paddle复现pytorch踩坑(二）：paddleAPI对照表（百度论文复现营活动提供）
2024-04-15 13:51:52
百度提供了paddlepaddle对应pytorch的API对照表
环境：paddlepaddle1.8
百度论文复现营活动：论文复现营
| Pytorch的API名称 | Paddle的API名称 |
|---|---|
| torch.abs | fluid.layers.abs(x, name=None) |
| torch.acos | fluid.layers.acos(x, name=None) |
| torch.add | fluid.layers.elementwise_add(x, y, axis=-1, act=None, name=None) |
| torch.allclose | fluid.layers.allclose |
| torch.arange | fluid.layers.range(start, end, step, dtype) |
| torch.argmax | fluid.layers.argmax |
| torch.argmin | fluid.layers.argmin |
| torch.argsort | fluid.layers.argsort(input, axis=-1, descending=False, name=None) |
| torch.as_strided | fluid.layers.strided_slice(input, axes, starts, ends, strides): |
| torch.asin | fluid.layers.asin(x, name=None) |
| torch.atan | fluid.layers.atan(x, name=None) |
| torch.cat | fluid.layers.concat(input, axis=0, name=None) |
| torch.ceil | fluid.layers.ceil |
| torch.chunk | fluid.layers.unstack(x, axis=0, num=None) |
| torch.cos | fluid.layers.cos |
| torch.cumsum | fluid.layers.cumsum(x, axis=None, exclusive=None, reverse=None) |
| torch.diag | fluid.layer.diag(diagonal) |
| torch.diag_embed | fluid.layer.diag_embed |
| torch.div | fluid.layers.elementwise_div(x, y, axis=-1, act=None, name=None) |
| torch.eq | fluid.layers.equal(x, y, cond=None) |
| torch.equal | fluid.layers.elementwise_equal(x, y, name=None) |
| torch.equal | fluid.layers.equal(x, y, cond=None) |
| torch.erf | fluid.layers.erf(x) |
| torch.exp | fluid.layers.exp(x, name=None) |
| torch.eye | fluid.layers.eye(num_rows, num_columns=None, batch_shape=None, dtype=‘float32’) |
| torch.flatten | fluid.layers.flatten |
| torch.flip | fluid.layers.flip |
| torch.floor | fluid.layers.floor(x, name=None) |
| torch.fmod | fluid.layers.elementwise_mod(x, y, axis=-1, act=None, name=None) |
| torch.from_numpy | fluid.dygraph.to_variable(value, block=None, name=None) |
| torch.full | fluid.layers.full(shape, fill_value, out=None, dtype=None, device=None, stop_gradient=True, name=None) |
| torch.full_like | fluid.layers.full_like |
| torch.gather | fluid.layers.gather(input, axis, index, out=None, sparse_grad=False, name=None) |
| torch.ge | fluid.layers.greater_equal |
| torch.gt | fluid.layers.greater_than |
| torch.le | fluid.layers.less_equal |
| torch.linspace | fluid.layers.linspace(start, stop, num, dtype) |
| torch.log | fluid.layers.log(x, name=None) |
| torch.logical_not | fluid.layers.logical_not(x, out=None, name=None) |
| torch.logical_xor | fluid.layers.logical_xor(x, y, out=None, name=None) |
| torch.lt | fluid.layers.less_than |
| torch.masked_select | fluid.layers.masked_select |
| torch.matmul | fluid.layers.matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None) |
| torch.max | fluid.layers.max(input, axis=None, keepdim=False, out=None, name=None) -> (Tensor, LongTensor) |
| torch.mean | fluid.layers.reduce_mean(input, dim=None, keep_dim=False, name=None) |
| torch.min | fluid.layers.min |
| torch.mm | fluid.layers.mm |
| torch.mul | fluid.layers.mul |
| torch.mv | fluid.layers.matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None) |
| torch.ne | fluid.layers.not_equal |
| torch.nn.AdaptiveAvgPool2d | fluid.layers.adaptive_pool2d |
| torch.nn.AdaptiveAvgPool3d | fluid.layers.adaptive_pool3d |
| torch.nn.AdaptiveMaxPool2d | fluid.layers.adaptive_pool2d |
| torch.nn.AdaptiveMaxPool3d | fluid.layers.adaptive_pool3d |
| torch.nn.add_module | fluid.dygraph.Layer.add_sublayer(name, sublayer) |
| torch.nn.append | fluid.dygraph.LayerList.append(sublayer) |
| torch.nn.append | fluid.dygraph.ParameterList.append(parameter) |
| torch.nn.AvgPool2d | fluid.layers.pool2d(pool_type为’avg’) |
| torch.nn.AvgPool3d | fluid.layers.pool3d(pool_type为’avg’) |
| torch.nn.BatchNorm1d | fluid.layers.batch_norm(input, act=None, is_test=False, momentum=0.9, epsilon=1e-05, param_attr=None, bias_attr=None, data_layout=‘NCHW’, in_place=False, name=None, moving_mean_name=None, moving_variance_name=None, do_model_average_for_mean_and_var=False, use_global_stats=False) |
| torch.nn.BatchNorm2d | fluid.layers.batch_norm(input, act=None, is_test=False, momentum=0.9, epsilon=1e-05, param_attr=None, bias_attr=None, data_layout=‘NCHW’, in_place=False, name=None, moving_mean_name=None, moving_variance_name=None, do_model_average_for_mean_and_var=False, use_global_stats=False) |
| torch.nn.BatchNorm3d | fluid.layers.batch_norm(input, act=None, is_test=False, momentum=0.9, epsilon=1e-05, param_attr=None, bias_attr=None, data_layout=‘NCHW’, in_place=False, name=None, moving_mean_name=None, moving_variance_name=None, do_model_average_for_mean_and_var=False, use_global_stats=False) |
| torch.nn.Bilinear | fluid.layers.bilinear_tensor_product(x, y, size, act=None, name=None, param_attr=None, bias_attr=None) |
| torch.nn.ConstantPad1d | fluid.layers.pad |
| torch.nn.ConstantPad2d | fluid.layers.pad / fluid.layers.pad2d |
| torch.nn.ConstantPad3d | fluid.layers.pad |
| torch.nn.Conv2d | fluid.dygraph.Conv2D |
| torch.nn.Conv3d | fluid.dygraph.Conv3D |
| torch.nn.ConvTranspose2d | fluid.layers.conv2d_transpose |
| torch.nn.ConvTranspose3d | fluid.layers.conv3d_transpose |
| torch.nn.double | fluid.layers.cast(x, dtype) |
| torch.nn.Dropout | fluid.layers.dropout(x, dropout_prob, is_test=False, seed=None, name=None, dropout_implementation=‘downgrade_in_infer’) |
| torch.nn.ELU | fluid.layers.elu(x, alpha=1.0, name=None) |
| torch.nn.Embedding | fluid.dygraph.Embedding |
| torch.nn.eval | fluid.dygraph.Layer.eval() |
| torch.nn.float | fluid.layers.cast(x, dtype) |
| torch.nn.Fold | fluid.layers.unfold |
| torch.nn.functional.adaptive_avg_pool2d | fluid.layers.adaptive_pool2d |
| torch.nn.functional.adaptive_avg_pool3d | fluid.layers.adaptive_pool3d |
| torch.nn.functional.adaptive_max_pool2d | fluid.layers.adaptive_pool2d |
| torch.nn.functional.adaptive_max_pool3d | fluid.layers.adaptive_pool3d |
| torch.nn.functional.affine_grid | fluid.layers.affine_grid |
| torch.nn.functional.avg_pool2d | fluid.layers.pool2d(input, pool_size=-1, pool_type=“max”, pool_stride=1, pool_padding=0, global_pooling=False, use_cudnn=True, ceil_mode=False, name=None, exclusive=True, data_format=“NCHW”)设置pool_type=“avg” |
| torch.nn.functional.avg_pool3d | fluid.layers.pool3d(input, pool_size=-1, pool_type=“max”, pool_stride=1, pool_padding=0, global_pooling=False, use_cudnn=True, ceil_mode=False, name=None, exclusive=True, data_format=“NCDHW”)设置pool_type=“avg” |
| torch.nn.functional.cross_entropy | fluid.layers.softmax_with_cross_entropy |
| torch.nn.functional.ctc_loss | fluid.layers.warpctc |
| torch.nn.functional.elu | fluid.layers.elu |
| torch.nn.functional.embedding | fluid.embedding |
| torch.nn.functional.embedding_bag | fluid.contrib.fused_embedding_seq_pool |
| torch.nn.functional.glu | fluid.nets.glu(input, dim=-1) |
| torch.nn.functional.grid_sample | fluid.layers.grid_sampler |
| torch.nn.functional.hardshrink | fluid.layers.hard_shrink |
| torch.nn.functional.interpolate | fluid.layers.interpolate |
| torch.nn.functional.kl_div | fluid.layers.kldiv_loss |
| torch.nn.functional.leaky_relu | fluid.layers.leaky_relu |
| torch.nn.functional.logsigmoid | fluid.layers.logsigmoid |
| torch.nn.functional.margin_ranking_loss | fluid.layers.margin_rank_loss(label, left, right, margin=0.1, name=None) |
| torch.nn.functional.max_pool2d | fluid.layers.pool2d(input, pool_size=-1, pool_type=“max”, pool_stride=1, pool_padding=0, global_pooling=False, use_cudnn=True, ceil_mode=False, name=None, exclusive=True, data_format=“NCHW”)设置pool_type=“max” |
| torch.nn.functional.max_pool3d | fluid.layers.pool3d(input, pool_size=-1, pool_type=“max”, pool_stride=1, pool_padding=0, global_pooling=False, use_cudnn=True, ceil_mode=False, name=None, exclusive=True, data_format=“NCDHW”)设置pool_type=“max” |
| torch.nn.functional.mse_loss | fluid.layers.mse_loss(input, label) |
| torch.nn.functional.one_hot | fluid.one_hot |
| torch.nn.functional.pad | fluid.layers.pad(x, paddings, pad_value=0.0, name=None) |
| torch.nn.functional.pixel_shuffle | fluid.layers.pixel_shuffle(x, upscale_factor) |
| torch.nn.functional.prelu | fluid.layers.prelu(x, mode, param_attr=None, name=None) |
| torch.nn.functional.relu | fluid.layers.relu(x, name=None) |
| torch.nn.functional.relu6 | fluid.layers.relu6 |
| torch.nn.functional.selu | fluid.layers.selu |
| torch.nn.functional.sigmoid | fluid.layers.sigmoid |
| torch.nn.functional.smooth_l1_loss | fluid.layers.smooth_l1(x, y, inside_weight=None, outside_weight=None, sigma=None) |
| torch.nn.functional.softmax | fluid.layers.softmax |
| torch.nn.functional.softplus | fluid.layers.softplus |
| torch.nn.functional.softshrink | fluid.layers.softshrink |
| torch.nn.functional.softsign | fluid.layers.softsign |
| torch.nn.functional.tanh | fluid.layers.tanh |
| torch.nn.functional.tanhshrink | fluid.layers.tanh_shrink |
| torch.nn.functional.threshold | fluid.layers.thresholded_relu |
| torch.nn.functional.unfold | fluid.layers.unfold(x, kernel_size, strides=1, paddings=0, dilation=1, name=None) |
| torch.nn.GroupNorm | fluid.layers.group_norm(input, groups, epsilon=1e-05, param_attr=None, bias_attr=None, act=None, data_layout=‘NCHW’, name=None) |
| torch.nn.GRU | fluid.layers.dynamic_gru |
| torch.nn.GRUCell | fluid.layers.GRUCell |
| torch.nn.half | fluid.layers.cast(x, dtype) |
| torch.nn.Hardshrink | fluid.layers.hard_shrink(x, threshold=None) |
| torch.nn.init.constant_ | fluid.initializer.ConstantInitializer(value=0.0, force_cpu=False) |
| torch.nn.init.eye_ | fluid.layers.eye(num_rows, num_columns=None, batch_shape=None, dtype=‘float32’) |
| torch.nn.init.kaiming_normal_ | fluid.initializer.MSRAInitializer(uniform=True, fan_in=None, seed=0) |
| torch.nn.init.kaiming_uniform_ | fluid.initializer.MSRAInitializer(uniform=True, fan_in=None, seed=0) |
| torch.nn.init.normal_ | fluid.initializer.NormalInitializer(loc=0.0, scale=1.0, seed=0) |
| torch.nn.init.ones_ | fluid.layers.ones(shape, dtype, force_cpu=False) |
| torch.nn.init.uniform_ | fluid.initializer.UniformInitializer(low=-1.0, high=1.0, seed=0) |
| torch.nn.init.xavier_normal_ | fluid.initializer.XavierInitializer(uniform=True, fan_in=None, fan_out=None, seed=0 |
| torch.nn.init.xavier_uniform_ | fluid.initializer.XavierInitializer(uniform=True, fan_in=None, fan_out=None, seed=0 |
| torch.nn.init.zeros_ | fluid.layers.zeros(shape, dtype, force_cpu=False) |
| torch.nn.InstanceNorm1d | fluid.layers.instance_norm |
| torch.nn.InstanceNorm2d | fluid.layers.instance_norm |
| torch.nn.InstanceNorm3d | fluid.layers.instance_norm |
| torch.nn.LayerNorm | fluid.layers.layer_norm(input, scale=True, shift=True, begin_norm_axis=1, epsilon=1e-05, param_attr=None, bias_attr=None, act=None, name=None) |
| torch.nn.LeakyReLU | fluid.layers.leaky_relu(x, alpha=0.02, name=None) |
| torch.nn.Linear | fluid.dygraph.Linear(input_dim, output_dim, param_attr=None, bias_attr=None, act=None, dtype=‘float32’) |
| torch.nn.load_state_dict | fluid.dygraph.Layer.set_dict(stat_dict, include_sublayers=True) |
| torch.nn.LogSigmoid | fluid.layers.logsigmoid(x, name=None) |
| torch.nn.LSTM | fluid.layers.lstm |
| torch.nn.LSTMCell | fluid.layers.LSTMCell |
| torch.nn.MarginRankingLoss | fluid.layers.margin_rank_loss(label, left, right, margin=0.1, name=None) |
| torch.nn.MaxPool2d | fluid.layers.pool2d(pool_type为’max’) |
| torch.nn.MaxPool3d | fluid.layers.pool3d(pool_type为’max’) |
| torch.nn.Module | fluid.dygraph.Layer() |
| torch.nn.modules | fluid.dygraph.Layer.sublayers(include_sublayers=True) |
| torch.nn.MSELoss | fluid.dygraph.MSELoss(input, label) |
| torch.nn.MultiheadAttention | fluid.nets.scaled_dot_product_attention |
| torch.nn.named_modules | fluid.dygraph.Layer.named_sublayers(prefix=’’,include_sublayers=True,include_self=False,layers_set=None) |
| torch.nn.named_parameters | fluid.dygraph.Layer.named_parameters(prefix=’’, include_sublayers=True) |
| torch.nn.ParameterList | fluid.dygraph.ParameterList |
| torch.nn.parameters | fluid.dygraph.Layer.parameters(include_sublayers=True) |
| torch.nn.PixelShuffle | fluid.layers.pixel_shuffle |
| torch.nn.PReLU | fluid.dygraph.Prelu |
| torch.nn.ReflectionPad2d | fluid.layers.pad2d |
| torch.nn.register_parameter | fluid.dygraph.Layer.add_parameter(name, parameter) |
| torch.nn.ReLU | fluid.layers.relu(x, name=None) |
| torch.nn.ReLU6 | fluid.layers.relu6 |
| torch.nn.ReplicationPad2d | fluid.layers.pad2d |
| torch.nn.requires_grad_ | fluid.Variable().stop_gradient |
| torch.nn.RNN | fluid.layers.rnn |
| torch.nn.RNNCell | fluid.layers.RNNCell |
| torch.nn.SELU | fluid.layers.selu |
| torch.nn.Sequential | fluid.dygraph.Sequential(*layers) |
| torch.nn.Sigmoid | fluid.layers.sigmoid(x, name=None) |
| torch.nn.SmoothL1Loss | fluid.layers.smooth_l1(x, y, inside_weight=None, outside_weight=None, sigma=None) |
| torch.nn.Softmax | fluid.layers.softmax(input, use_cudnn=False, name=None, axis=-1) |
| torch.nn.Softplus | fluid.layers.softplus |
| torch.nn.Softshrink | fluid.layers.softshrink(x, alpha=None) |
| torch.nn.Softsign | fluid.layers.softsign(x, name=None) |
| torch.nn.state_dict | fluid.dygraph.Layer.state_dict |
| torch.nn.Tanh | fluid.layers.tanh(x, name=None) |
| torch.nn.Tanhshrink | fluid.layers.tanh_shrin |
| torch.nn.Threshold | fluid.layers.thresholded_relu(x, threshold=None) |
| torch.nn.train | fluid.dygraph.Layer.train() |
| torch.nn.type | fluid.layers.cast(x, dtype) |
| torch.nn.UpsamplingBilinear2d | fluid.layers.resize_bilinear |
| torch.nn.UpsamplingNearest2d | fluid.layers.resize_nearest |
| torch.nn.zero_grad | fluid.dygraph.Layer.clear_gradients() |
| torch.nn.ZeroPad2d | fluid.layers.pad / fluid.layers.pad2d |
| torch.norm | fluid.layers.l2_normalize |
| torch.normal | fluid.layers.gaussian_random(shape, mean=0.0, std=1.0, seed=0, dtype=‘float32’) |
| torch.numel | fluid.layers.size(input) |
| torch.ones | fluid.layers.ones |
| torch.ones_like | fluid.layers.ones_like |
| torch.pow | fluid.layers.pow |
| torch.prod | fluid.layers.reduce_prod(input, dim=None, keep_dim=False, name=None) |
| torch.rand_like | fluid.layers.gaussian_random_batch_size_like(input, shape, input_dim_idx=0, output_dim_idx=0, mean=0.0, std=1.0, seed=0, dtype=‘float32’)[ |
| torch.randint | fluid.layers.randint(low, high=None, shape=None, out=None, dtype=None, device=None, stop_gradient=False, seed=0, name=None) |
| torch.randn | fluid.layers.randn |
| torch.randn_like | fluid.layers.gaussian_random_batch_size_like |
| torch.randperm | fluid.layers.randperm |
| torch.range | fluid.layers.range |
| torch.reciprocal | fluid.layers.reciprocal(x, name=None) |
| torch.reshape | fluid.layers.reshape(x, shape, actual_shape=None, act=None, inplace=False, name=None) |
| torch.roll | fluid.layers.roll |
| torch.round | fluid.layers.round(x, name=None) |
| torch.rsqrt | fluid.layers.rsqrt(x, name=None) |
| torch.seed | fluid.Program.random_seed |
| torch.set_num_interop_threads | fluid.ExecutionStrategy.num_threads |
| torch.set_num_threads | fluid.cpu_places(device_count=None) |
| torch.sigmoid | fluid.layers.sigmoid(x, name=None) |
| torch.sign | fluid.layers.sign(x) |
| torch.sin | fluid.layers.sin(x, name=None) |
| torch.split | fluid.layers.split(input, num_or_sections, dim=-1, name=None) |
| torch.sqrt | fluid.layers.squeeze(input, axes, name=None) |
| torch.squeeze | fluid.layers.squeeze(input, axes, name=None) |
| torch.stack | fluid.layers.stack(x, axis=0) |
| torch.std_mean | fluid.layers.mean |
| torch.sum | fluid.layers.sum(x) |
| torch.sum | fluid.layers.elementwise_add(x, y, axis=-1, act=None, name=None) |
| torch.tan | fluid.layers.tanh(x, name=None) |
| torch.tanh | fluid.layers.tanh(x, name=None) |
| torch.Tensor.abs | fluid.layers.abs(x, name=None) |
| torch.Tensor.abs_ | fluid.layers.abs(x, name=None) |
| torch.Tensor.acos | fluid.layers.acos(x, name=None) |
| torch.Tensor.acos_ | fluid.layers.acos(x, name=None) |
| torch.Tensor.add_ | fluid.layers.elementwise_add(x, y, axis=-1, act=None, name=None) |
| torch.Tensor.asin_ | fluid.layers.asin(x, name=None) |
| torch.Tensor.atan_ | fluid.layers.atan(x, name=None) |
| torch.Tensor.backward | fluid.gradients(targets, inputs, target_gradients=None, no_grad_set=None)paddle 目标变量和起始 |
| torch.Tensor.bfloat16 | fluid.layers.cast(x, float16) |
| torch.Tensor.bmm | fluid.layers.matmul |
| torch.Tensor.bool | fluid.layers.cast(x, bool) |
| torch.Tensor.byte | fluid.layers.cast(x, uint8) |
| torch.Tensor.ceil | fluid.layers.ceil |
| torch.Tensor.clamp | fluid.layers.clip |
| torch.Tensor.clamp_ | fluid.layers.clip(x, min, max, name=None) |
| torch.Tensor.clone | fluid.layers.assign(input, output=None) |
| torch.Tensor.cos | fluid.layers.cos |
| torch.Tensor.cumsum | fluid.layers.cumsum |
| torch.Tensor.dequantize | fluid.layers.dequantize() |
| torch.Tensor.diag | fluid.layers.diag |
| torch.Tensor.div | fluid.layers.elementwise_div(x, y, axis=-1, act=None, name=None) |
| torch.Tensor.eq | fluid.layers.equal |
| torch.Tensor.eq_ | fluid.layers.equal(x, y, cond=None) |
| torch.Tensor.exp | fluid.layers.exp(x, name=None) |
| torch.Tensor.exp_ | fluid.layers.exp(x, name=None) |
| torch.Tensor.expand | fluid.layers.expand(x, expand_times, name=None) |
| torch.Tensor.expand_as | fluid.layers.expand_as(x, target_tensor, name=None) |
| torch.Tensor.fill_diagonal_ | fluid.layers.diag(diagonal) |
| torch.Tensor.flatten | fluid.layers.flatten |
| torch.Tensor.flip | fluid.layers.reverse |
| torch.Tensor.floor | fluid.layers.floor |
| torch.Tensor.floor_ | fluid.layers.floor(x, name=None) |
| torch.Tensor.fmod_ | fluid.layers.elementwise_mod(x, y, axis=-1, act=None, name=None) |
| torch.Tensor.gather | fluid.layers.gather(input, index, overwrite=True) |
| torch.Tensor.ge | fluid.layers.greater_equal(x, y, cond=None) |
| torch.Tensor.ge_ | fluid.layers.greater_equal(x, y, cond=None) |
| torch.Tensor.gt | fluid.layers.greater_than(x, y, cond=None) |
| torch.Tensor.gt_ | fluid.layers.greater_than(x, y, cond=None) |
| torch.Tensor.hardshrink | fluid.layers.hard_shrink(x, threshold=None) |
| torch.Tensor.index_select | fluid.layers.multiplex(inputs, index) |
| torch.Tensor.int | fluid.layers.cast |
| torch.Tensor.le | fluid.layers.less_equal(x, y, cond=None) |
| torch.Tensor.le_ | fluid.layers.less_equal(x, y, cond=None) |
| torch.Tensor.log | fluid.layers.log(x, name=None) |
| torch.Tensor.log_ | fluid.layers.log(x, name=None) |
| torch.Tensor.logical_not | fluid.layers.logical_not(x, out=None, name=None) |
| torch.Tensor.logical_not_ | fluid.layers.logical_not(x, out=None, name=None) |
| torch.Tensor.logical_xor | fluid.layers.logical_xor(x, y, out=None, name=None) |
| torch.Tensor.logical_xor_ | fluid.layers.logical_xor(x, y, out=None, name=None) |
| torch.Tensor.lt | fluid.layers.less_than(x, y, cond=None) |
| torch.Tensor.lt_ | fluid.layers.less_than(x, y, force_cpu=None, cond=None) |
| torch.Tensor.masked_select | fluid.layers.masked_select(input, mask) |
| torch.Tensor.matmul | fluid.layers.matmul(x, y, transpose_x=False, transpose_y=False, alpha=1.0, name=None) |
| torch.Tensor.matrix_power | fluid.layers.pow(x, factor=1.0, name=None) |
| torch.Tensor.max | fluid.layers.reduce_max(input, dim=None, keep_dim=False, name=None) |
| torch.Tensor.mean | fluid.layers.mean(x, name=None) |
| torch.Tensor.min | "fluid.layers.reduce_min(input, dim=None, keep_dim=False, name=None) " |
| torch.Tensor.mul | fluid.layers.mul(x, y, x_num_col_dims=1, y_num_col_dims=1, name=None) |
| torch.Tensor.mul_ | fluid.layers.mul(x, y, x_num_col_dims=1, y_num_col_dims=1, name=None) |
| torch.Tensor.multinomial | fluid.layers.sampling_id(x, min=0.0, max=1.0, seed=0, dtype=‘float32’) |
| torch.Tensor.mv | fluid.layers.mul(x, y, x_num_col_dims=1, y_num_col_dims=1, name=None) |
| torch.Tensor.narrow_copy | fluid.layers.slice(input, axes, starts, ends) |
| torch.Tensor.ndimension | fluid.layers.rank(input) |
| torch.Tensor.ne_ | fluid.layers.not_equal(x, y, cond=None) |
| torch.Tensor.normal_ | fluid.layers.Normal(loc, scale) |
| torch.Tensor.numel | fluid.layers.size(input) |
| torch.Tensor.permute | fluid.layers.transpose(x, perm, name=None) |
| torch.Tensor.pow | fluid.layers.pow(x, factor=1.0, name=None) |
| torch.Tensor.pow_ | fluid.layers.pow(x, factor=1.0, name=None) |
| torch.Tensor.prod | fluid.layers.reduce_prod(input, dim=None, keep_dim=False, name=None)[ |
| torch.Tensor.put_ | fluid.layers.scatter(input, index, updates, name=None, overwrite=True)[源代码] |
| torch.Tensor.random_ | fluid.layers.uniform_random(shape, dtype=‘float32’, min=-1.0, max=1.0, seed=0) |
| torch.Tensor.reciprocal | fluid.layers.reciprocal(x, name=None) |
| torch.Tensor.reciprocal_ | fluid.layers.reciprocal(x, name=None) |
| torch.Tensor.remainder_ | fluid.layers.elementwise_mod(x, y, axis=-1, act=None, name=None) |
| torch.Tensor.requires_grad | fluid.Variable.stop_gradient(属性) |
| torch.Tensor.requires_grad_ | fluid.Variable.stop_gradient(属性) |
| torch.Tensor.reshape | fluid.layers.reshape(x, shape, actual_shape=None, act=None, inplace=False, name=None) |
| torch.Tensor.reshape_as | fluid.layers.reshape(x, shape, actual_shape=None, act=None, inplace=False, name=None) |
| torch.Tensor.resize_as_ | fluid.layers.reshape(x, shape, actual_shape=None, act=None, inplace=False, name=None) |
| torch.Tensor.round | fluid.layers.round(x, name=None) |
| torch.Tensor.round_ | fluid.layers.round(x, name=None) |
| torch.Tensor.rsqrt | fluid.layers.rsqrt(x, name=None) |
| torch.Tensor.rsqrt_ | fluid.layers.rsqrt(x, name=None) |
| torch.Tensor.scatter | fluid.layers.scatter(input, index, updates, name=None, overwrite=True) |
| torch.Tensor.scatter_ | fluid.layers.scatter(input, index, updates, name=None, overwrite=True) |
| torch.Tensor.scatter_add | fluid.layers.scatter(input, index, updates, name=None, overwrite=True) |
| torch.Tensor.scatter_add_ | fluid.layers.scatter(input, index, updates, name=None, overwrite=True) |
| torch.Tensor.select | fluid.layers.slice |
| torch.Tensor.short | fluid.layers.cast(x, dtype) |
| torch.Tensor.sigmoid | fluid.layers.sigmoid(x, name=None) |
| torch.Tensor.sigmoid_ | fluid.layers.sigmoid(x, name=None) |
| torch.Tensor.sign | fluid.layers.sign(x) |
| torch.Tensor.sign_ | fluid.layers.sign(x) |
| torch.Tensor.sin | fluid.layers.sin(x, name=None) |
| torch.Tensor.sin_ | fluid.layers.sin(x, name=None) |
| torch.Tensor.split | fluid.layers.split(input, num_or_sections, dim=-1, name=None) |
| torch.Tensor.sqrt | fluid.layers.sqrt(x, name=None) |
| torch.Tensor.sqrt_ | fluid.layers.sqrt(x, name=None) |
| torch.Tensor.squeeze | fluid.layers.squeeze(input, axes, name=None) |
| torch.Tensor.squeeze_ | fluid.layers.squeeze(input, axes, name=None) |
| torch.Tensor.sub | fluid.layers.elementwise_sub(x, y, axis=-1, act=None, name=None) |
| torch.Tensor.sub_ | fluid.layers.elementwise_sub(x, y, axis=-1, act=None, name=None) |
| torch.Tensor.sum | fluid.layers.sum(x) |
| torch.Tensor.t_ | fluid.layers.transpose(x, perm, name=None) |
| torch.Tensor.tanh | fluid.layers.tanh(x, name=None) |
| torch.Tensor.tanh_ | fluid.layers.tanh(x, name=None) |
| torch.Tensor.to | fluid.layers.cast(x, dtype) |
| torch.Tensor.topk | fluid.layers.topk(input, k, name=None) |
| torch.Tensor.transpose | fluid.layers.transpose(x, perm, name=None) |
| torch.Tensor.transpose_ | fluid.layers.transpose(x, perm, name=None) |
| torch.Tensor.trunc | fluid.layers.floor(x, name=None) |
| torch.Tensor.unbind | fluid.layers.unstack(x, axis=0, num=None) |
| torch.Tensor.unique | fluid.layers.unique |
| torch.Tensor.unsqueeze | fluid.layers.unsqueeze |
| torch.Tensor.unsqueeze_ | fluid.layers.unsqueeze(input, axes, name=None) |
| torch.Tensor.view | fluid.layers.reshape(x, shape, actual_shape=None, act=None, inplace=False, name=None) |
| torch.Tensor.where | fluid.layers.where |
| torch.Tensor.zero_ | fluid.layers.zeros(shape, dtype, force_cpu=False) |
| torch.topk | fluid.layers.topk(input, k, name=None) |
| torch.transpose | fluid.layers.transpose(x, perm, name=None) |
| torch.tril | fluid.layers.tril |
| torch.triu | fluid.layers.triu |
| torch.unique | fluid.layers.unique(x, dtype=‘int32’) |
| torch.unsqueeze | fluid.layers.unsqueeze(input, axes, name=None) |
| torch.var_mean | fluid.layers.mean |
| torch.where | fluid.layers.where(condition) |
| torch.zeros | fluid.layers.zeros(shape, dtype, force_cpu=False) |
| torch.zeros_like | fluid.layers.zeros_like(x, out=None) |
| torch.utils.data.dataloader.default_collate | fluid.io.default_collate_fn |
| torch.optim.lr_scheduler.StepLR | fluid.dygraph.StepDecay |
paddle复现pytorch踩坑(二）：paddleAPI对照表（百度论文复现营活动提供）相关推荐
- Pytorch踩坑记录：关于用net.eval()和with no grad装饰器计算结果不一样的问题
Pytorch踩坑记录 相同点 net.eval()和with toch.no_grad()的相同点:都停止反向传播 不同点: 1.net.eval() 用net.eval(),此时BN层会用训练时的 ...
- pytorch .item_从数据到模型，你可能需要1篇详实的pytorch踩坑指南
原创 · 作者 | Giant 学校 | 浙江大学 研究方向 | 对话系统.text2sql 熟悉DL的朋友应该知道Tensorflow.Pytorch.Caffe这些成熟的框架,它们让广大AI爱好者 ...
- PyTorch踩坑记录
使用Pytorch也一年多了,记录一下自己遇到的各种问题. 各种踩坑记录 loss.backward()报错 一种非常常见的错误,在网络前向传播时没问题,但是当loss.backward()时会报错 ...
- 【百度论文复现赛】ArcFace: Additive Angular Margin Loss for Deep Face Recognition
文章目录 摘要 Introduction Proposed Approach ArcFace SphereFace与CosFace的比较 与其它损失函数比较 实验 Implementation Det ...
- Pytorch踩坑记之交叉熵（nn.CrossEntropy，nn.NLLLoss，nn.BCELoss的区别和使用）
目录 nn.Softmax和nn.LogSoftmax nn.NLLLoss nn.CrossEntropy nn.BCELoss 总结 在Pytorch中的交叉熵函数的血泪史要从nn.CrossEn ...
- 在linux上运行python脚本(安装pytorch踩坑记录,pyinstaller使用方式,构建docker镜像)
背景 脚本需要导入pytorch等库才能运行. 脚本在windows上运行成功,尝试放到linux上运行. linux服务器内存较小. 方法一:在linux上安装依赖 把脚本放到linux上,直接安装 ...
- （微信支付签名失败）微信支付v2版本踩坑-二次签名不同类型签名参数key不同
小伙伴们在做微信支付对接的时候会遇到各种各样的坑,特别是微信支付v2版本,微信支付的二次签名官方文档没有给出明显的签名参数,导致在开发中很容易碰到微信支付的这个坑(微信支付签名失败),app签名key ...
- PyTorch踩坑记录——torch.functional 与 torch.nn.functional的区别
问题描述: 提示:刚入门深度学习,记录一些犯下的小错误: 由于本周开始试图复现华为的CTR库以增加记忆,熟悉代码细节,没想到第一天看基础模块的时候就遇到了麻烦,在torch.utils类中,有如下获取 ...
- Superpoint复现过程——踩坑实录分享
代码来自GitHub,搜索superpoint即可下载. 2022.5.10 环境配置 Anaconda虚拟环境下,下载相关包 python 3.6 tensorflow-gpu1.14 numpy ...
最新文章
- nn.BCELoss与nn.CrossEntropyLoss的区别
- org.junit.runners.model.InvalidTestClassError: Invalid test class ‘‘: 1. No runnable methods
- 漫画算法：无序数组排序后的最大相邻差值
- SpringBoot时间格式化的5种方法！
- 奇怪的比赛--蓝桥杯
- PIL库的总结与简单应用
- Github客户端下载以及使用方法
- 基于肌电信号(sEMG) 的深度学习手势分类
- 检查linux是否安装rar,rarlinux的安装及使用
- CAD必备插件分享 快速提高工作效率
- Xposed框架原理深入研究
- mybatisplus代码生成器构建项目以及lambd链式编程
- 软件开发过程回归与回溯之区别
- mpAndroidchart 坐标和图表距离_合二为一——在Excel中制作组合图表！
- 软件测试-5-移动应用测试
- Galaxy Nexus I9250一直处于充电状态的解决办法
- Hive时间格式转换
- Fiddler4.6中文版和Jmeter5.1集成导出
- Intranet+Intranet QA-11/20 游记
- 解决mac excel中文乱码的问题
热门文章
- vue日历，阳历， 阴历，周，年，月，下拉选择，引用element组件，
- 麒麟系统挂载文件、麒麟系统挂载windows共享文件夹
- 【论文泛读】Multi-modal Sarcasm Detection and Humor Classification in Code-mixed Conversations
- 基于FPGA的混沌信号发生器设计与实现
- python-opencv基础
- 小猫爪：S32K3学习笔记09-S32K3之Safety简介
- Android 插件化学习
- 哈尔滨新发屯双星计算机学院,家校合作，助力成长——计算机与信息工程学院召开2019级新生家长见面会...
- Nat. Commun.| 机器学习对可突变的治疗性抗体的亲和力和特异性进行共同优化
- 以太坊签名，验证签名， EIP712domain Permit授权并转账