index
int64
0
50.7k
project_path
stringclasses
279 values
completion_path
stringlengths
57
199
signature_postion
stringlengths
3
11
class_position
stringlengths
3
11
function_name
stringlengths
1
62
class_name
stringlengths
0
55
body_position
stringlengths
3
11
body_position_with_comments
stringlengths
3
11
masked_class
stringlengths
14
8.32k
masked_class_with_comment
stringlengths
14
8.32k
func_body
stringlengths
9
6.72k
func_body_with_comment
stringlengths
9
6.72k
len_func_body
int64
1
643
len_input
int64
10
2.02k
len_output
int64
4
1.9k
total
int64
20
2.05k
file_code
stringlengths
14
442k
file_code_with_comment
stringlengths
14
442k
45,321
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/mdetr/transformer.py
162,169
135,186
forward
TransformerEncoder
169,186
169,186
class TransformerEncoder(nn.Module): """ A transformer encoder. Args: encoder_layer (nn.Module): Module for an individual encoder layer. num_layers (int): Number of encoder layers. norm (Optional[nn.Module]): Normalization applied after last encoder layer. Default:...
class TransformerEncoder(nn.Module): """ A transformer encoder. Args: encoder_layer (nn.Module): Module for an individual encoder layer. num_layers (int): Number of encoder layers. norm (Optional[nn.Module]): Normalization applied after last encoder layer. Default:...
output = src for layer in self.layers: output = layer( output, src_mask=mask, src_key_padding_mask=src_key_padding_mask, pos=pos, ) if self.norm is not None: output = self.norm(output) ...
output = src for layer in self.layers: output = layer( output, src_mask=mask, src_key_padding_mask=src_key_padding_mask, pos=pos, ) if self.norm is not None: output = self.norm(output) ...
25
337
106
443
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, NamedTuple, Optional import torch from torch import nn, Tensor from torchmultimod...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, NamedTuple, Optional import torch from torch import nn, Tensor from torchmultimod...
45,322
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/mdetr/transformer.py
326,333
264,357
forward_pre
TransformerEncoderLayer
333,345
333,345
class TransformerEncoderLayer(nn.Module): """ A single layer from a transformer encoder. Args: d_model (int): Number of features in the input. num_heads (int): Number of heads in multi-head attention. dim_feedforward (int): Dimension of feedforward network. Default: 2048 ...
class TransformerEncoderLayer(nn.Module): """ A single layer from a transformer encoder. Args: d_model (int): Number of features in the input. num_heads (int): Number of heads in multi-head attention. dim_feedforward (int): Dimension of feedforward network. Default: 2048 ...
x = src x = self.norm1(x) q = k = self.with_pos_embed(x, pos) self_attention_outputs = self.self_attn( q, k, value=x, attn_mask=src_mask, key_padding_mask=src_key_padding_mask )[0] x = x + self.dropout1(self_attention_outputs) x = self.norm2(x) ...
x = src x = self.norm1(x) q = k = self.with_pos_embed(x, pos) self_attention_outputs = self.self_attn( q, k, value=x, attn_mask=src_mask, key_padding_mask=src_key_padding_mask )[0] x = x + self.dropout1(self_attention_outputs) x = self.norm2(x) ...
39
983
168
1,151
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, NamedTuple, Optional import torch from torch import nn, Tensor from torchmultimod...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, NamedTuple, Optional import torch from torch import nn, Tensor from torchmultimod...
45,324
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/omnivore.py
77,78
54,86
forward
Omnivore
78,86
78,86
class Omnivore(nn.Module): """Omnivore is a model that accept multiple vision modality. Omnivore (https://arxiv.org/abs/2201.08377) is a single model that able to do classification on images, videos, and single-view 3D data using the same shared parameters of the encoder. Args: encoder (nn.Mod...
class Omnivore(nn.Module): """Omnivore is a model that accept multiple vision modality. Omnivore (https://arxiv.org/abs/2201.08377) is a single model that able to do classification on images, videos, and single-view 3D data using the same shared parameters of the encoder. Args: encoder (nn.Mod...
x = self.encoder(x) assert ( input_type in self.heads ), f"Unsupported input_type: {input_type}, please use one of {list(self.heads.keys())}" x = self.heads[input_type](x) return x
x = self.encoder(x) assert ( input_type in self.heads ), f"Unsupported input_type: {input_type}, please use one of {list(self.heads.keys())}" x = self.heads[input_type](x) return x
22
324
86
410
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, Optional import torch import torchmultimodal.utils.common as common_utils ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, Optional import torch import torchmultimodal.utils.common as common_utils ...
45,325
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/omnivore.py
119,120
86,134
forward
PatchEmbedOmnivore
120,134
120,134
class PatchEmbedOmnivore(nn.Module): """Patch Embedding strategy for Omnivore model It will use common PatchEmbed3d for image and video, for single view depth image it will have separate embedding for the depth channel and add the embedding result with the RGB channel reference: https://arxiv.org/ab...
class PatchEmbedOmnivore(nn.Module): """Patch Embedding strategy for Omnivore model It will use common PatchEmbed3d for image and video, for single view depth image it will have separate embedding for the depth channel and add the embedding result with the RGB channel reference: https://arxiv.org/ab...
# x: B C D H W # Note: D here represent time assert x.ndim == 5 has_depth = x.shape[1] == 4 if has_depth: x_rgb = self.patch_embed(x[:, :3, ...]) x_d = self.depth_patch_embed(x[:, 3:, ...]) x = x_rgb + x_d else: x = self.pa...
# x: B C D H W # Note: D here represent time assert x.ndim == 5 has_depth = x.shape[1] == 4 if has_depth: x_rgb = self.patch_embed(x[:, :3, ...]) x_d = self.depth_patch_embed(x[:, 3:, ...]) x = x_rgb + x_d else: x = self.pa...
45
393
142
535
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, Optional import torch import torchmultimodal.utils.common as common_utils ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, Optional import torch import torchmultimodal.utils.common as common_utils ...
45,326
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/two_tower.py
73,79
19,92
_get_tower_input
TwoTower
79,92
79,92
class TwoTower(nn.Module): """ A two tower architecture with a pair of late fusion models (for now, can be extended) followed by a fusion for output of each tower. Args: tower_id_to_tower (Dict[str, LateFusion]): mapping of tower id to tower model. Size should be 2, same tower should be ...
class TwoTower(nn.Module): """ A two tower architecture with a pair of late fusion models (for now, can be extended) followed by a fusion for output of each tower. Args: tower_id_to_tower (Dict[str, LateFusion]): mapping of tower id to tower model. Size should be 2, same tower should be ...
tower_input = {} channel_name_mapping: Dict[str, str] = {} if self.shared_tower_id_to_channel_mapping is not None: if self.shared_tower_id_to_channel_mapping.get(tower_id) is not None: channel_name_mapping = self.shared_tower_id_to_channel_mapping[tower_id] f...
tower_input = {} channel_name_mapping: Dict[str, str] = {} if self.shared_tower_id_to_channel_mapping is not None: if self.shared_tower_id_to_channel_mapping.get(tower_id) is not None: channel_name_mapping = self.shared_tower_id_to_channel_mapping[tower_id] f...
42
809
199
1,008
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from collections import OrderedDict from typing import Dict, List, NamedTuple, Optional from torch import nn, ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from collections import OrderedDict from typing import Dict, List, NamedTuple, Optional from torch import nn, ...
45,328
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/gpt.py
424,438
385,519
forward
MultimodalTransformerDecoder
467,504
438,504
class MultimodalTransformerDecoder(nn.Module): """A transformer decoder for two modalities The token- and position- embedding layers are per modality: * During training both modalities are fed into the module and concatenated as a single sequence of tokenized embedding vectors * Dur...
class MultimodalTransformerDecoder(nn.Module): """A transformer decoder for two modalities The token- and position- embedding layers are per modality: * During training both modalities are fed into the module and concatenated as a single sequence of tokenized embedding vectors * Dur...
if (in_modality is None) and (out_modality is None): raise ValueError( "in_modality and out_modality sequences cannot be both empty" ) # Since generation is based on the previous data point (autoregressive) where # only one modality is needed at any point...
""" Args: in_modality (Tensor, optional): Tensor of dimension ``(b, in_seq_len, d_model)`` containing tokenized embeddings for the input modality. Defaults to ``None``. out_modality (Tensor, optional): Tensor of dimension ``(b, out_seq_len, d_model')`` containing ...
153
1,451
512
1,963
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
45,329
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/gpt.py
760,761
740,777
forward
RightShift
768,777
761,777
class RightShift(nn.Module): """Shifts the embedding vectors along the sequence dimension to the right. Since the decoder progresses by taking the token it generates in the previous step, before it has generated anything it needs a token to start with. Hence, the start-of-sentence (SOS) token. The SOS ...
class RightShift(nn.Module): """Shifts the embedding vectors along the sequence dimension to the right. Since the decoder progresses by taking the token it generates in the previous step, before it has generated anything it needs a token to start with. Hence, the start-of-sentence (SOS) token. The SOS ...
x_shape = list(x.shape) x = x.flatten(start_dim=1, end_dim=-2) # (batch, seq_len, emb) sos = self.sos.unsqueeze(0).unsqueeze(1).repeat(x_shape[0], 1, 1) # (b, 1, emb) # Shift one unit to the right along dim ``seq_len`` x = torch.cat( (sos.data, x[:, :-1, :]), dim=1 ...
""" Args: x (Tensor): An input tensor of shape ``(b, seq_len, emb_dim)``. Returns; A tensor of the same shape as that of the input with the ``sos`` token prepended. """ x_shape = list(x.shape) x = x.flatten(start_dim=1, end_dim=-2) # (batch, seq_...
48
360
181
541
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
45,330
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/gpt.py
537,547
519,609
forward
TransformerDecoder
567,609
547,609
class TransformerDecoder(nn.Module): """A transformer decoder. Args: decoder_layer (nn.Module): The transformer decoder layer. An instance of :class:`TransformerDecoderLayer`. num_layers (int): The number of transformer decoder layers to be stacked up. """ def __init__( ...
class TransformerDecoder(nn.Module): """A transformer decoder. Args: decoder_layer (nn.Module): The transformer decoder layer. An instance of :class:`TransformerDecoderLayer`. num_layers (int): The number of transformer decoder layers to be stacked up. """ def __init__( ...
if attn_mask is not None and attn_mask.dim() == 2: attn_mask = attn_mask[ None, None, :, : ] # (q_seq_len, k_seq_len) -> (1, 1, q_seq_len, k_seq_len) if head_mask is not None and head_mask.dim() == 3: head_mask = head_mask[None, :, :, :] all...
""" Args: hidden_states (Tensor): Tensor of the embedding vectors of dimension ``(b, seq_len, emb_dim)``. attn_mask (Tensor, optional): Tensor of dimension ``(q_seq_len, k_seq_len)`` or ``(b, q_seq_len, k_seq_len)`` where prefixes ``q`` and ``k`` stand for query a...
122
679
525
1,204
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
45,331
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/gpt.py
659,668
609,740
forward
TransformerDecoderLayer
685,716
668,716
class TransformerDecoderLayer(nn.Module): """A single layer from a GPT transformer decoder Layer norm is applied before the attention layer and the feedforward layer so that the gradients are well-behaved at initialization for training stability. This is also called "Pre-LN Transformer" studied in `"On...
class TransformerDecoderLayer(nn.Module): """A single layer from a GPT transformer decoder Layer norm is applied before the attention layer and the feedforward layer so that the gradients are well-behaved at initialization for training stability. This is also called "Pre-LN Transformer" studied in `"On...
attn_probs = None past_key_values = None attn_out = self._attn( self.norm_attn(x), attn_mask, head_mask, return_attn_weights, use_cache=use_cache, causal=causal, ) if return_attn_weights: attn_h...
""" Args: x (Tensor): input embedding vectors. attn_mask (Tensor, optional): Tensor of dimension ``(b, q_seq_len, k_seq_len)`` where prefixes ``q`` and ``k`` stand for query and key. Contains 1s for positions to attend to and 0s for masked positions. ...
50
1,224
304
1,528
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from functools import partial from typing import Any, Callable, Dict, NamedTuple, Optional, Tuple, Union impor...
45,334
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/video_vqvae.py
444,445
408,452
forward
AttentionResidualBlock
449,452
445,452
class AttentionResidualBlock(nn.Module): """Residual block with axial attention. Implements the component as proposed in `"VideoGPT: Video Generation using VQ-VAE and Transformers (Yan et al. 2022)"<https://arxiv.org/pdf/2104.10157.pdf>`_. Code reference: https://github.com/wilson1yan/VideoGPT...
class AttentionResidualBlock(nn.Module): """Residual block with axial attention. Implements the component as proposed in `"VideoGPT: Video Generation using VQ-VAE and Transformers (Yan et al. 2022)"<https://arxiv.org/pdf/2104.10157.pdf>`_. Code reference: https://github.com/wilson1yan/VideoGPT...
return x + self.block(x)
""" Args: x (Tensor): Input of shape ``(b, c, d1, d2, d3)``. """ return x + self.block(x)
4
535
16
551
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
45,335
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/video_vqvae.py
37,45
23,87
forward
AxialAttention
62,87
45,87
class AxialAttention(nn.Module): """Computes attention over a single axis of the input. Other dims are flattened into the batch dimension. Args: axial_dim (int): Dimension to compute attention on, indexed by input dimensions (i.e., ``0`` for first input dimension, ``1`` for second). ...
class AxialAttention(nn.Module): """Computes attention over a single axis of the input. Other dims are flattened into the batch dimension. Args: axial_dim (int): Dimension to compute attention on, indexed by input dimensions (i.e., ``0`` for first input dimension, ``1`` for second). ...
# Ensure axial dim is within right dimensions, should be between head dim and embedding dim if self.axial_dim >= len(q.shape) - 1: raise ValueError("axial dim does not match input shape") # flatten all dims into batch dimension except chosen axial dim and channel dim # b, h,...
""" Args: q (Tensor): Query input of shape ``(b, h, d1, ..., dn, dim_q)`` where ``h`` is the number of attention heads, ``(d1, ..., dn)`` are the query latent dimensions and ``dim_q`` is the dimension of the query embeddings. k, v (Tensor): Key/val...
106
610
385
995
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
45,336
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/video_vqvae.py
127,128
87,143
forward
AxialAttentionBlock
128,143
128,143
class AxialAttentionBlock(nn.Module): """Computes multihead axial attention across all dims of the input. Axial attention is an alternative to standard full attention, where instead of computing attention across the entire flattened input, you compute it for each dimension. To capture the global contex...
class AxialAttentionBlock(nn.Module): """Computes multihead axial attention across all dims of the input. Axial attention is an alternative to standard full attention, where instead of computing attention across the entire flattened input, you compute it for each dimension. To capture the global contex...
n_channel = x.shape[1] if n_channel != self.qkv_dim: raise ValueError( f"Input channel dimension is {n_channel}, expected {self.qkv_dim}" ) h = shift_dim(x, 1, -1) # (b, c, d1, ..., dn) -> (b, d1, ..., dn, c) attn_out = torch.zeros_like(h) ...
n_channel = x.shape[1] if n_channel != self.qkv_dim: raise ValueError( f"Input channel dimension is {n_channel}, expected {self.qkv_dim}" ) h = shift_dim(x, 1, -1) # (b, c, d1, ..., dn) -> (b, d1, ..., dn, c) attn_out = torch.zeros_like(h) ...
67
577
236
813
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
45,337
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/video_vqvae.py
392,393
320,408
forward
VideoDecoder
397,408
393,408
class VideoDecoder(nn.Module): """Decoder for Video VQVAE. Takes quantized output from codebook and applies a ``SamePadConv3d`` layer, a stack of ``AttentionResidualBlocks``, followed by a specified number of ``SamePadConvTranspose3d`` layers. The residual blocks use Axial Attention to enhance represen...
class VideoDecoder(nn.Module): """Decoder for Video VQVAE. Takes quantized output from codebook and applies a ``SamePadConv3d`` layer, a stack of ``AttentionResidualBlocks``, followed by a specified number of ``SamePadConvTranspose3d`` layers. The residual blocks use Axial Attention to enhance represen...
in_channel = x.shape[1] if in_channel != self.conv_in.conv.in_channels: raise ValueError( f"expected input channel dim to be {self.conv_in.conv.in_channels}, but got {in_channel}" ) h = self.conv_in(x) h = self.res_stack(h) h = self.convts(...
""" Args: x (Tensor): Input quantized embeddings with shape ``(b, emb_dim, d1, d2, d3)``. """ in_channel = x.shape[1] if in_channel != self.conv_in.conv.in_channels: raise ValueError( f"expected input channel dim to be {self.conv_in.conv.in...
31
996
125
1,121
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
45,338
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/video_gpt/video_vqvae.py
291,292
217,320
get_latent_shape
VideoEncoder
293,304
292,304
class VideoEncoder(nn.Module): """Encoder for Video VQVAE. Stacks specified number of ``SamePadConv3d`` layers followed by a stack of ``AttentionResidualBlocks`` and a final ``SamePadConv3d`` layer before the codebook. The residual blocks use Axial Attention to enhance representations of video data...
class VideoEncoder(nn.Module): """Encoder for Video VQVAE. Stacks specified number of ``SamePadConv3d`` layers followed by a stack of ``AttentionResidualBlocks`` and a final ``SamePadConv3d`` layer before the codebook. The residual blocks use Axial Attention to enhance representations of video data...
latent_shape = list(input_shape) for layer in self.convs: # ignore conv_out since it has a stride of 1 if isinstance(layer, SamePadConv3d): # SamePadConv should downsample input shape by factor of stride latent_shape = [ latent_shape[dim] ...
"""Return shape of encoder output based on number of downsampling conv layers""" latent_shape = list(input_shape) for layer in self.convs: # ignore conv_out since it has a stride of 1 if isinstance(layer, SamePadConv3d): # SamePadConv should downsample input shape by...
43
1,146
133
1,279
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, cast, List, Optional, Tuple, Union import torch from torch import nn, Size, Tensor f...
45,339
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/models/vqvae.py
72,75
26,119
encode
VQVAE
85,93
75,93
class VQVAE(nn.Module): """General model for VQVAE that provides codebook layer to link user specified encoder and decoder. Vector Quantized Variational Autoencoder is a type of autoencoder that defines an embedding of discrete vectors as the latent variables in the bottleneck layer instead of norm...
class VQVAE(nn.Module): """General model for VQVAE that provides codebook layer to link user specified encoder and decoder. Vector Quantized Variational Autoencoder is a type of autoencoder that defines an embedding of discrete vectors as the latent variables in the bottleneck layer instead of norm...
encoded = self.encoder(x) out = self.codebook(encoded) indices = out.codebook_indices quantized = out.quantized if return_embeddings: return indices, quantized return indices
"""Converts input data to token ids Args: x (Tensor): Input data of shape ``(b, c, d1, ..., dn)``. return_embeddings (bool): Flag to return also the quantized embeddings. Defaults to ``False``. Returns: * A tensor of token ids: ``(b, d1, ...., dn)`` ...
19
1,127
71
1,198
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Tuple, Union from torch import nn, Size, Tensor from torchmultimodal.modules.la...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Tuple, Union from torch import nn, Size, Tensor from torchmultimodal.modules.la...
45,341
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/encoders/bert_text_encoder.py
67,77
17,123
forward
BERTTextEncoder
77,123
77,123
class BERTTextEncoder(nn.Module): """ General text transformer encoder with embeddings, following BERT. Can be constructed with any user-provided embeddings and encoder. Based on https://github.com/huggingface/transformers/blob/main/src/transformers/models/bert/modeling_bert.py#L870 Attributes: ...
class BERTTextEncoder(nn.Module): """ General text transformer encoder with embeddings, following BERT. Can be constructed with any user-provided embeddings and encoder. Based on https://github.com/huggingface/transformers/blob/main/src/transformers/models/bert/modeling_bert.py#L870 Attributes: ...
if input_ids is not None: input_shape = input_ids.size() device = input_ids.device elif inputs_embeds is not None: input_shape = inputs_embeds.size()[:-1] device = inputs_embeds.device else: raise ValueError("input_ids or inputs_embeds ...
if input_ids is not None: input_shape = input_ids.size() device = input_ids.device elif inputs_embeds is not None: input_shape = inputs_embeds.size()[:-1] device = inputs_embeds.device else: raise ValueError("input_ids or inputs_embeds ...
108
830
532
1,362
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Optional import torch from torch import nn, Tensor from torchmultimodal.models.fl...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Optional import torch from torch import nn, Tensor from torchmultimodal.models.fl...
45,342
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/encoders/embedding_encoder.py
48,49
14,59
forward
EmbeddingEncoder
49,59
49,59
class EmbeddingEncoder(nn.Module): """Combine embeddings for tensor representing list of indices based on pooling type Args: embedding (nn.Embedding): embedding module pooling_type (str): pooling function to combine the embeddings like sum. Choose from pooling_types pooling_dim ...
class EmbeddingEncoder(nn.Module): """Combine embeddings for tensor representing list of indices based on pooling type Args: embedding (nn.Embedding): embedding module pooling_type (str): pooling function to combine the embeddings like sum. Choose from pooling_types pooling_dim ...
if self.use_hash: x = x % (self.embedding.num_embeddings - 1) + 1 out = self.embedding(x) if self.pooling_type == "sum": out = torch.sum(out, dim=self.pooling_dim) elif self.pooling_type == "mean": out = torch.mean(out, dim=self.pooling_dim) el...
if self.use_hash: x = x % (self.embedding.num_embeddings - 1) + 1 out = self.embedding(x) if self.pooling_type == "sum": out = torch.sum(out, dim=self.pooling_dim) elif self.pooling_type == "mean": out = torch.mean(out, dim=self.pooling_dim) el...
37
348
147
495
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import torch from torch import nn, Tensor POOLING_TYPES = ["sum", "mean", "max"] class EmbeddingEncoder(nn....
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import torch from torch import nn, Tensor POOLING_TYPES = ["sum", "mean", "max"] class EmbeddingEncoder(nn....
45,343
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/encoders/mil_encoder.py
85,86
17,100
forward
MILEncoder
86,100
86,100
class MILEncoder(nn.Module): """ Multi instance learning encoder that partitions the input into a set of inputs and uses a shared encoder followed by deepset fusion to get a pooled representation of the entire input. Example use is to build a single representation from embeddings of all images in a ...
class MILEncoder(nn.Module): """ Multi instance learning encoder that partitions the input into a set of inputs and uses a shared encoder followed by deepset fusion to get a pooled representation of the entire input. Example use is to build a single representation from embeddings of all images in a ...
idx = 0 input_size = x.size(dim=1) if input_size != sum(self.partition_sizes): raise ValueError( f"partition sizes should sum to the input size {input_size}" ) partitioned_input = torch.split(x, self.partition_sizes, dim=1) encoded_input =...
idx = 0 input_size = x.size(dim=1) if input_size != sum(self.partition_sizes): raise ValueError( f"partition sizes should sum to the input size {input_size}" ) partitioned_input = torch.split(x, self.partition_sizes, dim=1) encoded_input =...
43
809
173
982
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, Optional, Union import torch from torch import nn, Tensor from torchmultimo...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, Optional, Union import torch from torch import nn, Tensor from torchmultimo...
45,345
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/encoders/vision_transformer.py
119,120
91,133
forward
GlobalAveragePooler
127,133
120,133
class GlobalAveragePooler(nn.Module): """ Global average pooler that averages the embeddings over all the patches in a sample and applies layer norm and an optional linear head on top. Args: input_dim (int): hidden dim of the transformer last hidden state. output_dim (Optional[int]): out...
class GlobalAveragePooler(nn.Module): """ Global average pooler that averages the embeddings over all the patches in a sample and applies layer norm and an optional linear head on top. Args: input_dim (int): hidden dim of the transformer last hidden state. output_dim (Optional[int]): out...
out = x[:, 1:, :].mean(dim=1) out = self.norm(out) out = self.head(out) return out
""" Args: x (Tensor): Input tensor with shape bsz x len x input_dim. The first entry in assumed to be CLS and ignored during averaging Returns: Tensor: Output tensor with shape bsz x output_dim """ out = x[:, 1:, :].mean(dim=1) out ...
13
418
51
469
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Callable, Optional, Tuple, Union import torch from torch import nn, Tensor from torch...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Callable, Optional, Tuple, Union import torch from torch import nn, Tensor from torch...
45,346
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/encoders/vision_transformer.py
57,63
20,91
forward
VisionTransformer
63,91
63,91
class VisionTransformer(nn.Module): """ General image transformer encoder with embeddings. Similar to ``VisionTransformer`` in torchvision, but more composable. Can be constructed with any user-provided embeddings, encoder, and task head. Attributes: embeddings (nn.Module): Module that projects...
class VisionTransformer(nn.Module): """ General image transformer encoder with embeddings. Similar to ``VisionTransformer`` in torchvision, but more composable. Can be constructed with any user-provided embeddings, encoder, and task head. Attributes: embeddings (nn.Module): Module that projects...
embedding_output = self.embeddings( images, image_patches_mask=image_patches_mask ).embeddings encoder_output = self.encoder( embedding_output, attention_mask=attention_mask, return_hidden_states=True, ) last_hidden_state = encode...
embedding_output = self.embeddings( images, image_patches_mask=image_patches_mask ).embeddings encoder_output = self.encoder( embedding_output, attention_mask=attention_mask, return_hidden_states=True, ) last_hidden_state = encode...
50
580
259
839
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Callable, Optional, Tuple, Union import torch from torch import nn, Tensor from torch...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Callable, Optional, Tuple, Union import torch from torch import nn, Tensor from torch...
45,347
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/encoders/weighted_embedding_encoder.py
40,41
13,53
forward
WeightedEmbeddingEncoder
41,53
41,53
class WeightedEmbeddingEncoder(nn.Module): """Combine weighted embeddings for tensor representing list of indices based on pooling type. Args: embedding (nn.Embedding): embedding module pooling_function (Callable[[Tensor, int], Union[Tensor, Tuple]]): pooling function to combine the weighte...
class WeightedEmbeddingEncoder(nn.Module): """Combine weighted embeddings for tensor representing list of indices based on pooling type. Args: embedding (nn.Embedding): embedding module pooling_function (Callable[[Tensor, int], Union[Tensor, Tuple]]): pooling function to combine the weighte...
index = torch.arange(0, weights.size(1), dtype=torch.int) index = index.to(weights.device) weighted_embeddings = self.embedding(index) * weights.unsqueeze(-1) pooled_embeddings = self.pooling_function(weighted_embeddings, self.pooling_dim) if isinstance(pooled_embeddings, Tensor)...
index = torch.arange(0, weights.size(1), dtype=torch.int) index = index.to(weights.device) weighted_embeddings = self.embedding(index) * weights.unsqueeze(-1) pooled_embeddings = self.pooling_function(weighted_embeddings, self.pooling_dim) if isinstance(pooled_embeddings, Tensor)...
50
326
192
518
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Tuple, Union import torch from torch import nn, Tensor class WeightedEmbeddingE...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Tuple, Union import torch from torch import nn, Tensor class WeightedEmbeddingE...
45,348
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/fusions/attention_fusion.py
46,47
13,63
forward
AttentionFusionModule
47,63
47,63
class AttentionFusionModule(nn.Module): """ Fuse embeddings through weighted sum of the corresponding linear projections. Linear layer for learning the weights. Args: channel_to_encoder_dim: mapping of channel name to the encoding dimension encoding_projection_dim: common dimension to p...
class AttentionFusionModule(nn.Module): """ Fuse embeddings through weighted sum of the corresponding linear projections. Linear layer for learning the weights. Args: channel_to_encoder_dim: mapping of channel name to the encoding dimension encoding_projection_dim: common dimension to p...
concatenated_in = torch.cat( [embeddings[k] for k in sorted(embeddings.keys())], dim=-1 ) attention_weights = self.attention(concatenated_in) projected_embeddings: List[Tensor] = [] for channel, projection in self.encoding_projection.items(): projected_emb...
concatenated_in = torch.cat( [embeddings[k] for k in sorted(embeddings.keys())], dim=-1 ) attention_weights = self.attention(concatenated_in) projected_embeddings: List[Tensor] = [] for channel, projection in self.encoding_projection.items(): projected_emb...
44
385
217
602
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Dict, List, Optional import torch from torch import nn, Tensor class AttentionFusionModul...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Dict, List, Optional import torch from torch import nn, Tensor class AttentionFusionModul...
45,349
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/fusions/concat_fusion.py
29,30
13,34
forward
ConcatFusionModule
30,34
30,34
class ConcatFusionModule(nn.Module): """Module to fuse modalities via concatenation. Sorted by keys for consistency. Inputs: embeddings (Dict[str, Tensor]): A dictionary mapping modalities to their tensor representations. """ def __init__(self, projection: nn.Module = None): ...
class ConcatFusionModule(nn.Module): """Module to fuse modalities via concatenation. Sorted by keys for consistency. Inputs: embeddings (Dict[str, Tensor]): A dictionary mapping modalities to their tensor representations. """ def __init__(self, projection: nn.Module = None): ...
concatenated_in = torch.cat( [embeddings[k] for k in sorted(embeddings.keys())], dim=-1 ) return self.projection(concatenated_in)
concatenated_in = torch.cat( [embeddings[k] for k in sorted(embeddings.keys())], dim=-1 ) return self.projection(concatenated_in)
12
162
56
218
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Dict import torch from torch import nn class ConcatFusionModule(nn.Module): """Module...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Dict import torch from torch import nn class ConcatFusionModule(nn.Module): """Module...
45,351
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/fusions/deepset_fusion.py
130,131
14,142
_pool_features
DeepsetFusionModule
131,142
131,142
class DeepsetFusionModule(nn.Module): """ Fuse embeddings through stacking followed by pooling strategy and MLP See https://arxiv.org/pdf/2003.01607.pdf Args: channel_to_encoder_dim (Dict[str, int]): mapping of channel name to the\ encoding dimension mlp (nn.Module): MLP with in...
class DeepsetFusionModule(nn.Module): """ Fuse embeddings through stacking followed by pooling strategy and MLP See https://arxiv.org/pdf/2003.01607.pdf Args: channel_to_encoder_dim (Dict[str, int]): mapping of channel name to the\ encoding dimension mlp (nn.Module): MLP with in...
pooled_embeddings = self.pooling_function(embeddings, dim=1) if torch.jit.isinstance(pooled_embeddings, Tuple[Tensor, Tensor]): return pooled_embeddings.values if not isinstance(pooled_embeddings, Tensor): raise ValueError( f"Result from pooling function s...
pooled_embeddings = self.pooling_function(embeddings, dim=1) if torch.jit.isinstance(pooled_embeddings, Tuple[Tensor, Tensor]): return pooled_embeddings.values if not isinstance(pooled_embeddings, Tensor): raise ValueError( f"Result from pooling function s...
32
1,266
135
1,401
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Dict, Optional, Tuple import torch import torch.nn.functional as F from torch imp...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Dict, Optional, Tuple import torch import torch.nn.functional as F from torch imp...
45,352
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/fusions/deepset_fusion.py
165,166
142,171
_pool_features
DeepsetFusionWithTransformer
166,171
166,171
class DeepsetFusionWithTransformer(DeepsetFusionModule): def __init__( self, channel_to_encoder_dim: Dict[str, int], mlp: nn.Module, pooling_function: nn.TransformerEncoder, apply_attention: bool = False, attention_dim: Optional[int] = None, modality_normalize...
class DeepsetFusionWithTransformer(DeepsetFusionModule): def __init__( self, channel_to_encoder_dim: Dict[str, int], mlp: nn.Module, pooling_function: nn.TransformerEncoder, apply_attention: bool = False, attention_dim: Optional[int] = None, modality_normalize...
pooled = self.pooling_function(embeddings) # take representation of the first token as the pooled feature return pooled[:, 0, :]
pooled = self.pooling_function(embeddings) # take representation of the first token as the pooled feature return pooled[:, 0, :]
18
232
48
280
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Dict, Optional, Tuple import torch import torch.nn.functional as F from torch imp...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, Dict, Optional, Tuple import torch import torch.nn.functional as F from torch imp...
45,353
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/activation.py
42,43
28,45
forward
GEGLU
43,45
43,45
class GEGLU(nn.Module): """Gated Linear Unit with GELU activation function .. math:: \text{GEGLU}(a,b) = a * \text{GELU}(b) where :math:`a` is the first half of the input matrices and :math:`b` is the second half, as descibed in the paper: `"GLU Variants Improve Transformer"<https://arxiv.org/pdf/...
class GEGLU(nn.Module): """Gated Linear Unit with GELU activation function .. math:: \text{GEGLU}(a,b) = a * \text{GELU}(b) where :math:`a` is the first half of the input matrices and :math:`b` is the second half, as descibed in the paper: `"GLU Variants Improve Transformer"<https://arxiv.org/pdf/...
x, gate = x.chunk(2, dim=self.split_dim) return x * F.gelu(gate)
x, gate = x.chunk(2, dim=self.split_dim) return x * F.gelu(gate)
9
195
36
231
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import torch import torch.nn.functional as F from torch import nn, Tensor class SiLU(nn.Module): r"""Sigm...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import torch import torch.nn.functional as F from torch import nn, Tensor class SiLU(nn.Module): r"""Sigm...
45,354
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/activation.py
24,25
12,28
forward
SiLU
25,28
25,28
class SiLU(nn.Module): r"""Sigmoid Linear Unit .. math:: \text{SiLU}(x) = x * \sigma(1.702 * x) where :math:`\sigma(x)` is the cumulative distribution function for Logistic Distribution. Approximation of the exact GeLU for greater forward speed. Note that this is different from ``torch.nn.SiLU`` ...
class SiLU(nn.Module): r"""Sigmoid Linear Unit .. math:: \text{SiLU}(x) = x * \sigma(1.702 * x) where :math:`\sigma(x)` is the cumulative distribution function for Logistic Distribution. Approximation of the exact GeLU for greater forward speed. Note that this is different from ``torch.nn.SiLU`` ...
return torch.sigmoid(1.702 * x) * x
return torch.sigmoid(1.702 * x) * x
6
190
25
215
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import torch import torch.nn.functional as F from torch import nn, Tensor class SiLU(nn.Module): r"""Sigm...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import torch import torch.nn.functional as F from torch import nn, Tensor class SiLU(nn.Module): r"""Sigm...
45,356
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/attention.py
120,129
70,185
forward
MultiHeadAttention
144,185
129,185
class MultiHeadAttention(nn.Module): """Computes multihead attention with flexible attention mechanism and caching for fast decoding. Multihead attention linearly projects and divides queries, keys, and values into multiple 'heads'. This enables the computation of attention multiple times in parallel, ...
class MultiHeadAttention(nn.Module): """Computes multihead attention with flexible attention mechanism and caching for fast decoding. Multihead attention linearly projects and divides queries, keys, and values into multiple 'heads'. This enables the computation of attention multiple times in parallel, ...
# If kv is specified use those inputs for cross-attention, otherwise use q k = v = q if kv is None else kv # compute q q = split_multihead(self.query(q), self.n_head) # For causal k, v are provided step-wise so we should always compute them # For non-causal skip computin...
""" Args: q (Tensor): Query of shape ``(b, d1, ..., dn, dim_q)`` or ``(b, seq_len, dim_q)`` (for autoregressive decoding it's typical to pass in flattened tensors). kv (Tensor, optional): Key (and value) of shape ``(b, d1', ..., dn', dim_kv)`` or `...
197
1,084
553
1,637
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Dict, Optional, Tuple, Union import torch from torch import nn, Tensor from torch.nn i...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Dict, Optional, Tuple, Union import torch from torch import nn, Tensor from torch.nn i...
45,357
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/attention.py
26,34
15,70
forward
SelfAttention
51,70
34,70
class SelfAttention(nn.Module): """Computes attention over the entire n-dimensional input. Args: attn_dropout (float, optional): Probability of dropout after softmax. Default is ``0.0``. """ def __init__(self, attn_dropout: float = 0.0) -> None: super().__init__() self.attn_dro...
class SelfAttention(nn.Module): """Computes attention over the entire n-dimensional input. Args: attn_dropout (float, optional): Probability of dropout after softmax. Default is ``0.0``. """ def __init__(self, attn_dropout: float = 0.0) -> None: super().__init__() self.attn_dro...
_, _, *shape, _ = q.shape # flatten to b, h, (d1, ..., dn), dim_q/dim_kv q = q.flatten(start_dim=2, end_dim=-2) k = k.flatten(start_dim=2, end_dim=-2) v = v.flatten(start_dim=2, end_dim=-2) out, attn_probs = scaled_dot_product_attention( q, k, ...
""" Args: q (Tensor): Query input of shape ``(b, h, d1, ..., dn, dim_q)`` where ``h`` is the number of attention heads, ``(d1, ..., dn)`` are the query latent dimensions and ``dim_q`` is the dimension of the query embeddings. k, v (Tensor): Key/val...
46
522
225
747
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Dict, Optional, Tuple, Union import torch from torch import nn, Tensor from torch.nn i...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, Dict, Optional, Tuple, Union import torch from torch import nn, Tensor from torch.nn i...
45,358
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/attention_pooler.py
47,48
16,71
forward
AttentionPooler
55,67
48,67
class AttentionPooler(nn.Module): """ Attention pooling layer: pools inputs to sequence length n_queries by performing cross-attention with learned query embeddings. Originally proposed in https://arxiv.org/abs/1810.00825. This implementation is based on the one in open_clip repo: https:...
class AttentionPooler(nn.Module): """ Attention pooling layer: pools inputs to sequence length n_queries by performing cross-attention with learned query embeddings. Originally proposed in https://arxiv.org/abs/1810.00825. This implementation is based on the one in open_clip repo: https:...
x = self.ln_k(x) query = self.ln_q(self.query) batch_size = x.shape[0] # (n_queries, output_embed_dim) -> (batch_size, n_queries, output_embed_dim) query = self._repeat(query, batch_size) out = self.attn(query, x, x) assert isinstance(out, Tensor) out = ...
""" Inputs: x (Tensor): Input tensor of shape (batch_size, seq_len, input_embed_dim). Returns: Attention pooled tensor with shape (batch_size, n_queries, output_embed_dim). """ x = self.ln_k(x) query = self.ln_q(self.query) ...
33
578
142
720
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import List import torch from torch import nn, Tensor from torchmultimodal.modules.layers.multi_he...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import List import torch from torch import nn, Tensor from torchmultimodal.modules.layers.multi_he...
45,359
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/attention_pooler.py
88,89
71,102
forward
CascadedAttentionPooler
97,102
89,102
class CascadedAttentionPooler(nn.Module): """ Wrapper class to perform cascaded attention pooling given multiple attention poolers. E.g. in CoCa the contrastive pooler is applied on top of the outputs of the captioning pooler. Args: poolers (List[AttentionPooler]): List of individual attent...
class CascadedAttentionPooler(nn.Module): """ Wrapper class to perform cascaded attention pooling given multiple attention poolers. E.g. in CoCa the contrastive pooler is applied on top of the outputs of the captioning pooler. Args: poolers (List[AttentionPooler]): List of individual attent...
pooler_outs = [] for pooler in self.poolers: x = pooler(x) pooler_outs.append(x) return pooler_outs
""" Inputs: x (Tensor): Input tensor of shape (batch_size, seq_len, input_embed_dim). Returns: List[Tensor] containing attention pooled tensors with shapes (batch_size, n_queries, output_embed_dim), where n_queries and output_embed_dim are ...
13
272
51
323
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import List import torch from torch import nn, Tensor from torchmultimodal.modules.layers.multi_he...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import List import torch from torch import nn, Tensor from torchmultimodal.modules.layers.multi_he...
45,362
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/conv.py
62,63
15,75
forward
SamePadConv3d
67,75
63,75
class SamePadConv3d(nn.Module): """Performs a same padded convolution on a 3D input. This maintains input shape with unit stride, and divides input dims by non-unit stride. Code reference: https://github.com/wilson1yan/VideoGPT/blob/master/videogpt/vqvae.py Args: in_channels (int): Nu...
class SamePadConv3d(nn.Module): """Performs a same padded convolution on a 3D input. This maintains input shape with unit stride, and divides input dims by non-unit stride. Code reference: https://github.com/wilson1yan/VideoGPT/blob/master/videogpt/vqvae.py Args: in_channels (int): Nu...
# Calculate padding needed based on input shape only once to reduce run time if self.pad_input is None: self.pad_input = calculate_same_padding( self.kernel_size, self.stride, x.shape[2:] ) return self.conv(F.pad(x, self.pad_input))
""" Args: x (Tensor): Input of shape ``(b, c, d1, d2, d3)``. """ # Calculate padding needed based on input shape only once to reduce run time if self.pad_input is None: self.pad_input = calculate_same_padding( self.kernel_size, self.stride,...
28
579
93
672
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import warnings from itertools import repeat from typing import Any, Tuple, Union from torch import nn, Size, ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import warnings from itertools import repeat from typing import Any, Tuple, Union from torch import nn, Size, ...
45,363
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/conv.py
117,118
75,133
forward
SamePadConvTranspose3d
122,133
118,133
class SamePadConvTranspose3d(nn.Module): """Performs a same padded transposed convolution on a 3D input. This ensures output shape in input shape multiplied by stride. Code reference: https://github.com/wilson1yan/VideoGPT/blob/master/videogpt/vqvae.py Args: in_channels (int): Number ...
class SamePadConvTranspose3d(nn.Module): """Performs a same padded transposed convolution on a 3D input. This ensures output shape in input shape multiplied by stride. Code reference: https://github.com/wilson1yan/VideoGPT/blob/master/videogpt/vqvae.py Args: in_channels (int): Number ...
# Calculate padding needed based on input shape only once to reduce run time if self.pad_input is None: self.pad_input = calculate_same_padding( self.kernel_size, self.stride, x.shape[2:] ) self.convt.padding, self.convt.output_padding = calculate_tran...
""" Args: x (Tensor): Input of shape ``(b, c, d1, d2, d3)``. """ # Calculate padding needed based on input shape only once to reduce run time if self.pad_input is None: self.pad_input = calculate_same_padding( self.kernel_size, self.stride,...
37
544
155
699
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import warnings from itertools import repeat from typing import Any, Tuple, Union from torch import nn, Size, ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import warnings from itertools import repeat from typing import Any, Tuple, Union from torch import nn, Size, ...
45,364
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/multi_head_attention.py
118,128
84,182
forward
MultiHeadAttentionWithCache
148,182
128,182
class MultiHeadAttentionWithCache(nn.Module): """ MultiHeadAttention module for both self-attention(SA) and cross-attention(CA). This class supports a cache mechanism for decoders to store previous states through "past_key_value". Key/value states should be only cached for self-attention cases. q, k...
class MultiHeadAttentionWithCache(nn.Module): """ MultiHeadAttention module for both self-attention(SA) and cross-attention(CA). This class supports a cache mechanism for decoders to store previous states through "past_key_value". Key/value states should be only cached for self-attention cases. q, k...
bsz = query.size(0) embed_dim = query.size(-1) head_dim = embed_dim // self.num_heads query = self.q_proj(query) key = self.k_proj(key) value = self.v_proj(value) # bsz x seq_len x embed_dim => bsz x num_heads x seq_len x head_dim query = query.view(bsz, ...
""" Args: query (Tensor): input query of shape bsz x target_seq_len x embed_dim key (Tensor): key of shape bsz x source_seq_len x embed_dim value (Tensor): value of shape bsz x source_seq_len x embed_dim attn_mask (optional Tensor): Attention mask of shape...
163
929
563
1,492
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Optional, Tuple, Union import torch import torch.nn.functional as F from torch...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Optional, Tuple, Union import torch import torch.nn.functional as F from torch...
45,365
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/multi_head_attention.py
40,46
20,84
forward
MultiHeadSelfAttention
59,84
46,84
class MultiHeadSelfAttention(nn.Module): """ Multihead self attention. Similar to the self attention variant of MHA in attention.py but uses the scaled_dot_product_attention from PyTorch (which uses flash or memory efficient version for certain conditions). TODO: merge this into attention.py once ot...
class MultiHeadSelfAttention(nn.Module): """ Multihead self attention. Similar to the self attention variant of MHA in attention.py but uses the scaled_dot_product_attention from PyTorch (which uses flash or memory efficient version for certain conditions). TODO: merge this into attention.py once ot...
bsz = query.size(0) embed_dim = query.size(-1) projected_query = self.input_proj(query) query, key, value = projected_query.chunk(3, dim=-1) head_dim = embed_dim // self.num_heads # bsz x seq len x embed_dim => bsz x num_heads x seq len x head_dim query = query....
""" Args: query (Tensor): input query of shape bsz x seq_len x embed_dim attn_mask (optional Tensor): attention mask of shape bsz x num_heads x seq_len x seq_len. Note that the num_heads dimension can equal 1 and the mask will be broadcasted to all heads. ...
99
500
379
879
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Optional, Tuple, Union import torch import torch.nn.functional as F from torch...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Optional, Tuple, Union import torch import torch.nn.functional as F from torch...
45,366
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/normalizations.py
40,41
28,51
forward
Fp32GroupNorm
41,51
41,51
class Fp32GroupNorm(nn.GroupNorm): """ GroupNorm that supports mixed-precision / fp16 training by performing normalization in fp32 and converting back. Code ref: https://github.com/facebookresearch/fairseq/blob/0338cdc3094ca7d29ff4d36d64791f7b4e4b5e6e/fairseq/modules/fp32_group_norm.py#L13 """ ...
class Fp32GroupNorm(nn.GroupNorm): """ GroupNorm that supports mixed-precision / fp16 training by performing normalization in fp32 and converting back. Code ref: https://github.com/facebookresearch/fairseq/blob/0338cdc3094ca7d29ff4d36d64791f7b4e4b5e6e/fairseq/modules/fp32_group_norm.py#L13 """ ...
output = nn.functional.group_norm( x.float(), self.num_groups, self.weight.float() if self.weight is not None else None, self.bias.float() if self.bias is not None else None, self.eps, ) return output.type_as(x)
output = nn.functional.group_norm( x.float(), self.num_groups, self.weight.float() if self.weight is not None else None, self.bias.float() if self.bias is not None else None, self.eps, ) return output.type_as(x)
25
193
98
291
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
45,367
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/normalizations.py
17,18
13,28
forward
Fp32LayerNorm
18,28
18,28
class Fp32LayerNorm(nn.LayerNorm): def __init__(self, *args: Any, **kwargs: Any) -> None: super().__init__(*args, **kwargs) def forward(self, x: Tensor) -> Tensor: FILL_FUNC_BODY
class Fp32LayerNorm(nn.LayerNorm): def __init__(self, *args: Any, **kwargs: Any) -> None: super().__init__(*args, **kwargs) def forward(self, x: Tensor) -> Tensor: FILL_FUNC_BODY
output = nn.functional.layer_norm( x.float(), self.normalized_shape, self.weight.float() if self.weight is not None else None, self.bias.float() if self.bias is not None else None, self.eps, ) return output.type_as(x)
output = nn.functional.layer_norm( x.float(), self.normalized_shape, self.weight.float() if self.weight is not None else None, self.bias.float() if self.bias is not None else None, self.eps, ) return output.type_as(x)
25
77
99
176
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
45,368
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/normalizations.py
69,70
51,77
_norm
RMSNorm
70,72
70,72
class RMSNorm(nn.Module): """Root Mean Square Layer Normalization as proposed in: https://arxiv.org/abs/1910.07467 Calcs are done in fp32. original impl: https://github.com/facebookresearch/llama/blob/main/llama/model.py Args: dim(int) = model size eps(float) = epsilon """ ...
class RMSNorm(nn.Module): """Root Mean Square Layer Normalization as proposed in: https://arxiv.org/abs/1910.07467 Calcs are done in fp32. original impl: https://github.com/facebookresearch/llama/blob/main/llama/model.py Args: dim(int) = model size eps(float) = epsilon """ ...
return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
return x * torch.rsqrt(x.pow(2).mean(-1, keepdim=True) + self.eps)
7
254
34
288
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
45,369
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/normalizations.py
94,95
77,97
forward
SimpleRMSNorm
95,97
95,97
class SimpleRMSNorm(nn.Module): """Simple RMSNorm SRMSNorm(x) = (x / ∥x∥2) /√d as proposed in: Scaling TransNormer to 175 Billion Parameters https://arxiv.org/abs/2307.14995 Usage: use as drop in replacement for RMSNorm. """ def __init__(self, dim: int, eps: float = 1e-12): s...
class SimpleRMSNorm(nn.Module): """Simple RMSNorm SRMSNorm(x) = (x / ∥x∥2) /√d as proposed in: Scaling TransNormer to 175 Billion Parameters https://arxiv.org/abs/2307.14995 Usage: use as drop in replacement for RMSNorm. """ def __init__(self, dim: int, eps: float = 1e-12): s...
denom = x.norm(p=2, dim=-1, keepdim=True).clamp_min(self.eps).expand_as(x) return (x / denom) * self.scaling
denom = x.norm(p=2, dim=-1, keepdim=True).clamp_min(self.eps).expand_as(x) return (x / denom) * self.scaling
11
193
54
247
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any import torch from torch import nn, Tensor class Fp32LayerNorm(nn.LayerNorm): def ...
45,370
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/patch_embedding.py
86,87
26,159
_init_conv_weights
PatchEmbeddings
87,96
87,96
class PatchEmbeddings(nn.Module): """ Construct the CLS token, position and patch embeddings for vision transformer Args: image_size (Union[int, Tuple[int, int]]): Size of the input. If set to an int, we assume a square input. Defaults to 224. patch_size (int): Size of the patch. Defaults to...
class PatchEmbeddings(nn.Module): """ Construct the CLS token, position and patch embeddings for vision transformer Args: image_size (Union[int, Tuple[int, int]]): Size of the input. If set to an int, we assume a square input. Defaults to 224. patch_size (int): Size of the patch. Defaults to...
fan_in = ( self.conv_projection.in_channels * self.conv_projection.kernel_size[0] * self.conv_projection.kernel_size[1] ) nn.init.trunc_normal_(self.conv_projection.weight, std=math.sqrt(1 / fan_in)) assert self.conv_projection.bias is not None ...
fan_in = ( self.conv_projection.in_channels * self.conv_projection.kernel_size[0] * self.conv_projection.kernel_size[1] ) nn.init.trunc_normal_(self.conv_projection.weight, std=math.sqrt(1 / fan_in)) assert self.conv_projection.bias is not None ...
19
1,746
133
1,879
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from typing import Iterable, NamedTuple, Optional, Tuple, Union import torch from ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from typing import Iterable, NamedTuple, Optional, Tuple, Union import torch from ...
45,371
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/text_embedding.py
70,77
13,105
forward
BERTTextEmbeddings
77,105
77,105
class BERTTextEmbeddings(nn.Module): """Construct word, position, and token type embeddings following BERT, similar to HuggingFace BertEmbeddings Attributes: hidden_size (int): size of embedding space. Default is 768. vocab_size (int): size of vocabulary. Default is 30522. pad_token_id ...
class BERTTextEmbeddings(nn.Module): """Construct word, position, and token type embeddings following BERT, similar to HuggingFace BertEmbeddings Attributes: hidden_size (int): size of embedding space. Default is 768. vocab_size (int): size of vocabulary. Default is 30522. pad_token_id ...
if input_ids is not None: input_shape = input_ids.size() device = input_ids.device elif inputs_embeds is not None: input_shape = inputs_embeds.size()[:-1] device = inputs_embeds.device else: raise ValueError("input_ids or inputs_embeds ...
if input_ids is not None: input_shape = input_ids.size() device = input_ids.device elif inputs_embeds is not None: input_shape = inputs_embeds.size()[:-1] device = inputs_embeds.device else: raise ValueError("input_ids or inputs_embeds ...
89
988
399
1,387
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch from torch import nn, Tensor class BERTTextEmbeddings(nn.Module): ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch from torch import nn, Tensor class BERTTextEmbeddings(nn.Module): ...
45,372
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/transformer.py
592,602
529,663
forward
TransformerDecoder
627,663
602,663
class TransformerDecoder(nn.Module): """ Transformer decoder: n transformer decoder layers and an optional final LN Args: n_layer (int): number of transformer decoder layers d_model (int): size of hidden dimension of input n_head (int): number of attention heads dim_feedforw...
class TransformerDecoder(nn.Module): """ Transformer decoder: n transformer decoder layers and an optional final LN Args: n_layer (int): number of transformer decoder layers d_model (int): size of hidden dimension of input n_head (int): number of attention heads dim_feedforw...
all_hidden_states = [] current_key_values = torch.jit.annotate(List[Tuple[Tensor, Tensor]], []) for i, layer_module in enumerate(self.layer): if return_hidden_states: all_hidden_states.append(hidden_states) past_key_value = past_key_values[i] if past_k...
""" Inputs: hidden_states (Tensor): input query of shape bsz x seq_len x embed_dim encoder_hidden_states (Optional[Tensor]): input key/values of shape bsz x seq_len x embed_dim, only used for cross-attention. Default is None. attention_...
62
1,193
382
1,575
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, NamedTuple, Optional, Tuple import torch from torch import nn, Tensor fro...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, NamedTuple, Optional, Tuple import torch from torch import nn, Tensor fro...
45,374
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/transformer.py
214,220
158,259
forward
TransformerEncoder
233,259
220,259
class TransformerEncoder(nn.Module): """ Transformer encoder consisting of n Transformer encoder layers and an optional final LN Args: n_layer (int): number of Transformer encoder layers d_model (int): size of hidden dimension of input n_head (int): number of attention heads ...
class TransformerEncoder(nn.Module): """ Transformer encoder consisting of n Transformer encoder layers and an optional final LN Args: n_layer (int): number of Transformer encoder layers d_model (int): size of hidden dimension of input n_head (int): number of attention heads ...
all_hidden_states = [] for layer_module in self.layer: if return_hidden_states: all_hidden_states.append(hidden_states) layer_outputs = layer_module( hidden_states, attention_mask=attention_mask, ) hidden...
""" Args: hidden_states (Tensor): input to the transformer encoder of shape bsz x seq_len x d_model attention_mask (Optional[Tensor]): attention mask of shape bsz x seq_len x seq_len. Same format as MultiHeadSelfAttention class. return_hidden_states (bool)...
39
922
215
1,137
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, NamedTuple, Optional, Tuple import torch from torch import nn, Tensor fro...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, NamedTuple, Optional, Tuple import torch from torch import nn, Tensor fro...
45,375
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/layers/transformer.py
81,86
32,158
_attention_block
TransformerEncoderLayer
86,93
86,93
class TransformerEncoderLayer(nn.Module): """Transformer encoder layer: transformer block consisting of multihead self-attention and feedforward blocks, based on "Attention Is All You Need" (Vaswani et al. 2017). Args: d_model (int): size of hidden dimension of input n_head (int): number of...
class TransformerEncoderLayer(nn.Module): """Transformer encoder layer: transformer block consisting of multihead self-attention and feedforward blocks, based on "Attention Is All You Need" (Vaswani et al. 2017). Args: d_model (int): size of hidden dimension of input n_head (int): number of...
output = self.attention( hidden_states, attn_mask=attention_mask, ) output = self.attention_dropout(output) return output
output = self.attention( hidden_states, attn_mask=attention_mask, ) output = self.attention_dropout(output) return output
11
1,305
53
1,358
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, NamedTuple, Optional, Tuple import torch from torch import nn, Tensor fro...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Callable, List, NamedTuple, Optional, Tuple import torch from torch import nn, Tensor fro...
45,376
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/albef.py
105,112
80,139
forward
CausalLanguageModelingLoss
112,139
112,139
class CausalLanguageModelingLoss(nn.Module): """ Compute the autoregressive masked language modeling loss by predicting the next token, as used in VQA. Support loss distillation for non-zero alpha. Compute standard mlm loss for zero alpha. Args: mask_token_id (int): The token id indicating a ma...
class CausalLanguageModelingLoss(nn.Module): """ Compute the autoregressive masked language modeling loss by predicting the next token, as used in VQA. Support loss distillation for non-zero alpha. Compute standard mlm loss for zero alpha. Args: mask_token_id (int): The token id indicating a ma...
batch_size = labels.size(0) # shift prediction scores and labels by one for next-token predict prediction_scores = prediction_scores[:, :-1, :].contiguous() labels = labels[:, 1:].contiguous() mlm_loss = F.cross_entropy( prediction_scores.view(-1, prediction_scores.sh...
batch_size = labels.size(0) # shift prediction scores and labels by one for next-token predict prediction_scores = prediction_scores[:, :-1, :].contiguous() labels = labels[:, 1:].contiguous() mlm_loss = F.cross_entropy( prediction_scores.view(-1, prediction_scores.sh...
90
383
377
760
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch import torch.nn.functional as F from torch import nn, Tensor class ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch import torch.nn.functional as F from torch import nn, Tensor class ...
45,377
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/albef.py
36,45
14,80
forward
ImageTextContrastiveLoss
45,80
45,80
class ImageTextContrastiveLoss(nn.Module): """ Compute the image-text contrastive loss from image-text similarity, as used in ALBEF. Support loss distillation with pseudo-targets for non-zero alpha. Compute standard contrastive loss for zero alpha. Inputs: image_to_text_sim (Tensor): Image to t...
class ImageTextContrastiveLoss(nn.Module): """ Compute the image-text contrastive loss from image-text similarity, as used in ALBEF. Support loss distillation with pseudo-targets for non-zero alpha. Compute standard contrastive loss for zero alpha. Inputs: image_to_text_sim (Tensor): Image to t...
if sim_targets is None: sim_targets = torch.zeros(image_to_text_sim.size()).to( image_to_text_sim.device ) sim_targets.fill_diagonal_(1) if alpha != 0: assert ( image_to_text_sim_m is not None and text_to_image_sim_m is not...
if sim_targets is None: sim_targets = torch.zeros(image_to_text_sim.size()).to( image_to_text_sim.device ) sim_targets.fill_diagonal_(1) if alpha != 0: assert ( image_to_text_sim_m is not None and text_to_image_sim_m is not...
99
407
463
870
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch import torch.nn.functional as F from torch import nn, Tensor class ...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch import torch.nn.functional as F from torch import nn, Tensor class ...
45,379
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/blip2_losses.py
292,299
257,364
forward
Blip2Phase1Loss
309,364
299,364
class Blip2Phase1Loss(nn.Module): """ Blip2 phase 1 loss module Args: dim_q (int): Dimension of query tensor, this value should be the same as dim_q in qformer. default value is 768 as in the paper. enable_itc (bool): enable image-text contrastive loss, default is True e...
class Blip2Phase1Loss(nn.Module): """ Blip2 phase 1 loss module Args: dim_q (int): Dimension of query tensor, this value should be the same as dim_q in qformer. default value is 768 as in the paper. enable_itc (bool): enable image-text contrastive loss, default is True e...
# calculate similarities assert model_output.text_features is not None ( sim_i2t, sim_t2i, ) = compute_image_text_similarity( model_output.image_features, model_output.text_features, temp=self.temp, ) # calcula...
""" Inputs: model_output (Blip2Output): model output from BLIP2 (see blip2.py) blip2 (nn.Module): BLIP2 model with updated params input_ids (Optional[torch.Tensor]): text input ids of shape [bsz, seq_len]. attention_mask (Optional[torch.Tensor]): text inpu...
111
632
677
1,309
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from dataclasses import dataclass from typing import Optional, OrderedDict, Tuple import torch from torch impo...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from dataclasses import dataclass from typing import Optional, OrderedDict, Tuple import torch from torch impo...
45,381
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/contrastive_loss_with_temperature.py
185,193
121,203
forward
ContrastiveLossWithTemperature
193,203
193,203
class ContrastiveLossWithTemperature(nn.Module): """Contrastive loss with a temperature parameter, as used in CLIP and FLAVA. CLIP: https://arxiv.org/pdf/2103.00020.pdf FLAVA: https://arxiv.org/pdf/2112.04482.pdf A contrastive loss over pairs of input embeddings a and b. For each input_a embedding...
class ContrastiveLossWithTemperature(nn.Module): """Contrastive loss with a temperature parameter, as used in CLIP and FLAVA. CLIP: https://arxiv.org/pdf/2103.00020.pdf FLAVA: https://arxiv.org/pdf/2112.04482.pdf A contrastive loss over pairs of input embeddings a and b. For each input_a embedding...
self.logit_scale.data.clamp_(self.logit_scale_min, self.logit_scale_max) return contrastive_loss_with_temperature( embeddings_a=embeddings_a, embeddings_b=embeddings_b, logit_scale=self.logit_scale, backprop_type=backprop_type, cross_entropy_k...
self.logit_scale.data.clamp_(self.logit_scale_min, self.logit_scale_max) return contrastive_loss_with_temperature( embeddings_a=embeddings_a, embeddings_b=embeddings_b, logit_scale=self.logit_scale, backprop_type=backprop_type, cross_entropy_k...
11
1,065
131
1,196
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math from dataclasses import dataclass from typing import Any, Dict, Optional, OrderedDict, Tuple, Union...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math from dataclasses import dataclass from typing import Any, Dict, Optional, OrderedDict, Tuple, Union...
45,383
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/flava.py
261,267
241,297
forward
FLAVAGlobalContrastiveLoss
267,297
267,297
class FLAVAGlobalContrastiveLoss(nn.Module): def __init__( self, logit_scale: Union[float, nn.Parameter] = None, image_embedding_size: int = 768, text_embedding_size: int = 768, projection_size: int = 768, image_embedding_index: int = 0, text_embedding_index: ...
class FLAVAGlobalContrastiveLoss(nn.Module): def __init__( self, logit_scale: Union[float, nn.Parameter] = None, image_embedding_size: int = 768, text_embedding_size: int = 768, projection_size: int = 768, image_embedding_index: int = 0, text_embedding_index: ...
text_embedding = nn.functional.normalize(text_sequence, dim=-1) image_embedding = nn.functional.normalize( image_sequence, dim=-1, ) self.logit_scale.data.clamp_(0, 4.6052) output = contrastive_loss_with_temperature( embeddings_a=image_embed...
text_embedding = nn.functional.normalize(text_sequence, dim=-1) image_embedding = nn.functional.normalize( image_sequence, dim=-1, ) self.logit_scale.data.clamp_(0, 4.6052) output = contrastive_loss_with_temperature( embeddings_a=image_embed...
40
293
325
618
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
45,385
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/flava.py
122,127
110,143
forward
ITMLoss
127,143
127,143
class ITMLoss(nn.Module): def __init__( self, hidden_size: int = 768, ignore_index: int = -1, **kwargs: Any, ): super().__init__() self.pooler = Pooler(hidden_size=hidden_size) self.cls = TwoWayHead(hidden_size=hidden_size) self.ce_loss = nn.CrossE...
class ITMLoss(nn.Module): def __init__( self, hidden_size: int = 768, ignore_index: int = -1, **kwargs: Any, ): super().__init__() self.pooler = Pooler(hidden_size=hidden_size) self.cls = TwoWayHead(hidden_size=hidden_size) self.ce_loss = nn.CrossE...
if self.training: assert_labels_are_present(labels, "itm labels") pooled_output = self.pooler(hidden_states) scores = self.cls(pooled_output) if labels is None: loss = pooled_output.sum() * 0 else: loss = self.ce_loss( scores....
if self.training: assert_labels_are_present(labels, "itm labels") pooled_output = self.pooler(hidden_states) scores = self.cls(pooled_output) if labels is None: loss = pooled_output.sum() * 0 else: loss = self.ce_loss( scores....
31
170
148
318
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
45,386
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/flava.py
174,175
143,182
forward
MaskedPredictionHead
175,182
175,182
class MaskedPredictionHead(nn.Module): def __init__( self, hidden_size: int = 768, vocab_size: int = 30522, transform_act_fn: Callable[[Tensor], Tensor] = nn.functional.gelu, layer_norm_eps: float = 1e-5, use_fp32_layer_norm: bool = True, **kwargs: Any, ):...
class MaskedPredictionHead(nn.Module): def __init__( self, hidden_size: int = 768, vocab_size: int = 30522, transform_act_fn: Callable[[Tensor], Tensor] = nn.functional.gelu, layer_norm_eps: float = 1e-5, use_fp32_layer_norm: bool = True, **kwargs: Any, ):...
hidden_states = self.dense(hidden_states) hidden_states = self.transform_act_fn(hidden_states) hidden_states = self.layer_norm(hidden_states) hidden_states = self.decoder(hidden_states) return hidden_states
hidden_states = self.dense(hidden_states) hidden_states = self.transform_act_fn(hidden_states) hidden_states = self.layer_norm(hidden_states) hidden_states = self.decoder(hidden_states) return hidden_states
14
406
79
485
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
45,387
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/flava.py
206,209
182,241
forward
MaskedPredictionLoss
209,241
209,241
class MaskedPredictionLoss(nn.Module): def __init__( self, hidden_size: int = 768, vocab_size: int = 30522, transform_act_fn: Callable[[Tensor], Tensor] = nn.functional.gelu, layer_norm_eps: float = 1e-5, ignore_index: int = -1, ignore_nan: bool = False, ...
class MaskedPredictionLoss(nn.Module): def __init__( self, hidden_size: int = 768, vocab_size: int = 30522, transform_act_fn: Callable[[Tensor], Tensor] = nn.functional.gelu, layer_norm_eps: float = 1e-5, ignore_index: int = -1, ignore_nan: bool = False, ...
if self.training: assert_labels_are_present(masked_labels, "masked labels") if masked_labels is not None: masked_tokens = masked_labels.ne(self.ignore_index) masked_labels = masked_labels[masked_tokens] sequence_output = hidden_states[masked_tokens, :] ...
if self.training: assert_labels_are_present(masked_labels, "masked labels") if masked_labels is not None: masked_tokens = masked_labels.ne(self.ignore_index) masked_labels = masked_labels[masked_tokens] sequence_output = hidden_states[masked_tokens, :] ...
82
316
369
685
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
45,388
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/flava.py
91,92
84,100
forward
Pooler
92,100
92,100
class Pooler(nn.Module): def __init__(self, hidden_size: int = 768, **kwargs: Any): super().__init__() self.dense = nn.Linear(hidden_size, hidden_size) self.activation = nn.Tanh() def forward(self, hidden_states: Tensor) -> Tensor: FILL_FUNC_BODY
class Pooler(nn.Module): def __init__(self, hidden_size: int = 768, **kwargs: Any): super().__init__() self.dense = nn.Linear(hidden_size, hidden_size) self.activation = nn.Tanh() def forward(self, hidden_states: Tensor) -> Tensor: FILL_FUNC_BODY
# We "pool" the model by simply taking the hidden state corresponding # to the first token. first_token_tensor = hidden_states[:, 0] pooled_output = self.dense(first_token_tensor) pooled_output = self.activation(pooled_output) return pooled_output
# We "pool" the model by simply taking the hidden state corresponding # to the first token. first_token_tensor = hidden_states[:, 0] pooled_output = self.dense(first_token_tensor) pooled_output = self.activation(pooled_output) return pooled_output
29
105
92
197
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
45,389
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/flava.py
106,107
100,110
forward
TwoWayHead
107,110
107,110
class TwoWayHead(nn.Module): def __init__(self, hidden_size: int = 768, **kwargs: Any): super().__init__() self.seq_relationship = nn.Linear(hidden_size, 2) def forward(self, pooled_output: Tensor) -> Tensor: FILL_FUNC_BODY
class TwoWayHead(nn.Module): def __init__(self, hidden_size: int = 768, **kwargs: Any): super().__init__() self.seq_relationship = nn.Linear(hidden_size, 2) def forward(self, pooled_output: Tensor) -> Tensor: FILL_FUNC_BODY
return self.seq_relationship(pooled_output)
return self.seq_relationship(pooled_output)
2
95
20
115
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import math import warnings from dataclasses import dataclass, field from typing import Any, Callable, Optional...
45,391
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/reconstruction_loss.py
24,25
10,48
forward
ReconstructionLoss
34,48
25,48
class ReconstructionLoss(nn.Module): """ Reconstruction loss that computed MSE between predicted and target values as described in MAE paper https://arxiv.org/abs/2111.06377. Loss is averaged only over masked patches. Args: normalize_target (bool) : Whether target should be normalized. Defaults...
class ReconstructionLoss(nn.Module): """ Reconstruction loss that computed MSE between predicted and target values as described in MAE paper https://arxiv.org/abs/2111.06377. Loss is averaged only over masked patches. Args: normalize_target (bool) : Whether target should be normalized. Defaults...
if mask.sum() == 0: raise ValueError("At least one patch must be masked") if self.normalize_target: mean = target.mean(dim=-1, keepdim=True) var = target.var(dim=-1, keepdim=True) # pyre-fixme[58]: `**` is not supported for operand types `Tensor` and ...
""" Args: pred (Tensor): predicted tensor with shape bsz x num_patches x (patch_size ** 2 * channels) target (Tensor): patchified input with the same shape as pred mask (Tensor): Tensor of shape bsz x num_patches indicating which patches are masked. 1 ind...
74
283
217
500
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from torch import nn, Tensor class ReconstructionLoss(nn.Module): """ Reconstruction loss that comput...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from torch import nn, Tensor class ReconstructionLoss(nn.Module): """ Reconstruction loss that comput...
45,392
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/losses/vqvae.py
26,27
13,31
forward
CommitmentLoss
27,31
27,31
class CommitmentLoss(nn.Module): """Commitment loss calculates the mean Euclidean distance between pairs of encoder output vectors and their corresponding quantized vectors. It encourages an encoder to generate outputs closer to an embedding. This is the beta in Eq. 3 of Oord et al. 2017 (https://arxiv.org/...
class CommitmentLoss(nn.Module): """Commitment loss calculates the mean Euclidean distance between pairs of encoder output vectors and their corresponding quantized vectors. It encourages an encoder to generate outputs closer to an embedding. This is the beta in Eq. 3 of Oord et al. 2017 (https://arxiv.org/...
# Quantized vectors must be detached because commitment loss only lets gradient flow through encoder output loss = F.mse_loss(quantized.detach(), encoded) * self.commitment_cost return loss
# Quantized vectors must be detached because commitment loss only lets gradient flow through encoder output loss = F.mse_loss(quantized.detach(), encoded) * self.commitment_cost return loss
24
206
55
261
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any from torch import nn, Tensor from torch.nn import functional as F class CommitmentLos...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any from torch import nn, Tensor from torch.nn import functional as F class CommitmentLos...
45,393
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/modules/masking/random_masking.py
95,101
1,110
_random_masking_1d
101,110
101,110
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Tuple import torch from torch import Tensor class RandomMaskingOutput(NamedT...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Tuple import torch from torch import Tensor class RandomMaskingOutput(NamedT...
# x shape : bsz x h x w x embed_dim n, _, _, d = x.shape len_keep = int(num_patches_h * (1 - mask_ratio)) noise = torch.rand(n, num_patches_h, device=x.device) ids_shuffle = torch.argsort(noise, dim=1) ids_keep = ids_shuffle[:, :len_keep] index = ids_keep.unsqueeze(-1).unsqueeze(-1).repeat(1...
# x shape : bsz x h x w x embed_dim n, _, _, d = x.shape len_keep = int(num_patches_h * (1 - mask_ratio)) noise = torch.rand(n, num_patches_h, device=x.device) ids_shuffle = torch.argsort(noise, dim=1) ids_keep = ids_shuffle[:, :len_keep] index = ids_keep.unsqueeze(-1).unsqueeze(-1).repeat(1...
51
1,149
194
1,343
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Tuple import torch from torch import Tensor class RandomMaskingOutput(NamedT...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import NamedTuple, Tuple import torch from torch import Tensor class RandomMaskingOutput(NamedT...
45,396
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/clip_transform.py
144,145
84,205
bpe
CLIPBPETokenizer
145,185
145,185
class CLIPBPETokenizer: """ Construct a CLIP tokenizer. Based on byte-level Byte-Pair-Encoding. This implementation is adapted from https://git.io/JDTuJ. Example usage: tokenizer = CLIPBPETokenizer() sentence = "Hello I am using CLIP tokenizer." tokens = tokenizer.encode(sentence) token...
class CLIPBPETokenizer: """ Construct a CLIP tokenizer. Based on byte-level Byte-Pair-Encoding. This implementation is adapted from https://git.io/JDTuJ. Example usage: tokenizer = CLIPBPETokenizer() sentence = "Hello I am using CLIP tokenizer." tokens = tokenizer.encode(sentence) token...
if token in self.cache: return self.cache[token] word = tuple(token[:-1]) + (token[-1] + "</w>",) pairs = get_pairs(word) if not pairs: return token + "</w>" while True: bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float("inf"...
if token in self.cache: return self.cache[token] word = tuple(token[:-1]) + (token[-1] + "</w>",) pairs = get_pairs(word) if not pairs: return token + "</w>" while True: bigram = min(pairs, key=lambda pair: self.bpe_ranks.get(pair, float("inf"...
118
1,223
402
1,625
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
45,397
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/clip_transform.py
236,237
205,243
forward
CLIPBPETransform
237,243
237,243
class CLIPBPETransform(nn.Module): """ nn.Module wrapper around CLIPBPETokenizer. Supports either a single string or list of strings for tokenization. Args: bpe_path (Optional[str]): path to the BPE file. Defaults to CLIP_DEFAULT_VOCAB_BPE_PATH bos_token (Optional[str]): beg...
class CLIPBPETransform(nn.Module): """ nn.Module wrapper around CLIPBPETokenizer. Supports either a single string or list of strings for tokenization. Args: bpe_path (Optional[str]): path to the BPE file. Defaults to CLIP_DEFAULT_VOCAB_BPE_PATH bos_token (Optional[str]): beg...
if isinstance(text, str): return self.bpe.encode(text) else: return [self.bpe.encode(t) for t in text]
if isinstance(text, str): return self.bpe.encode(text) else: return [self.bpe.encode(t) for t in text]
12
400
52
452
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
45,398
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/clip_transform.py
350,351
303,357
forward
CLIPImageTransform
351,357
351,357
class CLIPImageTransform(nn.Module): """CLIP image transform random resized crop (train mode) or resize and center crop, followed by RGB conversion, tensor conversion, and normalization. Args: image_size (Union[int, Tuple[int]): desired output image size. image_interpolation (torchvision.tr...
class CLIPImageTransform(nn.Module): """CLIP image transform random resized crop (train mode) or resize and center crop, followed by RGB conversion, tensor conversion, and normalization. Args: image_size (Union[int, Tuple[int]): desired output image size. image_interpolation (torchvision.tr...
if isinstance(image, Image): return self.image_transform(image) image_result = torch.stack([self.image_transform(x) for x in image]) return image_result
if isinstance(image, Image): return self.image_transform(image) image_result = torch.stack([self.image_transform(x) for x in image]) return image_result
14
563
59
622
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
45,399
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/clip_transform.py
297,298
243,303
forward
CLIPTextTransform
298,303
298,303
class CLIPTextTransform(nn.Module): """CLIP text transform CLIP BPE tokenizer transform, adds start and end tokens, then pads/truncates to text_max_length as necessary. This transform is torch scriptable. Args: text_max_length (int): Maximum length of text token sequences. text_start_tok...
class CLIPTextTransform(nn.Module): """CLIP text transform CLIP BPE tokenizer transform, adds start and end tokens, then pads/truncates to text_max_length as necessary. This transform is torch scriptable. Args: text_max_length (int): Maximum length of text token sequences. text_start_tok...
text_result = self.text_transform(text) assert torch.jit.isinstance(text_result, Tensor) return text_result
text_result = self.text_transform(text) assert torch.jit.isinstance(text_result, Tensor) return text_result
8
764
46
810
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
45,400
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/clip_transform.py
416,419
357,420
forward
CLIPTransform
419,420
419,420
class CLIPTransform(nn.Module): """Image and text transform for CLIP model. Image transform: either random resized crop (train mode) or resize and center crop, followed by RGB conversion, tensor conversion, and normalization. Text transform: applies CLIP's BPE tokenizer transform, adds start and en...
class CLIPTransform(nn.Module): """Image and text transform for CLIP model. Image transform: either random resized crop (train mode) or resize and center crop, followed by RGB conversion, tensor conversion, and normalization. Text transform: applies CLIP's BPE tokenizer transform, adds start and en...
return self.image_transform(image), self.text_transform(text)
return self.image_transform(image), self.text_transform(text)
3
864
20
884
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. import html from functools import lru_cache from typing import Callable, Dict, List, Optional, Set, Tuple, Unio...
45,402
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/text_transforms.py
49,50
35,58
forward
AddToken
54,58
50,58
class AddToken(nn.Module): """Add token to beginning or end of sequence :param token: The token to be added :type token: Union[int, str] :param begin: Whether to insert token at start or end or sequence, defaults to True :type begin: bool, optional """ def __init__(self, token: Union[int, ...
class AddToken(nn.Module): """Add token to beginning or end of sequence :param token: The token to be added :type token: Union[int, str] :param begin: Whether to insert token at start or end or sequence, defaults to True :type begin: bool, optional """ def __init__(self, token: Union[int, ...
return add_token(input, self.token, self.begin)
""" :param input: Input sequence or batch :type input: Union[List[Union[str, int]], List[List[Union[str, int]]]] """ return add_token(input, self.token, self.begin)
4
197
24
221
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
45,403
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/text_transforms.py
72,73
58,86
forward
PadTransform
79,86
73,86
class PadTransform(nn.Module): """Pad tensor to a fixed length with given padding value. :param max_length: Maximum length to pad to :type max_length: int :param pad_value: Value to pad the tensor with :type pad_value: bool """ def __init__(self, max_length: int, pad_value: int) -> None: ...
class PadTransform(nn.Module): """Pad tensor to a fixed length with given padding value. :param max_length: Maximum length to pad to :type max_length: int :param pad_value: Value to pad the tensor with :type pad_value: bool """ def __init__(self, max_length: int, pad_value: int) -> None: ...
max_encoded_length = x.size(-1) if max_encoded_length < self.max_length: pad_amount = self.max_length - max_encoded_length x = torch.nn.functional.pad(x, (0, pad_amount), value=self.pad_value) return x
""" :param x: The tensor to pad :type x: Tensor :return: Tensor padded up to max_length with pad_value :rtype: Tensor """ max_encoded_length = x.size(-1) if max_encoded_length < self.max_length: pad_amount = self.max_length - max_encoded_length...
20
209
93
302
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
45,404
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/text_transforms.py
102,103
86,111
forward
ToTensor
108,111
103,111
class ToTensor(nn.Module): r"""Convert input to torch tensor :param padding_value: Pad value to make each input in the batch of length equal to the longest sequence in the batch. :type padding_value: Optional[int] :param dtype: :class:`torch.dtype` of output tensor :type dtype: :class:`torch.dtype`...
class ToTensor(nn.Module): r"""Convert input to torch tensor :param padding_value: Pad value to make each input in the batch of length equal to the longest sequence in the batch. :type padding_value: Optional[int] :param dtype: :class:`torch.dtype` of output tensor :type dtype: :class:`torch.dtype`...
return to_tensor(input, padding_value=self.padding_value, dtype=self.dtype)
""" :param input: Sequence or batch of token ids :type input: Union[List[int], List[List[int]]] :rtype: Tensor """ return to_tensor(input, padding_value=self.padding_value, dtype=self.dtype)
4
245
32
277
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
45,405
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/text_transforms.py
25,26
14,35
forward
Truncate
32,35
26,35
class Truncate(nn.Module): r"""Truncate input sequence :param max_seq_len: The maximum allowable length for input sequence :type max_seq_len: int """ def __init__(self, max_seq_len: int) -> None: super().__init__() self.max_seq_len = max_seq_len def forward(self, x: Any) -> An...
class Truncate(nn.Module): r"""Truncate input sequence :param max_seq_len: The maximum allowable length for input sequence :type max_seq_len: int """ def __init__(self, max_seq_len: int) -> None: super().__init__() self.max_seq_len = max_seq_len def forward(self, x: Any) -> An...
return truncate(x, self.max_seq_len)
""" :param input: Input sequence or batch of sequence to be truncated :type input: Union[List[Union[str, int]], List[List[Union[str, int]]]] :return: Truncated sequence :rtype: Union[List[Union[str, int]], List[List[Union[str, int]]]] """ return truncate(x, self.m...
3
213
21
234
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Any, List, Optional, Union import torch from torch import nn, Tensor from torch.nn.utils.rn...
45,406
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/transforms/video_transform.py
85,86
21,97
normalize
VideoTransform
87,97
86,97
class VideoTransform: """Transform videos for encoding Args: time_samples (int): number of frames to sample in the time dimension mean (Tuple[float, float, float]): sequence of means of each channel std (Tuple[float, float, float]): sequence of standard deviations of each channel ...
class VideoTransform: """Transform videos for encoding Args: time_samples (int): number of frames to sample in the time dimension mean (Tuple[float, float, float]): sequence of means of each channel std (Tuple[float, float, float]): sequence of standard deviations of each channel ...
b, t, _, _, _ = video.shape video = video.flatten(start_dim=0, end_dim=1) # "b t h w c -> (b t) h w c" video = shift_dim(video, -1, 1) # "(b t) h w c -> (b t) c h w" video = video.float() / 255.0 video = normalize(video, mean=self.mean, std=self.std) video = video.unf...
"""Normalizes video of dims (b, t, h, w, c) to mean 0, std 1""" b, t, _, _, _ = video.shape video = video.flatten(start_dim=0, end_dim=1) # "b t h w c -> (b t) h w c" video = shift_dim(video, -1, 1) # "(b t) h w c -> (b t) c h w" video = video.float() / 255.0 video = n...
86
899
211
1,110
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Tuple import torch import torch.nn.functional as F from torchmultimodal.utils.common import...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Tuple import torch import torch.nn.functional as F from torchmultimodal.utils.common import...
45,407
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/utils/assertion.py
10,13
1,18
assert_equal_lengths
13,18
13,18
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Sized def assert_equal_lengths( *args: Sized, msg: str = "iterable arguments must have...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Sized def assert_equal_lengths( *args: Sized, msg: str = "iterable arguments must have...
lengths = set() for item in args: lengths.add(len(item)) if len(lengths) != 1: raise ValueError(msg)
lengths = set() for item in args: lengths.add(len(item)) if len(lengths) != 1: raise ValueError(msg)
14
112
48
160
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Sized def assert_equal_lengths( *args: Sized, msg: str = "iterable arguments must have...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Sized def assert_equal_lengths( *args: Sized, msg: str = "iterable arguments must have...
45,408
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/utils/attention.py
13,14
1,66
get_extended_attention_mask
30,56
14,56
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch from torch import Tensor def get_extended_attention_mask(attention_...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch from torch import Tensor def get_extended_attention_mask(attention_...
if attention_mask.dim() == 4: # Mask has already been broadcasted to the correct shape (either # [batch_size, num_heads, query_seq_length, key_seq_length] for causal case or # [batch_size, num_heads, seq_length, seq_length] for padding case) extended_attention_mask = attention_mask ...
"""Makes attention masks broadcastable along head and sequence dimensions. Accepting two types of attention masks: - Causal: masks that prevent attending to future positions of dimensions ``(batch_size, query_seq_len, key_seq_len)`` - Padding: masks that prevent attending to token p...
132
410
399
809
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch from torch import Tensor def get_extended_attention_mask(attention_...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from typing import Optional import torch from torch import Tensor def get_extended_attention_mask(attention_...
45,409
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/utils/distributed.py
61,66
1,91
concat_gather_all_gpu
78,86
66,86
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from enum import Enum from typing import List import torch from torch import Tensor from torch.distributed imp...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from enum import Enum from typing import List import torch from torch import Tensor from torch.distributed imp...
if not torch.distributed.is_available() or not torch.distributed.is_initialized(): return tensor tensors_all_gpus = gather_tensor(tensor, backprop_type) return torch.cat(tensors_all_gpus, dim=dim)
"""Gathers a tensor across all GPUs. Inputs: tensor (Tensor): Tensors that need to be gathered. backprop_type (BackpropType): whether to backpropagate gradients to all workers (GLOBAL), just the local worker (LOCAL), or not at all (NONE). Default: BackpropType.GLOBAL ...
15
848
85
933
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from enum import Enum from typing import List import torch from torch import Tensor from torch.distributed imp...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from enum import Enum from typing import List import torch from torch import Tensor from torch.distributed imp...
45,410
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal
/home/hieuvd/dung_repo/raw/stars_03/facebookresearch_multimodal/multimodal/torchmultimodal/utils/file_io.py
10,11
1,21
_get_path_manager
11,21
11,21
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from iopath.common.file_io import HTTPURLHandler, PathManager def _get_path_manager() -> PathManager: FILL_FU...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from iopath.common.file_io import HTTPURLHandler, PathManager def _get_path_manager() -> PathManager: FILL_FU...
try: from torchmultimodal.fb.utils.file_io import FBPathManager, register_handlers pm = FBPathManager() register_handlers(pm) return pm except ImportError: pm = PathManager() pm.register_handler(HTTPURLHandler()) return pm
try: from torchmultimodal.fb.utils.file_io import FBPathManager, register_handlers pm = FBPathManager() register_handlers(pm) return pm except ImportError: pm = PathManager() pm.register_handler(HTTPURLHandler()) return pm
20
102
93
195
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from iopath.common.file_io import HTTPURLHandler, PathManager def _get_path_manager() -> PathManager: FILL_FU...
# Copyright (c) Meta Platforms, Inc. and affiliates. # All rights reserved. # # This source code is licensed under the BSD-style license found in the # LICENSE file in the root directory of this source tree. from iopath.common.file_io import HTTPURLHandler, PathManager def _get_path_manager() -> PathManager: FILL_FU...
45,415
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/examples/simulate.py
16,17
1,94
parse_args
24,56
17,56
"""Runner script for non-RL simulations in flow. Usage python simulate.py EXP_CONFIG --no_render """ import argparse import sys import json import os from flow.core.experiment import Experiment from flow.core.params import AimsunParams from flow.utils.rllib import FlowParamsEncoder def parse_args(args): FILL_FU...
"""Runner script for non-RL simulations in flow. Usage python simulate.py EXP_CONFIG --no_render """ import argparse import sys import json import os from flow.core.experiment import Experiment from flow.core.params import AimsunParams from flow.utils.rllib import FlowParamsEncoder def parse_args(args): """...
parser = argparse.ArgumentParser( description="Parse argument used when running a Flow simulation.", epilog="python simulate.py EXP_CONFIG --num_runs INT --no_render") # required input parameters parser.add_argument( 'exp_config', type=str, help='Name of the experiment confi...
"""Parse training options user can specify in command line. Returns ------- argparse.Namespace the output parser object """ parser = argparse.ArgumentParser( description="Parse argument used when running a Flow simulation.", epilog="python simulate.py EXP_CONFIG --num_ru...
100
613
356
969
"""Runner script for non-RL simulations in flow. Usage python simulate.py EXP_CONFIG --no_render """ import argparse import sys import json import os from flow.core.experiment import Experiment from flow.core.params import AimsunParams from flow.utils.rllib import FlowParamsEncoder def parse_args(args): FILL_FU...
"""Runner script for non-RL simulations in flow. Usage python simulate.py EXP_CONFIG --no_render """ import argparse import sys import json import os from flow.core.experiment import Experiment from flow.core.params import AimsunParams from flow.utils.rllib import FlowParamsEncoder def parse_args(args): """...
45,416
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/controllers/base_lane_changing_controller.py
63,64
6,82
get_action
BaseLaneChangeController
78,82
64,82
class BaseLaneChangeController(metaclass=ABCMeta): """Base class for lane-changing controllers. Instantiates a controller and forces the user to pass a lane_changing duration to the controller. Usage ----- >>> from flow.core.params import VehicleParams >>> from flow.controllers import SimL...
class BaseLaneChangeController(metaclass=ABCMeta): """Base class for lane-changing controllers. Instantiates a controller and forces the user to pass a lane_changing duration to the controller. Usage ----- >>> from flow.core.params import VehicleParams >>> from flow.controllers import SimL...
lc_action = self.get_lane_change_action(env) # TODO(ak): add failsafe return lc_action
"""Return the action of the lane change controller. Modifies the lane change action to ensure safety, if requested. Parameters ---------- env : flow.envs.Env state of the environment at the current time step Returns ------- float or int ...
9
553
42
595
"""Contains the base lane change controller class.""" from abc import ABCMeta, abstractmethod class BaseLaneChangeController(metaclass=ABCMeta): """Base class for lane-changing controllers. Instantiates a controller and forces the user to pass a lane_changing duration to the controller. Usage -...
"""Contains the base lane change controller class.""" from abc import ABCMeta, abstractmethod class BaseLaneChangeController(metaclass=ABCMeta): """Base class for lane-changing controllers. Instantiates a controller and forces the user to pass a lane_changing duration to the controller. Usage -...
45,417
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/controllers/lane_change_controllers.py
13,14
7,18
get_lane_change_action
SimLaneChangeController
15,18
14,18
class SimLaneChangeController(BaseLaneChangeController): """A controller used to enforce sumo lane-change dynamics on a vehicle. Usage: See base class for usage example. """ def get_lane_change_action(self, env): FILL_FUNC_BODY
class SimLaneChangeController(BaseLaneChangeController): """A controller used to enforce sumo lane-change dynamics on a vehicle. Usage: See base class for usage example. """ def get_lane_change_action(self, env): """See parent class.""" FILL_FUNC_BODY
return None
"""See parent class.""" return None
2
82
9
91
"""Contains a list of custom lane change controllers.""" from flow.controllers.base_lane_changing_controller import \ BaseLaneChangeController class SimLaneChangeController(BaseLaneChangeController): """A controller used to enforce sumo lane-change dynamics on a vehicle. Usage: See base class for usage ...
"""Contains a list of custom lane change controllers.""" from flow.controllers.base_lane_changing_controller import \ BaseLaneChangeController class SimLaneChangeController(BaseLaneChangeController): """A controller used to enforce sumo lane-change dynamics on a vehicle. Usage: See base class for usage ...
45,418
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/controllers/lane_change_controllers.py
24,25
18,27
get_lane_change_action
StaticLaneChanger
26,27
25,27
class StaticLaneChanger(BaseLaneChangeController): """A lane-changing model used to keep a vehicle in the same lane. Usage: See base class for usage example. """ def get_lane_change_action(self, env): FILL_FUNC_BODY
class StaticLaneChanger(BaseLaneChangeController): """A lane-changing model used to keep a vehicle in the same lane. Usage: See base class for usage example. """ def get_lane_change_action(self, env): """See parent class.""" FILL_FUNC_BODY
return 0
"""See parent class.""" return 0
2
83
6
89
"""Contains a list of custom lane change controllers.""" from flow.controllers.base_lane_changing_controller import \ BaseLaneChangeController class SimLaneChangeController(BaseLaneChangeController): """A controller used to enforce sumo lane-change dynamics on a vehicle. Usage: See base class for usage ...
"""Contains a list of custom lane change controllers.""" from flow.controllers.base_lane_changing_controller import \ BaseLaneChangeController class SimLaneChangeController(BaseLaneChangeController): """A controller used to enforce sumo lane-change dynamics on a vehicle. Usage: See base class for usage ...
45,419
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/kernel/simulation/aimsun.py
115,116
9,132
close
AimsunKernelSimulation
117,132
116,132
class AimsunKernelSimulation(KernelSimulation): """Aimsun simulation kernel. Extends KernelSimulation. Attributes ---------- master_kernel : flow.core.kernel.Kernel the higher level kernel (used to call methods from other sub-kernels) kernel_api : any an API that may be used to...
class AimsunKernelSimulation(KernelSimulation): """Aimsun simulation kernel. Extends KernelSimulation. Attributes ---------- master_kernel : flow.core.kernel.Kernel the higher level kernel (used to call methods from other sub-kernels) kernel_api : any an API that may be used to...
# save the emission data to a csv if self.emission_path is not None: name = "%s_emission.csv" % self.master_kernel.network.network.name with open(osp.join(self.emission_path, name), "w") as f: writer = csv.writer(f, delimiter=',') writer.writerow(s...
"""See parent class.""" # save the emission data to a csv if self.emission_path is not None: name = "%s_emission.csv" % self.master_kernel.network.network.name with open(osp.join(self.emission_path, name), "w") as f: writer = csv.writer(f, delimiter=',') ...
54
1,023
208
1,231
"""Script containing the base simulation kernel class.""" from flow.core.kernel.simulation.base import KernelSimulation from flow.utils.aimsun.api import FlowAimsunAPI import os.path as osp import csv from flow.core.util import ensure_dir class AimsunKernelSimulation(KernelSimulation): """Aimsun simulation kernel...
"""Script containing the base simulation kernel class.""" from flow.core.kernel.simulation.base import KernelSimulation from flow.utils.aimsun.api import FlowAimsunAPI import os.path as osp import csv from flow.core.util import ensure_dir class AimsunKernelSimulation(KernelSimulation): """Aimsun simulation kernel...
45,420
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/kernel/simulation/base.py
52,53
4,86
simulation_step
KernelSimulation
57,59
53,59
class KernelSimulation(object): """Base simulation kernel. The simulation kernel is responsible for generating the simulation and passing to all other kernel the API that they can use to interact with the simulation. The simulation kernel is also responsible for advancing, resetting, and stori...
class KernelSimulation(object): """Base simulation kernel. The simulation kernel is responsible for generating the simulation and passing to all other kernel the API that they can use to interact with the simulation. The simulation kernel is also responsible for advancing, resetting, and stori...
raise NotImplementedError
"""Advance the simulation by one step. This is done in most cases by calling a relevant simulator API method. """ raise NotImplementedError
2
593
11
604
"""Script containing the base simulation kernel class.""" class KernelSimulation(object): """Base simulation kernel. The simulation kernel is responsible for generating the simulation and passing to all other kernel the API that they can use to interact with the simulation. The simulation kernel...
"""Script containing the base simulation kernel class.""" class KernelSimulation(object): """Base simulation kernel. The simulation kernel is responsible for generating the simulation and passing to all other kernel the API that they can use to interact with the simulation. The simulation kernel...
45,421
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/kernel/traffic_light/aimsun.py
56,57
5,73
get_state
AimsunKernelTrafficLight
72,73
57,73
class AimsunKernelTrafficLight(KernelTrafficLight): """Aimsun traffic light kernel. Implements all methods discussed in the base traffic light kernel class. """ def __init__(self, master_kernel): """Instantiate the sumo traffic light kernel. Parameters ---------- maste...
class AimsunKernelTrafficLight(KernelTrafficLight): """Aimsun traffic light kernel. Implements all methods discussed in the base traffic light kernel class. """ def __init__(self, master_kernel): """Instantiate the sumo traffic light kernel. Parameters ---------- maste...
return self.kernel_api.get_traffic_light_state(meter_aimsun_id)
"""Return the state of the traffic light(s) at the specified node. Parameters ---------- meter_aimsun_id: int aimsun id of the meter Returns ------- state : int desired state(s) for the traffic light 0: red 1: gree...
2
522
26
548
"""Script containing the base traffic light kernel class.""" from flow.core.kernel.traffic_light.base import KernelTrafficLight class AimsunKernelTrafficLight(KernelTrafficLight): """Aimsun traffic light kernel. Implements all methods discussed in the base traffic light kernel class. """ def __init_...
"""Script containing the base traffic light kernel class.""" from flow.core.kernel.traffic_light.base import KernelTrafficLight class AimsunKernelTrafficLight(KernelTrafficLight): """Aimsun traffic light kernel. Implements all methods discussed in the base traffic light kernel class. """ def __init_...
45,422
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/kernel/traffic_light/base.py
53,54
4,101
update
KernelTrafficLight
65,67
54,67
class KernelTrafficLight(object): """Base traffic light kernel. This kernel sub-class is used to interact with the simulator with regards to all traffic light -dependent components. Specifically, this class contains methods for: * Interacting with the simulator: This consisting specifying the stat...
class KernelTrafficLight(object): """Base traffic light kernel. This kernel sub-class is used to interact with the simulator with regards to all traffic light -dependent components. Specifically, this class contains methods for: * Interacting with the simulator: This consisting specifying the stat...
raise NotImplementedError
"""Update the states and phases of the traffic lights. This ensures that the traffic light variables match current traffic light data. Parameters ---------- reset : bool specifies whether the simulator was reset in the last simulation step ...
2
793
11
804
"""Script containing the base traffic light kernel class.""" class KernelTrafficLight(object): """Base traffic light kernel. This kernel sub-class is used to interact with the simulator with regards to all traffic light -dependent components. Specifically, this class contains methods for: * Inte...
"""Script containing the base traffic light kernel class.""" class KernelTrafficLight(object): """Base traffic light kernel. This kernel sub-class is used to interact with the simulator with regards to all traffic light -dependent components. Specifically, this class contains methods for: * Inte...
45,424
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/params.py
678,679
629,683
get_additional_param
EnvParams
680,683
679,683
class EnvParams: """Environment and experiment-specific parameters. This includes specifying the bounds of the action space and relevant coefficients to the reward function, as well as specifying how the positions of vehicles are modified in between rollouts. Attributes ---------- addition...
class EnvParams: """Environment and experiment-specific parameters. This includes specifying the bounds of the action space and relevant coefficients to the reward function, as well as specifying how the positions of vehicles are modified in between rollouts. Attributes ---------- addition...
return self.additional_params[key]
"""Return a variable from additional_params.""" return self.additional_params[key]
2
498
17
515
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
45,425
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/params.py
1146,1159
1136,1284
add
InFlows
1159,1281
1159,1281
class InFlows: """Used to add inflows to a network. Inflows can be specified for any edge that has a specified route or routes. """ def __init__(self): """Instantiate Inflows.""" self.__flows = [] def add(self, edge, veh_type, vehs_per_hour=None...
class InFlows: """Used to add inflows to a network. Inflows can be specified for any edge that has a specified route or routes. """ def __init__(self): """Instantiate Inflows.""" self.__flows = [] def add(self, edge, veh_type, vehs_per_hour=None...
r"""Specify a new inflow for a given type of vehicles and edge. Parameters ---------- edge : str starting edge for the vehicles in this inflow veh_type : str type of the vehicles entering the edge. Must match one of the types set in the Vehicl...
r"""Specify a new inflow for a given type of vehicles and edge. Parameters ---------- edge : str starting edge for the vehicles in this inflow veh_type : str type of the vehicles entering the edge. Must match one of the types set in the Vehicl...
643
198
1,519
1,717
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
45,426
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/params.py
70,81
49,219
add
TrafficLightParams
135,163
81,163
class TrafficLightParams: """Base traffic light. This class is used to place traffic lights in the network and describe the state of these traffic lights. In addition, this class supports modifying the states of certain lights via TraCI. """ def __init__(self, baseline=False): """Insta...
class TrafficLightParams: """Base traffic light. This class is used to place traffic lights in the network and describe the state of these traffic lights. In addition, this class supports modifying the states of certain lights via TraCI. """ def __init__(self, baseline=False): """Insta...
# prepare the data needed to generate xml files self.__tls_properties[node_id] = {"id": node_id, "type": tls_type} if programID: self.__tls_properties[node_id]["programID"] = programID if offset: self.__tls_properties[node_id]["offset"] = offset if phas...
"""Add a traffic light component to the network. When generating networks using xml files, using this method to add a traffic light will explicitly place the traffic light in the requested node of the generated network. If traffic lights are not added here but are already prese...
74
1,202
334
1,536
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
45,427
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/core/params.py
256,266
219,377
add
VehicleParams
293,366
266,366
class VehicleParams: """Base vehicle class. This is used to describe the state of all vehicles in the network. State information on the vehicles for a given time step can be set or retrieved from this class. """ def __init__(self): """Instantiate the base vehicle class.""" self...
class VehicleParams: """Base vehicle class. This is used to describe the state of all vehicles in the network. State information on the vehicles for a given time step can be set or retrieved from this class. """ def __init__(self): """Instantiate the base vehicle class.""" self...
if car_following_params is None: # FIXME: depends on simulator car_following_params = SumoCarFollowingParams() if lane_change_params is None: # FIXME: depends on simulator lane_change_params = SumoLaneChangeParams() type_params = {} type_...
"""Add a sequence of vehicles to the list of vehicles in the network. Parameters ---------- veh_id : str base vehicle ID for the vehicles (will be appended by a number) acceleration_controller : tup, optional 1st element: flow-specified acceleration contr...
210
822
780
1,602
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
"""Objects that define the various meta-parameters of an experiment.""" import logging import collections from flow.utils.flow_warnings import deprecated_attribute from flow.controllers.car_following_models import SimCarFollowingController from flow.controllers.rlcontroller import RLController from flow.controllers.l...
45,429
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/networks/bay_bridge_toll.py
55,56
32,98
specify_routes
BayBridgeTollNetwork
61,98
56,98
class BayBridgeTollNetwork(Network): """A network used to simulate the bottleneck portion of the Bay Bridge. The bay bridge was originally imported from OpenStreetMap and subsequently modified to more closely match the network geometry of the actual Bay Bridge. As opposed to BayBridgeNetwork, this netw...
class BayBridgeTollNetwork(Network): """A network used to simulate the bottleneck portion of the Bay Bridge. The bay bridge was originally imported from OpenStreetMap and subsequently modified to more closely match the network geometry of the actual Bay Bridge. As opposed to BayBridgeNetwork, this netw...
rts = { "11198593": ["11198593", "11198595.0"], "157598960": ["157598960", "11198595.0"], "11198595.0": ["11198595.0", "11198595.656.0"], "11198595.656.0": ["11198595.656.0", "gneE5"], "gneE5": ["gneE5", "340686911#2.0.13"], "124952171": ["...
"""See parent class. Routes for vehicles moving through the bay bridge from Oakland to San Francisco. """ rts = { "11198593": ["11198593", "11198595.0"], "157598960": ["157598960", "11198595.0"], "11198595.0": ["11198595.0", "11198595.656.0"],...
95
279
1,321
1,600
"""Contains the Bay Bridge toll network class.""" from flow.networks.base import Network # Use this to ensure that vehicles are only placed in the edges of the Bay # Bridge moving from Oakland to San Francisco. EDGES_DISTRIBUTION = [ '157598960', '11198599', '11198595.0', '11198595.656.0', '124952...
"""Contains the Bay Bridge toll network class.""" from flow.networks.base import Network # Use this to ensure that vehicles are only placed in the edges of the Bay # Bridge moving from Oakland to San Francisco. EDGES_DISTRIBUTION = [ '157598960', '11198599', '11198595.0', '11198595.656.0', '124952...
45,432
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/utils/flow_warnings.py
30,31
1,106
deprecated
36,106
31,106
"""Warnings that may be printed by Flow (e.g. deprecation warnings).""" import functools import inspect import warnings string_types = (type(b''), type(u'')) def deprecated_attribute(obj, dep_from, dep_to): """Print a deprecation warning. Parameters ---------- obj : class The class with the...
"""Warnings that may be printed by Flow (e.g. deprecation warnings).""" import functools import inspect import warnings string_types = (type(b''), type(u'')) def deprecated_attribute(obj, dep_from, dep_to): """Print a deprecation warning. Parameters ---------- obj : class The class with the...
# if isinstance(base, string_types): # The @deprecated is used with a 'reason'. # # .. code-block:: python # # @deprecated("please, use another function") # def old_function(x, y): # pass def decorator(func1): if inspect.isclass(func1): fmt1 = "The c...
"""Print a deprecation warning. This is a decorator which can be used to mark functions as deprecated. It will result in a warning being emitted when the function is used. """ # if isinstance(base, string_types): # The @deprecated is used with a 'reason'. # # .. code-block:: python ...
182
279
696
975
"""Warnings that may be printed by Flow (e.g. deprecation warnings).""" import functools import inspect import warnings string_types = (type(b''), type(u'')) def deprecated_attribute(obj, dep_from, dep_to): """Print a deprecation warning. Parameters ---------- obj : class The class with the...
"""Warnings that may be printed by Flow (e.g. deprecation warnings).""" import functools import inspect import warnings string_types = (type(b''), type(u'')) def deprecated_attribute(obj, dep_from, dep_to): """Print a deprecation warning. Parameters ---------- obj : class The class with the...
45,433
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/flow/utils/leaderboard/solution.py.template
13,14
1,22
get_actions
14,19
14,19
""" A solution should include: 1. Benchmark network; 2. get_actions() method; 3. get_states() method. """ # Specify benchmark network below. BENCHMARK = "" # Benchmark name goes here... # Specify get_action() method below. def get_actions(state): FILL_FUNC_BODYdef get_states(env, **kwargs): # get_st...
""" A solution should include: 1. Benchmark network; 2. get_actions() method; 3. get_states() method. """ # Specify benchmark network below. BENCHMARK = "" # Benchmark name goes here... # Specify get_action() method below. def get_actions(state): FILL_FUNC_BODYdef get_states(env, **kwargs): # get_st...
# get_actions() code goes here... return # Specify get_state() method below.
# get_actions() code goes here... return # Specify get_state() method below.
11
121
32
153
""" A solution should include: 1. Benchmark network; 2. get_actions() method; 3. get_states() method. """ # Specify benchmark network below. BENCHMARK = "" # Benchmark name goes here... # Specify get_action() method below. def get_actions(state): FILL_FUNC_BODYdef get_states(env, **kwargs): # get_st...
""" A solution should include: 1. Benchmark network; 2. get_actions() method; 3. get_states() method. """ # Specify benchmark network below. BENCHMARK = "" # Benchmark name goes here... # Specify get_action() method below. def get_actions(state): FILL_FUNC_BODYdef get_states(env, **kwargs): # get_st...
45,434
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/setup.py
11,12
1,52
_read_requirements_file
13,18
12,18
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): FILL...
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): ...
req_file_path = '%s/requirements.txt' % dirname(realpath(__file__)) with open(req_file_path) as f: return [line.strip() for line in f]
"""Return the elements in requirements.txt.""" req_file_path = '%s/requirements.txt' % dirname(realpath(__file__)) with open(req_file_path) as f: return [line.strip() for line in f]
15
437
60
497
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): FILL...
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): ...
45,435
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/setup.py
32,33
29,37
has_ext_modules
BinaryDistribution
34,37
33,37
class BinaryDistribution(Distribution): """See parent class.""" def has_ext_modules(self): FILL_FUNC_BODY
class BinaryDistribution(Distribution): """See parent class.""" def has_ext_modules(self): """Return True for external modules.""" FILL_FUNC_BODY
return True
"""Return True for external modules.""" return True
2
48
9
57
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): ...
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): ...
45,436
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow
/home/hieuvd/dung_repo/raw/stars_03/flow-project_flow/flow/setup.py
21,22
18,29
run
build_ext
23,29
22,29
class build_ext(_build_ext.build_ext): """External buid commands.""" def run(self): FILL_FUNC_BODY
class build_ext(_build_ext.build_ext): """External buid commands.""" def run(self): """Install traci wheels.""" FILL_FUNC_BODY
subprocess.check_call( ['python3','-m','pip', 'install', 'https://akreidieh.s3.amazonaws.com/sumo/flow-0.4.0/' 'sumotools-0.4.0-py3-none-any.whl'])
"""Install traci wheels.""" subprocess.check_call( ['python3','-m','pip', 'install', 'https://akreidieh.s3.amazonaws.com/sumo/flow-0.4.0/' 'sumotools-0.4.0-py3-none-any.whl'])
5
48
85
133
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): ...
#!/usr/bin/env python3 # flake8: noqa """Setup script for the Flow repository.""" from os.path import dirname, realpath from setuptools import find_packages, setup, Distribution import setuptools.command.build_ext as _build_ext import subprocess from flow.version import __version__ def _read_requirements_file(): ...
45,460
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/demo/image_demo.py
7,8
1,30
main
8,28
8,28
from argparse import ArgumentParser from mmseg.apis import inference_segmentor, init_segmentor, show_result_pyplot from mmseg.core.evaluation import get_palette def main(): FILL_FUNC_BODYif __name__ == '__main__': main()
from argparse import ArgumentParser from mmseg.apis import inference_segmentor, init_segmentor, show_result_pyplot from mmseg.core.evaluation import get_palette def main(): FILL_FUNC_BODYif __name__ == '__main__': main()
parser = ArgumentParser() parser.add_argument('img', help='Image file') parser.add_argument('config', help='Config file') parser.add_argument('checkpoint', help='Checkpoint file') parser.add_argument( '--device', default='cuda:0', help='Device used for inference') parser.add_argument( ...
parser = ArgumentParser() parser.add_argument('img', help='Image file') parser.add_argument('config', help='Config file') parser.add_argument('checkpoint', help='Checkpoint file') parser.add_argument( '--device', default='cuda:0', help='Device used for inference') parser.add_argument( ...
65
79
254
333
from argparse import ArgumentParser from mmseg.apis import inference_segmentor, init_segmentor, show_result_pyplot from mmseg.core.evaluation import get_palette def main(): FILL_FUNC_BODYif __name__ == '__main__': main()
from argparse import ArgumentParser from mmseg.apis import inference_segmentor, init_segmentor, show_result_pyplot from mmseg.core.evaluation import get_palette def main(): FILL_FUNC_BODYif __name__ == '__main__': main()
45,461
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/docs/conf.py
26,27
1,80
get_version
27,33
27,33
# Configuration file for the Sphinx documentation builder. # # This file only contains a selection of the most common options. For a full # list see the documentation: # https://www.sphinx-doc.org/en/master/usage/configuration.html # -- Path setup -------------------------------------------------------------- # If ex...
# Configuration file for the Sphinx documentation builder. # # This file only contains a selection of the most common options. For a full # list see the documentation: # https://www.sphinx-doc.org/en/master/usage/configuration.html # -- Path setup -------------------------------------------------------------- # If ex...
with open(version_file, 'r') as f: exec(compile(f.read(), version_file, 'exec')) return locals()['__version__'] # The full version, including alpha/beta/rc tags
with open(version_file, 'r') as f: exec(compile(f.read(), version_file, 'exec')) return locals()['__version__'] # The full version, including alpha/beta/rc tags
17
692
63
755
# Configuration file for the Sphinx documentation builder. # # This file only contains a selection of the most common options. For a full # list see the documentation: # https://www.sphinx-doc.org/en/master/usage/configuration.html # -- Path setup -------------------------------------------------------------- # If ex...
# Configuration file for the Sphinx documentation builder. # # This file only contains a selection of the most common options. For a full # list see the documentation: # https://www.sphinx-doc.org/en/master/usage/configuration.html # -- Path setup -------------------------------------------------------------- # If ex...
45,462
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/core/bbox/coder/distance_point_bbox_coder.py
41,42
7,64
decode
DistancePointBBoxCoder
58,64
42,64
@BBOX_CODERS.register_module() class DistancePointBBoxCoder(BaseBBoxCoder): """Distance Point BBox coder. This coder encodes gt bboxes (x1, y1, x2, y2) into (top, bottom, left, right) and decode it back to the original. Args: clip_border (bool, optional): Whether clip the objects outside the ...
@BBOX_CODERS.register_module() class DistancePointBBoxCoder(BaseBBoxCoder): """Distance Point BBox coder. This coder encodes gt bboxes (x1, y1, x2, y2) into (top, bottom, left, right) and decode it back to the original. Args: clip_border (bool, optional): Whether clip the objects outside the ...
assert points.size(0) == pred_bboxes.size(0) assert points.size(-1) == 2 assert pred_bboxes.size(-1) == 4 if self.clip_border is False: max_shape = None return distance2bbox(points, pred_bboxes, max_shape)
"""Decode distance prediction to bounding box. Args: points (Tensor): Shape (B, N, 2) or (N, 2). pred_bboxes (Tensor): Distance from the given point to 4 boundaries (left, top, right, bottom). Shape (B, N, 4) or (N, 4) max_shape (Seque...
23
668
90
758
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import BBOX_CODERS from ..transforms import bbox2distance, distance2bbox from .base_bbox_coder import BaseBBoxCoder @BBOX_CODERS.register_module() class DistancePointBBoxCoder(BaseBBoxCoder): """Distance Point BBox coder. This coder encodes gt bb...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import BBOX_CODERS from ..transforms import bbox2distance, distance2bbox from .base_bbox_coder import BaseBBoxCoder @BBOX_CODERS.register_module() class DistancePointBBoxCoder(BaseBBoxCoder): """Distance Point BBox coder. This coder encodes gt bb...
45,463
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/core/bbox/coder/pseudo_bbox_coder.py
17,18
6,20
decode
PseudoBBoxCoder
19,20
18,20
@BBOX_CODERS.register_module() class PseudoBBoxCoder(BaseBBoxCoder): """Pseudo bounding box coder.""" def __init__(self, **kwargs): super(BaseBBoxCoder, self).__init__(**kwargs) def encode(self, bboxes, gt_bboxes): """torch.Tensor: return the given ``bboxes``""" return gt_bboxes ...
@BBOX_CODERS.register_module() class PseudoBBoxCoder(BaseBBoxCoder): """Pseudo bounding box coder.""" def __init__(self, **kwargs): super(BaseBBoxCoder, self).__init__(**kwargs) def encode(self, bboxes, gt_bboxes): """torch.Tensor: return the given ``bboxes``""" return gt_bboxes ...
return pred_bboxes
"""torch.Tensor: return the given ``pred_bboxes``""" return pred_bboxes
2
167
9
176
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import BBOX_CODERS from .base_bbox_coder import BaseBBoxCoder @BBOX_CODERS.register_module() class PseudoBBoxCoder(BaseBBoxCoder): """Pseudo bounding box coder.""" def __init__(self, **kwargs): super(BaseBBoxCoder, self).__init__(**kwargs...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import BBOX_CODERS from .base_bbox_coder import BaseBBoxCoder @BBOX_CODERS.register_module() class PseudoBBoxCoder(BaseBBoxCoder): """Pseudo bounding box coder.""" def __init__(self, **kwargs): super(BaseBBoxCoder, self).__init__(**kwargs...
45,464
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/core/bbox/samplers/combined_sampler.py
19,20
6,22
_sample_neg
CombinedSampler
21,22
20,22
@BBOX_SAMPLERS.register_module() class CombinedSampler(BaseSampler): """A sampler that combines positive sampler and negative sampler.""" def __init__(self, pos_sampler, neg_sampler, **kwargs): super(CombinedSampler, self).__init__(**kwargs) self.pos_sampler = build_sampler(pos_sampler, **kwarg...
@BBOX_SAMPLERS.register_module() class CombinedSampler(BaseSampler): """A sampler that combines positive sampler and negative sampler.""" def __init__(self, pos_sampler, neg_sampler, **kwargs): super(CombinedSampler, self).__init__(**kwargs) self.pos_sampler = build_sampler(pos_sampler, **kwarg...
raise NotImplementedError
"""Sample negative samples.""" raise NotImplementedError
2
185
9
194
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import BBOX_SAMPLERS, build_sampler from .base_sampler import BaseSampler @BBOX_SAMPLERS.register_module() class CombinedSampler(BaseSampler): """A sampler that combines positive sampler and negative sampler.""" def __init__(self, pos_sampler, ne...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import BBOX_SAMPLERS, build_sampler from .base_sampler import BaseSampler @BBOX_SAMPLERS.register_module() class CombinedSampler(BaseSampler): """A sampler that combines positive sampler and negative sampler.""" def __init__(self, pos_sampler, ne...
45,465
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/core/evaluation/bbox_overlaps.py
5,10
1,66
bbox_overlaps
28,66
10,66
# Copyright (c) OpenMMLab. All rights reserved. import numpy as np def bbox_overlaps(bboxes1, bboxes2, mode='iou', eps=1e-6, use_legacy_coordinate=False): FILL_FUNC_BODY
# Copyright (c) OpenMMLab. All rights reserved. import numpy as np def bbox_overlaps(bboxes1, bboxes2, mode='iou', eps=1e-6, use_legacy_coordinate=False): """Calculate the ious between each bbox of bboxes1 and bboxes2. Args: bbox...
assert mode in ['iou', 'iof'] if not use_legacy_coordinate: extra_length = 0. else: extra_length = 1. bboxes1 = bboxes1.astype(np.float32) bboxes2 = bboxes2.astype(np.float32) rows = bboxes1.shape[0] cols = bboxes2.shape[0] ious = np.zeros((rows, cols), dtype=np.float32)...
"""Calculate the ious between each bbox of bboxes1 and bboxes2. Args: bboxes1 (ndarray): Shape (n, 4) bboxes2 (ndarray): Shape (k, 4) mode (str): IOU (intersection over union) or IOF (intersection over foreground) use_legacy_coordinate (bool): Whether to use coordina...
177
338
655
993
# Copyright (c) OpenMMLab. All rights reserved. import numpy as np def bbox_overlaps(bboxes1, bboxes2, mode='iou', eps=1e-6, use_legacy_coordinate=False): FILL_FUNC_BODY
# Copyright (c) OpenMMLab. All rights reserved. import numpy as np def bbox_overlaps(bboxes1, bboxes2, mode='iou', eps=1e-6, use_legacy_coordinate=False): """Calculate the ious between each bbox of bboxes1 and bboxes2. Args: bbox...
45,466
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/datasets/pipelines/instaboost.py
56,57
7,119
_load_anns
InstaBoost
57,78
57,78
@PIPELINES.register_module() class InstaBoost: r"""Data augmentation method in `InstaBoost: Boosting Instance Segmentation Via Probability Map Guided Copy-Pasting <https://arxiv.org/abs/1908.07801>`_. Refer to https://github.com/GothicAi/Instaboost for implementation details. Args: action_...
@PIPELINES.register_module() class InstaBoost: r"""Data augmentation method in `InstaBoost: Boosting Instance Segmentation Via Probability Map Guided Copy-Pasting <https://arxiv.org/abs/1908.07801>`_. Refer to https://github.com/GothicAi/Instaboost for implementation details. Args: action_...
labels = results['ann_info']['labels'] masks = results['ann_info']['masks'] bboxes = results['ann_info']['bboxes'] n = len(labels) anns = [] for i in range(n): label = labels[i] bbox = bboxes[i] mask = masks[i] x1, y1, x2, ...
labels = results['ann_info']['labels'] masks = results['ann_info']['masks'] bboxes = results['ann_info']['bboxes'] n = len(labels) anns = [] for i in range(n): label = labels[i] bbox = bboxes[i] mask = masks[i] x1, y1, x2, ...
67
1,182
221
1,403
# Copyright (c) OpenMMLab. All rights reserved. import numpy as np from ..builder import PIPELINES @PIPELINES.register_module() class InstaBoost: r"""Data augmentation method in `InstaBoost: Boosting Instance Segmentation Via Probability Map Guided Copy-Pasting <https://arxiv.org/abs/1908.07801>`_. ...
# Copyright (c) OpenMMLab. All rights reserved. import numpy as np from ..builder import PIPELINES @PIPELINES.register_module() class InstaBoost: r"""Data augmentation method in `InstaBoost: Boosting Instance Segmentation Via Probability Map Guided Copy-Pasting <https://arxiv.org/abs/1908.07801>`_. ...
45,467
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/models/detectors/cascade_rcnn.py
30,31
6,50
show_result
CascadeRCNN
41,50
31,50
@DETECTORS.register_module() class CascadeRCNN(TwoStageDetector): r"""Implementation of `Cascade R-CNN: Delving into High Quality Object Detection <https://arxiv.org/abs/1906.09756>`_""" def __init__(self, backbone, neck=None, rpn_head=None, ...
@DETECTORS.register_module() class CascadeRCNN(TwoStageDetector): r"""Implementation of `Cascade R-CNN: Delving into High Quality Object Detection <https://arxiv.org/abs/1906.09756>`_""" def __init__(self, backbone, neck=None, rpn_head=None, ...
if self.with_mask: ms_bbox_result, ms_segm_result = result if isinstance(ms_bbox_result, dict): result = (ms_bbox_result['ensemble'], ms_segm_result['ensemble']) else: if isinstance(result, dict): result = resu...
"""Show prediction results of the detector. Args: data (str or np.ndarray): Image filename or loaded image. result (Tensor or tuple): The results to draw over `img` bbox_result or (bbox_result, segm_result). Returns: np.ndarray: The image wit...
25
356
127
483
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .two_stage import TwoStageDetector @DETECTORS.register_module() class CascadeRCNN(TwoStageDetector): r"""Implementation of `Cascade R-CNN: Delving into High Quality Object Detection <https://arxiv.org/abs/1906.09756>`_""" ...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .two_stage import TwoStageDetector @DETECTORS.register_module() class CascadeRCNN(TwoStageDetector): r"""Implementation of `Cascade R-CNN: Delving into High Quality Object Detection <https://arxiv.org/abs/1906.09756>`_""" ...
45,468
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/models/detectors/fast_rcnn.py
27,28
6,56
forward_test
FastRCNN
41,56
28,56
@DETECTORS.register_module() class FastRCNN(TwoStageDetector): """Implementation of `Fast R-CNN <https://arxiv.org/abs/1504.08083>`_""" def __init__(self, backbone, roi_head, train_cfg, test_cfg, neck=None, pr...
@DETECTORS.register_module() class FastRCNN(TwoStageDetector): """Implementation of `Fast R-CNN <https://arxiv.org/abs/1504.08083>`_""" def __init__(self, backbone, roi_head, train_cfg, test_cfg, neck=None, pr...
for var, name in [(imgs, 'imgs'), (img_metas, 'img_metas')]: if not isinstance(var, list): raise TypeError(f'{name} must be a list, but got {type(var)}') num_augs = len(imgs) if num_augs != len(img_metas): raise ValueError(f'num of augmentations ({len(img...
""" Args: imgs (List[Tensor]): the outer list indicates test-time augmentations and inner Tensor should have a shape NxCxHxW, which contains all images in the batch. img_metas (List[List[dict]]): the outer list indicates test-time a...
57
394
212
606
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .two_stage import TwoStageDetector @DETECTORS.register_module() class FastRCNN(TwoStageDetector): """Implementation of `Fast R-CNN <https://arxiv.org/abs/1504.08083>`_""" def __init__(self, backbone, ...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .two_stage import TwoStageDetector @DETECTORS.register_module() class FastRCNN(TwoStageDetector): """Implementation of `Fast R-CNN <https://arxiv.org/abs/1504.08083>`_""" def __init__(self, backbone, ...
45,469
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/models/detectors/htc.py
13,15
6,17
with_semantic
HybridTaskCascade
16,17
15,17
@DETECTORS.register_module() class HybridTaskCascade(CascadeRCNN): """Implementation of `HTC <https://arxiv.org/abs/1901.07518>`_""" def __init__(self, **kwargs): super(HybridTaskCascade, self).__init__(**kwargs) @property def with_semantic(self): FILL_FUNC_BODY
@DETECTORS.register_module() class HybridTaskCascade(CascadeRCNN): """Implementation of `HTC <https://arxiv.org/abs/1901.07518>`_""" def __init__(self, **kwargs): super(HybridTaskCascade, self).__init__(**kwargs) @property def with_semantic(self): """bool: whether the detector has a se...
return self.roi_head.with_semantic
"""bool: whether the detector has a semantic head""" return self.roi_head.with_semantic
2
128
15
143
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .cascade_rcnn import CascadeRCNN @DETECTORS.register_module() class HybridTaskCascade(CascadeRCNN): """Implementation of `HTC <https://arxiv.org/abs/1901.07518>`_""" def __init__(self, **kwargs): super(HybridTaskCasc...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .cascade_rcnn import CascadeRCNN @DETECTORS.register_module() class HybridTaskCascade(CascadeRCNN): """Implementation of `HTC <https://arxiv.org/abs/1901.07518>`_""" def __init__(self, **kwargs): super(HybridTaskCasc...
45,470
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR
/home/hieuvd/dung_repo/raw/stars_03/fudan-zvg_SETR/SETR/hlg-detection/mmdet/models/detectors/sparse_rcnn.py
67,68
6,112
simple_test
SparseRCNN
81,93
68,93
@DETECTORS.register_module() class SparseRCNN(TwoStageDetector): r"""Implementation of `Sparse R-CNN: End-to-End Object Detection with Learnable Proposals <https://arxiv.org/abs/2011.12450>`_""" def __init__(self, *args, **kwargs): super(SparseRCNN, self).__init__(*args, **kwargs) assert se...
@DETECTORS.register_module() class SparseRCNN(TwoStageDetector): r"""Implementation of `Sparse R-CNN: End-to-End Object Detection with Learnable Proposals <https://arxiv.org/abs/2011.12450>`_""" def __init__(self, *args, **kwargs): super(SparseRCNN, self).__init__(*args, **kwargs) assert se...
x = self.extract_feat(img) proposal_boxes, proposal_features, imgs_whwh = \ self.rpn_head.simple_test_rpn(x, img_metas) results = self.roi_head.simple_test( x, proposal_boxes, proposal_features, img_metas, imgs_whwh=imgs_whw...
"""Test function without test time augmentation. Args: imgs (list[torch.Tensor]): List of multiple images img_metas (list[dict]): List of image information. rescale (bool): Whether to rescale the results. Defaults to False. Returns: ...
21
1,149
133
1,282
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .two_stage import TwoStageDetector @DETECTORS.register_module() class SparseRCNN(TwoStageDetector): r"""Implementation of `Sparse R-CNN: End-to-End Object Detection with Learnable Proposals <https://arxiv.org/abs/2011.12450>`...
# Copyright (c) OpenMMLab. All rights reserved. from ..builder import DETECTORS from .two_stage import TwoStageDetector @DETECTORS.register_module() class SparseRCNN(TwoStageDetector): r"""Implementation of `Sparse R-CNN: End-to-End Object Detection with Learnable Proposals <https://arxiv.org/abs/2011.12450>`...