -> np.ndarray: image = to_numpy_array(image) image = self._preprocess(image, do_reduce_labels=False, do_resize=do_resize, size=size, resample=resample, do_center_crop=do_center_crop, crop_size=crop_size, do_rescale=do_rescale, rescale_factor=rescale_factor, do_normalize=do_normalize, image_mean=image_mean, image_std=image_std) if (data_format is not None): image = to_channel_dimension_format(image, data_format) return image def _preprocess_segmentation_map(self, segmentation_map: ImageInput, do_resize: bool=None, size: Dict[(str, int)]=None, resample: PILImageResampling=None, do_center_crop: bool=None, crop_size: Dict[(str, int)]=None, do_reduce_labels: bool=None): segmentation_map = to_numpy_array(segmentation_map) if (segmentation_map.ndim == 2): segmentation_map = segmentation_map[(None, ...)] added_dimension = True else: added_dimension = False segmentation_map = self._preprocess(image=segmentation_map, do_reduce_labels=do_reduce_labels, do_resize=do_resize, resample=resample, size=size, do_center_crop=do_center_crop, crop_size=crop_size, do_normalize=False, do_rescale=False) if added_dimension: segmentation_map = np.squeeze(segmentation_map, axis=0) segmentation_map = segmentation_map.astype(np.int64) return segmentation_map def __call__(self, images, segmentation_maps=None, **kwargs): return super().__call__(images, segmentation_maps=segmentation_maps, **kwargs) def preprocess(self, images: ImageInput, segmentation_maps: Optional[ImageInput]=None, do_resize: bool=None, size: Dict[(str, int)]=None, resample: PILImageResampling=None, do_center_crop: bool=None, crop_size: Dict[(str, int)]=None, do_rescale: bool=None, rescale_factor: float=None, do_normalize: bool=None, image_mean: Optional[Union[(float, List[float])]]=None, image_std: Optional[Union[(float, List[float])]]=None, do_reduce_labels: Optional[bool]=None, return_tensors: Optional[Union[(str, TensorType)]]=None, data_format: ChannelDimension=ChannelDimension.FIRST, **kwargs) -> PIL.Image.Image: do_resize = (do_resize if (do_resize is not None) else self.do_resize) size = (size if (size is not None) else self.size) size = get_size_dict(size, default_to_square=True, param_name='size') resample = (resample if (resample is not None) else self.resample) do_center_crop = (do_center_crop if (do_center_crop is not None) else self.do_center_crop) crop_size = (crop_size if (crop_size is not None) else self.crop_size) crop_size = get_size_dict(crop_size, default_to_square=True, param_name='crop_size') do_rescale = (do_rescale if (do_rescale is not None) else self.do_rescale) rescale_factor = (rescale_factor if (rescale_factor is not None) else self.rescale_factor) do_normalize = (do_normalize if (do_normalize is not None) else self.do_normalize) image_mean = (image_mean if (image_mean is not None) else self.image_mean) image_std = (image_std if (image_std is not None) else self.image_std) do_reduce_labels = (do_reduce_labels if (do_reduce_labels is not None) else self.do_reduce_labels) images = make_list_of_images(images) if (segmentation_maps is not None): segmentation_maps = make_list_of_images(segmentation_maps, expected_ndims=2) if (not valid_images(images)): raise ValueError('Invalid image type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.') if ((segmentation_maps is not None) and (not valid_images(segmentation_maps))): raise ValueError('Invalid segmentation map type. Must be of type PIL.Image.Image, numpy.ndarray, torch.Tensor, tf.Tensor or jax.ndarray.') if ((do_resize and (size is None)) or (resample is None)): raise ValueError('Size and resample must be specified if do_resize is True.') if (do_center_crop and (crop_size is None)): raise ValueError('Crop size must be specified if do_center_crop is True.') if (do_rescale and (rescale_factor is None)): raise ValueError('Rescale factor must be specified if do_rescale is True.') if (do_normalize and ((image_mean is None) or (image_std is None))): raise ValueError('Image mean and std must be specified if do_normalize is True.') images = [self._preprocess_image(image=img, do_resize=do_resize, do_center_crop=do_center_crop, do_rescale=do_rescale, do_normalize=do_normalize, resample=resample, size=size, rescale_factor=rescale_factor, crop_size=crop_size, image_mean=image_mean, image_std=image_std, data_format=data_format) for img in images] data = {'pixel_values': images} if (segmentation_maps is not None): segmentation_maps = [self._preprocess_segmentation_map(segmentation_map=segmentation_map, do_reduce_labels=do_reduce_labels, do_resize=do_resize, resample=resample, size=size, do_center_crop=do_center_crop, crop_size=crop_size) for segmentation_map in segmentation_maps] data['labels'] = segmentation_maps return BatchFeature(data=data, tensor_type=return_tensors) def post_process_semantic_segmentation(self, outputs, target_sizes: List[Tuple]=None): logits = outputs.logits if (target_sizes is not None): if (len(logits) != len(target_sizes)): raise ValueError('Make sure that you pass in as many target sizes as the batch dimension of the logits') if is_torch_tensor(target_sizes): target_sizes = target_sizes.numpy() semantic_segmentation = [] for idx in range(len(logits)): resized_logits = torch.nn.functional.interpolate(logits[idx].unsqueeze(dim=0), size=target_sizes[idx], mode='bilinear', align_corners=False) semantic_map = resized_logits[0].argmax(dim=0) semantic_segmentation.append(semantic_map) else: semantic_segmentation = logits.argmax(dim=1) semantic_segmentation = [semantic_segmentation[i] for i in range(semantic_segmentation.shape[0])] return semantic_segmentation
def test_sanitize_date_range_bad_start_dt() -> None: with pytest.raises(ValueError) as ex_info: sanitize_date_range('INVALID', '2020-06-06') assert (str(ex_info.value) == 'Incorrect data format, should be YYYY-MM-DD')
def extract_answers(solver, data_path): answers = [] best_choices = [] num_corrects = 0 with open(data_path, 'r') as f: for (idx, line) in enumerate(f): question = MultipleChoiceQuestion.from_jsonl_ours(line, idx) answer = solver.answer_question(question) answers.append(answer) best_choice = answer_to_selection(answer) if (best_choice.choice.label == question.answerKey): num_corrects += 1 print('accuracy is: {} / {} = {:.1f}%'.format(num_corrects, len(answers), ((num_corrects * 100.0) / len(answers)))) return (answers, best_choices)
def compare_files(gold_file, pred_file, up_ignore_layer=0): (gold_entity, pred_entity, match_entity) = get_matched_ner_from_file(gold_file, pred_file, up_ignore_layer) match_num = len(match_entity) gold_num = len(gold_entity) pred_num = len(pred_entity) return get_final_score(gold_num, pred_num, match_num)
def count_model_param_flops(model=None, dataset=None, multiply_adds=True, full=False): prods = {} def save_hook(name): def hook_per(self, input, output): prods[name] = np.prod(input[0].shape) return hook_per list_1 = [] def simple_hook(self, input, output): list_1.append(np.prod(input[0].shape)) list_2 = {} def simple_hook2(self, input, output): list_2['names'] = np.prod(input[0].shape) list_conv = [] def conv_hook(self, input, output): (batch_size, input_channels, input_height, input_width) = input[0].size() (output_channels, output_height, output_width) = output[0].size() kernel_ops = ((self.kernel_size[0] * self.kernel_size[1]) * (self.in_channels / self.groups)) bias_ops = (1 if (self.bias is not None) else 0) if (not full): num_weight_params = (self.weight.data != 0).float().sum() else: num_weight_params = torch.numel(self.weight.data) assert (self.weight.numel() == (kernel_ops * output_channels)), 'Not match' flops = (((((num_weight_params * (2 if multiply_adds else 1)) + (bias_ops * output_channels)) * output_height) * output_width) * batch_size) list_conv.append(flops) list_linear = [] def linear_hook(self, input, output): batch_size = (input[0].size(0) if (input[0].dim() == 2) else 1) if (not full): weight_ops = ((self.weight.data != 0).float().sum() * (2 if multiply_adds else 1)) bias_ops = ((self.bias.data != 0).float().sum() if (self.bias is not None) else 0) else: weight_ops = (torch.numel(self.weight.data) * (2 if multiply_adds else 1)) bias_ops = (torch.numel(self.bias.data) if (self.bias is not None) else 0) flops = (batch_size * (weight_ops + bias_ops)) list_linear.append(flops) list_bn = [] def bn_hook(self, input, output): list_bn.append((input[0].nelement() * 2)) list_relu = [] def relu_hook(self, input, output): list_relu.append(input[0].nelement()) list_pooling = [] def pooling_hook(self, input, output): (batch_size, input_channels, input_height, input_width) = input[0].size() (output_channels, output_height, output_width) = output[0].size() kernel_ops = (self.kernel_size * self.kernel_size) bias_ops = 0 params = 0 flops = (((((kernel_ops + bias_ops) * output_channels) * output_height) * output_width) * batch_size) list_pooling.append(flops) list_upsample = [] def upsample_hook(self, input, output): (batch_size, input_channels, input_height, input_width) = input[0].size() (output_channels, output_height, output_width) = output[0].size() flops = ((((output_height * output_width) * output_channels) * batch_size) * 12) list_upsample.append(flops) def foo(handles, net): childrens = list(net.children()) if (not childrens): if isinstance(net, torch.nn.Conv2d): handles += [net.register_forward_hook(conv_hook)] if isinstance(net, torch.nn.Linear): handles += [net.register_forward_hook(linear_hook)] return for c in childrens: foo(handles, c) handles = [] foo(handles, model) if (dataset == 'emnist'): input_channel = 1 input_res = 28 elif (dataset == 'cifar10'): input_channel = 3 input_res = 32 elif (dataset == 'cifar100'): input_channel = 3 input_res = 32 elif (dataset == 'tiny'): input_channel = 3 input_res = 64 device = next(model.parameters()).device input = Variable(torch.rand(input_channel, input_res, input_res).unsqueeze(0), requires_grad=True).to(device) out = model(input) total_flops = (((((sum(list_conv) + sum(list_linear)) + sum(list_bn)) + sum(list_relu)) + sum(list_pooling)) + sum(list_upsample)) for handle in handles: handle.remove() return total_flops
def classify(info, gold, test): coord_tags = ['CC'] info['classified_type'] = ('UNSET ' + info['type']) if value_present(info, ['type'], ['move']): if ('start left siblings' in info): if ((len(info['start left siblings']) > 0) and (info['start left siblings'][(- 1)] in coord_tags)): info['classified_type'] = 'Co-ordination' return if ('start right siblings' in info): if ((len(info['start right siblings']) > 0) and (info['start right siblings'][0] in coord_tags)): info['classified_type'] = 'Co-ordination' return if ('end left siblings' in info): if ((len(info['end left siblings']) > 0) and (info['end left siblings'][(- 1)] in coord_tags)): info['classified_type'] = 'Co-ordination' return if ('end right siblings' in info): if ((len(info['end right siblings']) > 0) and (info['end right siblings'][0] in coord_tags)): info['classified_type'] = 'Co-ordination' return if ('movers' in info): if ((len(info['movers']) > 0) and ((info['movers'][(- 1)] in coord_tags) or (info['movers'][0] in coord_tags))): info['classified_type'] = 'Co-ordination' return multi_case = False if ('movers' in info): if (len(info['movers']) > 1): multi_case = True for label in info['movers']: if (label not in phrase_labels): multi_case = False break if value_present(info, ['movers'], ['PP']): info['classified_type'] = 'PP Attachment' return if value_present(info, ['movers'], ['NP']): info['classified_type'] = 'NP Attachment' return if value_present(info, ['movers'], ['VP']): info['classified_type'] = 'VP Attachment' return if value_present(info, ['movers'], ['S', 'SINV', 'SBAR']): info['classified_type'] = 'Clause Attachment' return if value_present(info, ['movers'], ['RB', 'ADVP', 'ADJP']): info['classified_type'] = 'Modifier Attachment' return if value_present(info, ['old_parent'], ['NP', 'QP']): if value_present(info, ['new_parent'], ['NP', 'QP']): info['classified_type'] = 'NP Internal Structure' return if (('over_word' in info) and info['over_word']): info['classified_type'] = 'Single Word Phrase' return if value_present(info, ['type'], ['relabel']): info['classified_type'] = 'Different label' return if (info['type'] == 'add'): if (('subtrees' in info) and (len(info['subtrees']) == 1)): if (info['subtrees'][0] == info['label']): info['classified_type'] = 'XoverX Unary' return info['classified_type'] = 'Unary' return if (info['type'] == 'remove'): if (('family' in info) and (len(info['family']) == 1)): if (info['parent'] == info['label']): info['classified_type'] = 'XoverX Unary' return info['classified_type'] = 'Unary' return if (('subtrees' in info) and (len(info['subtrees']) == 1)): info['classified_type'] = 'Unary' return if value_present(info, ['label'], ['UCP']): info['classified_type'] = 'Co-ordination' return if ('right siblings' in info): if ((len(info['right siblings']) > 0) and (info['right siblings'][0] in coord_tags)): info['classified_type'] = 'Co-ordination' return if (('subtrees' in info) and ('PP' in info['subtrees'][1:])): info['classified_type'] = 'PP Attachment' return if ('subtrees' in info): if ('S' in info['subtrees'][1:]): info['classified_type'] = 'Clause Attachment' return if ('SBAR' in info['subtrees'][1:]): info['classified_type'] = 'Clause Attachment' return if ('SINV' in info['subtrees'][1:]): info['classified_type'] = 'Clause Attachment' return if value_present(info, ['parent'], ['NP']): all_words = True if ('subtrees' in info): for label in info['subtrees']: if (label in phrase_labels): all_words = False break if all_words: info['classified_type'] = 'NP Internal Structure' return if value_present(info, ['label'], ['ADVP', 'ADJP']): info['classified_type'] = 'Modifier Attachment' return if ('subtrees' in info): if (('ADVP' in info['subtrees'][1:]) or ('ADJP' in info['subtrees'][1:])): info['classified_type'] = 'Modifier Attachment' return if ('label' in info): label = info['label'] if ('subtrees' in info): all_same = True for slabel in info['subtrees']: if (slabel != label): all_same = False break if all_same: if (label == 'NP'): info['classified_type'] = 'NP Internal Structure' return else: info['classified_type'] = 'Co-ordination' return
('l2_side') class TFL2SideBoxRegularizer(TFBoxRegularizer): def __init__(self, weight: float, log_scale: bool=False, reduction: str='sum') -> None: super().__init__(weight, log_scale=log_scale, reduction=reduction) def _forward(self, box_tensor: TFBoxTensor) -> tf.Tensor: return tf_l2_side_regularizer(box_tensor, log_scale=self.log_scale)
def get_optimizer(model): parameters = _get_paramters(model) opt_lower = cfg.SOLVER.OPTIMIZER.lower() if (opt_lower == 'sgd'): optimizer = optim.SGD(parameters, lr=cfg.SOLVER.LR, momentum=cfg.SOLVER.MOMENTUM, weight_decay=cfg.SOLVER.WEIGHT_DECAY) elif (opt_lower == 'adam'): optimizer = optim.Adam(parameters, lr=cfg.SOLVER.LR, eps=cfg.SOLVER.EPSILON, weight_decay=cfg.SOLVER.WEIGHT_DECAY) elif (opt_lower == 'adadelta'): optimizer = optim.Adadelta(parameters, lr=cfg.SOLVER.LR, eps=cfg.SOLVER.EPSILON, weight_decay=cfg.SOLVER.WEIGHT_DECAY) elif (opt_lower == 'rmsprop'): optimizer = optim.RMSprop(parameters, lr=cfg.SOLVER.LR, alpha=0.9, eps=cfg.SOLVER.EPSILON, momentum=cfg.SOLVER.MOMENTUM, weight_decay=cfg.SOLVER.WEIGHT_DECAY) else: raise ValueError('Expected optimizer method in [sgd, adam, adadelta, rmsprop], but received {}'.format(opt_lower)) return optimizer
class ConvBlock(nn.Module): def __init__(self, in_channels, out_channels, kernel_size=3, add_bias=True, scale_factor=1, filtering_kernel=(1, 3, 3, 1), use_wscale=True, wscale_gain=_WSCALE_GAIN, lr_mul=1.0, activation_type='lrelu', minibatch_std_group_size=0, minibatch_std_channels=1): super().__init__() if (minibatch_std_group_size > 1): in_channels = (in_channels + minibatch_std_channels) self.mbstd = MiniBatchSTDLayer(group_size=minibatch_std_group_size, new_channels=minibatch_std_channels) else: self.mbstd = nn.Identity() if (scale_factor > 1): extra_padding = (kernel_size - scale_factor) self.filter = DownsamplingLayer(scale_factor=1, kernel=filtering_kernel, extra_padding=extra_padding) self.stride = scale_factor self.padding = 0 else: self.filter = nn.Identity() assert ((kernel_size % 2) == 1) self.stride = 1 self.padding = (kernel_size // 2) weight_shape = (out_channels, in_channels, kernel_size, kernel_size) fan_in = ((kernel_size * kernel_size) * in_channels) wscale = (wscale_gain / np.sqrt(fan_in)) if use_wscale: self.weight = nn.Parameter((torch.randn(*weight_shape) / lr_mul)) self.wscale = (wscale * lr_mul) else: self.weight = nn.Parameter(((torch.randn(*weight_shape) * wscale) / lr_mul)) self.wscale = lr_mul if add_bias: self.bias = nn.Parameter(torch.zeros(out_channels)) else: self.bias = None self.bscale = lr_mul if (activation_type == 'linear'): self.activate = nn.Identity() self.activate_scale = 1.0 elif (activation_type == 'lrelu'): self.activate = nn.LeakyReLU(negative_slope=0.2, inplace=True) self.activate_scale = np.sqrt(2.0) else: raise NotImplementedError(f'Not implemented activation function: `{activation_type}`!') def forward(self, x): x = self.mbstd(x) x = self.filter(x) weight = (self.weight * self.wscale) bias = ((self.bias * self.bscale) if (self.bias is not None) else None) x = F.conv2d(x, weight=weight, bias=bias, stride=self.stride, padding=self.padding) x = (self.activate(x) * self.activate_scale) return x
def test_simple_creation() -> None: tensor = tf.constant(np.random.rand(3, 2, 3)) box_tensor = TFBoxTensor(tensor) assert (tensor.numpy() == box_tensor.data.numpy()).all() assert isinstance(box_tensor, TFBoxTensor) tensor = tf.constant(np.random.rand(2, 10)) box_tensor = TFBoxTensor(tensor) assert (tensor.numpy() == box_tensor.data.numpy()).all() assert isinstance(box_tensor, TFBoxTensor)
def process_images(files): for file in tqdm(files, mininterval=10): if ('.jpg' in file): continue try: im = Image.open(file) process_one_img(file, im) except: print(file)
class StackedConvLayers(nn.Module): def __init__(self, input_feature_channels, output_feature_channels, num_convs, conv_op=nn.Conv2d, conv_kwargs=None, norm_op=nn.BatchNorm2d, norm_op_kwargs=None, dropout_op=nn.Dropout2d, dropout_op_kwargs=None, nonlin=nn.LeakyReLU, nonlin_kwargs=None, first_stride=None, basic_block=ConvDropoutNormNonlin): self.input_channels = input_feature_channels self.output_channels = output_feature_channels if (nonlin_kwargs is None): nonlin_kwargs = {'negative_slope': 0.01, 'inplace': True} if (dropout_op_kwargs is None): dropout_op_kwargs = {'p': 0.5, 'inplace': True} if (norm_op_kwargs is None): norm_op_kwargs = {'eps': 1e-05, 'affine': True, 'momentum': 0.1} if (conv_kwargs is None): conv_kwargs = {'kernel_size': 3, 'stride': 1, 'padding': 1, 'dilation': 1, 'bias': True} self.nonlin_kwargs = nonlin_kwargs self.nonlin = nonlin self.dropout_op = dropout_op self.dropout_op_kwargs = dropout_op_kwargs self.norm_op_kwargs = norm_op_kwargs self.conv_kwargs = conv_kwargs self.conv_op = conv_op self.norm_op = norm_op if (first_stride is not None): self.conv_kwargs_first_conv = deepcopy(conv_kwargs) self.conv_kwargs_first_conv['stride'] = first_stride else: self.conv_kwargs_first_conv = conv_kwargs super(StackedConvLayers, self).__init__() self.blocks = nn.Sequential(*([basic_block(input_feature_channels, output_feature_channels, self.conv_op, self.conv_kwargs_first_conv, self.norm_op, self.norm_op_kwargs, self.dropout_op, self.dropout_op_kwargs, self.nonlin, self.nonlin_kwargs)] + [basic_block(output_feature_channels, output_feature_channels, self.conv_op, self.conv_kwargs, self.norm_op, self.norm_op_kwargs, self.dropout_op, self.dropout_op_kwargs, self.nonlin, self.nonlin_kwargs) for _ in range((num_convs - 1))])) def forward(self, x): return self.blocks(x)
_module() class LCLDataset(MInstrDataset): def __init__(self, *args, **kwargs): super().__init__(*args, **kwargs, placeholders=(IMAGE_PLACEHOLDER, EXPR_PLACEHOLDER)) self.data = self._get_annos(self.filename) self.cls_neg_label = None self.cls_idx = None self.cls_name = None def _get_annos(self, filename): cls_metas = [] with jsonlines.open(filename) as reader: for metas in reader: cls_metas.append(metas) return cls_metas def get_raw_item(self, index): return self.data[index] def get_ret(self, image, question, answer, conv_mode=None): ret = {'image': image, 'conversations': [{'from': 'human', 'value': question}, {'from': 'gpt', 'value': f'{answer}'}]} if (conv_mode is not None): ret['mode'] = conv_mode return ret def get_samples(self, index, mode): raise NotImplementedError def __getitem__(self, index): raise NotImplementedError def __get_icl_item__(self, index, shot): raise NotImplementedError
def main(): parser = HfArgumentParser((ModelArguments, DataTrainingArguments, TrainingArguments)) (model_args, data_args, training_args) = parser.parse_args_into_dataclasses() configure_logger(model_args, training_args) datasets = load_dataset(data_args.dataset_name, data_args.dataset_config_name, cache_dir=model_args.cache_dir) if ('validation' not in datasets.keys()): datasets = DatasetDict() datasets['validation'] = load_dataset(data_args.dataset_name, data_args.dataset_config_name, split=f'{data_args.train_split_name}[:{data_args.validation_split_percentage}%]', cache_dir=model_args.cache_dir) datasets['train'] = load_dataset(data_args.dataset_name, data_args.dataset_config_name, split=f'{data_args.train_split_name}[{data_args.validation_split_percentage}%:]', cache_dir=model_args.cache_dir) else: datasets = DatasetDict() datasets['validation'] = load_dataset(data_args.dataset_name, data_args.dataset_config_name, split='validation', cache_dir=model_args.cache_dir) datasets['train'] = load_dataset(data_args.dataset_name, data_args.dataset_config_name, split=f'{data_args.train_split_name}', cache_dir=model_args.cache_dir) feature_extractor = Wav2Vec2FeatureExtractor.from_pretrained(model_args.model_name_or_path, cache_dir=model_args.cache_dir, do_normalize=True) def prepare_dataset(batch): (batch['speech'], _) = librosa.load(batch[data_args.speech_file_column], sr=feature_extractor.sampling_rate) return batch vectorized_datasets = datasets.map(prepare_dataset, num_proc=data_args.preprocessing_num_workers, remove_columns=datasets['train'].column_names) vectorized_datasets = vectorized_datasets.filter((lambda data: (len(data['speech']) < int((data_args.max_duration_in_seconds * feature_extractor.sampling_rate))))) def normalize(batch): return feature_extractor(batch['speech'], sampling_rate=feature_extractor.sampling_rate) vectorized_datasets = vectorized_datasets.map(normalize, batched=True, num_proc=data_args.preprocessing_num_workers, load_from_cache_file=(not data_args.overwrite_cache), remove_columns=vectorized_datasets['train'].column_names) config = Wav2Vec2Config.from_pretrained(model_args.model_name_or_path, cache_dir=model_args.cache_dir, gradient_checkpointing=training_args.gradient_checkpointing) if ((not config.do_stable_layer_norm) or (config.feat_extract_norm != 'layer')): raise ValueError("PreTraining is only supported for ``config.do_stable_layer_norm=True`` and ``config.feat_extract_norm='layer'") model = Wav2Vec2ForPreTraining(config) data_collator = DataCollatorForWav2Vec2Pretraining(model=model, feature_extractor=feature_extractor) trainer = Wav2Vec2PreTrainer(model=model, data_collator=data_collator, args=training_args, train_dataset=vectorized_datasets['train'], eval_dataset=vectorized_datasets['validation'], tokenizer=feature_extractor, max_gumbel_temp=model_args.max_gumbel_temperature, min_gumbel_temp=model_args.min_gumbel_temperature, gumbel_temp_decay=model_args.gumbel_temperature_decay) trainer.train()
def eye_like(x: Tensor, /) -> Tensor: ndim = x.ndim if (ndim < 2): raise ValueError(f'Input must have at least two dimensions! Got "{ndim}"') (n, n2) = (x.shape[(- 2)], x.shape[(- 1)]) if (n != n2): raise ValueError(f'Input last two dimensions must be square (*, n, n)! Got "{x.shape}"') view = (([1] * (ndim - 2)) + [n, n]) I = torch.eye(n, dtype=x.dtype, device=x.device).view(view).expand_as(x).clone() return I
class LibrispeechASR(datasets.GeneratorBasedBuilder): DEFAULT_WRITER_BATCH_SIZE = 256 DEFAULT_CONFIG_NAME = 'all' BUILDER_CONFIGS = [LibrispeechASRConfig(name='clean', description="'Clean' speech."), LibrispeechASRConfig(name='other', description="'Other', more challenging, speech."), LibrispeechASRConfig(name='all', description='Combined clean and other dataset.')] def _info(self): return datasets.DatasetInfo(description=_DESCRIPTION, features=datasets.Features({'file': datasets.Value('string'), 'audio': datasets.Audio(sampling_rate=16000), 'text': datasets.Value('string'), 'speaker_id': datasets.Value('int64'), 'chapter_id': datasets.Value('int64'), 'id': datasets.Value('string')}), supervised_keys=('file', 'text'), homepage=_URL, citation=_CITATION, task_templates=[AutomaticSpeechRecognition(audio_column='audio', transcription_column='text')]) def _split_generators(self, dl_manager): archive_path = dl_manager.download(_DL_URLS[self.config.name]) local_extracted_archive = (dl_manager.extract(archive_path) if (not dl_manager.is_streaming) else {}) if (self.config.name == 'clean'): train_splits = [datasets.SplitGenerator(name='train.100', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('train.100'), 'files': dl_manager.iter_archive(archive_path['train.100'])}), datasets.SplitGenerator(name='train.360', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('train.360'), 'files': dl_manager.iter_archive(archive_path['train.360'])})] dev_splits = [datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={'local_extracted_archive': local_extracted_archive.get('dev'), 'files': dl_manager.iter_archive(archive_path['dev'])})] test_splits = [datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={'local_extracted_archive': local_extracted_archive.get('test'), 'files': dl_manager.iter_archive(archive_path['test'])})] elif (self.config.name == 'other'): train_splits = [datasets.SplitGenerator(name='train.500', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('train.500'), 'files': dl_manager.iter_archive(archive_path['train.500'])})] dev_splits = [datasets.SplitGenerator(name=datasets.Split.VALIDATION, gen_kwargs={'local_extracted_archive': local_extracted_archive.get('dev'), 'files': dl_manager.iter_archive(archive_path['dev'])})] test_splits = [datasets.SplitGenerator(name=datasets.Split.TEST, gen_kwargs={'local_extracted_archive': local_extracted_archive.get('test'), 'files': dl_manager.iter_archive(archive_path['test'])})] elif (self.config.name == 'all'): dev_splits = [datasets.SplitGenerator(name='validation.clean', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('dev.clean'), 'files': dl_manager.iter_archive(archive_path['dev.clean'])}), datasets.SplitGenerator(name='validation.other', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('dev.other'), 'files': dl_manager.iter_archive(archive_path['dev.other'])})] test_splits = [datasets.SplitGenerator(name='test.clean', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('test.clean'), 'files': dl_manager.iter_archive(archive_path['test.clean'])}), datasets.SplitGenerator(name='test.other', gen_kwargs={'local_extracted_archive': local_extracted_archive.get('test.other'), 'files': dl_manager.iter_archive(archive_path['test.other'])})] return ((train_splits + dev_splits) + test_splits) def _generate_examples(self, files, local_extracted_archive): key = 0 audio_data = {} transcripts = [] for (path, f) in files: if path.endswith('.flac'): id_ = path.split('/')[(- 1)][:(- len('.flac'))] audio_data[id_] = f.read() elif path.endswith('.trans.txt'): for line in f: if line: line = line.decode('utf-8').strip() (id_, transcript) = line.split(' ', 1) audio_file = f'{id_}.flac' (speaker_id, chapter_id) = [int(el) for el in id_.split('-')[:2]] audio_file = (os.path.join(local_extracted_archive, audio_file) if local_extracted_archive else audio_file) transcripts.append({'id': id_, 'speaker_id': speaker_id, 'chapter_id': chapter_id, 'file': audio_file, 'text': transcript}) if (audio_data and (len(audio_data) == len(transcripts))): for transcript in transcripts: audio = {'path': transcript['file'], 'bytes': audio_data[transcript['id']]} (yield (key, {'audio': audio, **transcript})) key += 1 audio_data = {} transcripts = []