SYMBOL INDEX (110 symbols across 5 files) FILE: imagebind/data.py function return_bpe_path (line 26) | def return_bpe_path(): function waveform2melspec (line 32) | def waveform2melspec(waveform, sample_rate, num_mel_bins, target_length): function get_clip_timepoints (line 71) | def get_clip_timepoints(clip_sampler, duration): function load_and_transform_vision_data (line 82) | def load_and_transform_vision_data(image_paths, device): function load_and_transform_text (line 109) | def load_and_transform_text(text, device): function load_and_transform_audio_data (line 118) | def load_and_transform_audio_data( function crop_boxes (line 168) | def crop_boxes(boxes, x_offset, y_offset): function uniform_crop (line 187) | def uniform_crop(images, size, spatial_idx, boxes=None, scale_size=None): class SpatialCrop (line 246) | class SpatialCrop(nn.Module): method __init__ (line 255) | def __init__(self, crop_size: int = 224, num_crops: int = 3): method forward (line 267) | def forward(self, videos): class NormalizeVideo (line 289) | class NormalizeVideo: method __init__ (line 290) | def __init__(self, mean, std, inplace=False): method __call__ (line 295) | def __call__(self, clip): function load_and_transform_video_data (line 304) | def load_and_transform_video_data( FILE: imagebind/models/helpers.py class Normalize (line 15) | class Normalize(nn.Module): method __init__ (line 16) | def __init__(self, dim: int) -> None: method forward (line 20) | def forward(self, x): class LearnableLogitScaling (line 24) | class LearnableLogitScaling(nn.Module): method __init__ (line 25) | def __init__( method forward (line 41) | def forward(self, x): method extra_repr (line 44) | def extra_repr(self): class EinOpsRearrange (line 50) | class EinOpsRearrange(nn.Module): method __init__ (line 51) | def __init__(self, rearrange_expr: str, **kwargs) -> None: method forward (line 56) | def forward(self, x): class VerboseNNModule (line 61) | class VerboseNNModule(nn.Module): method get_readable_tensor_repr (line 67) | def get_readable_tensor_repr(name: str, tensor: torch.Tensor) -> str: method extra_repr (line 80) | def extra_repr(self) -> str: function cast_if_src_dtype (line 99) | def cast_if_src_dtype( class QuickGELU (line 109) | class QuickGELU(nn.Module): method forward (line 111) | def forward(self, x: torch.Tensor): class SelectElement (line 115) | class SelectElement(nn.Module): method __init__ (line 116) | def __init__(self, index) -> None: method forward (line 120) | def forward(self, x): class SelectEOSAndProject (line 125) | class SelectEOSAndProject(nn.Module): method __init__ (line 130) | def __init__(self, proj: nn.Module) -> None: method forward (line 134) | def forward(self, x, seq_len): FILE: imagebind/models/imagebind_model.py class ImageBindModel (line 37) | class ImageBindModel(nn.Module): method __init__ (line 38) | def __init__( method _create_modality_preprocessors (line 131) | def _create_modality_preprocessors( method _create_modality_trunks (line 265) | def _create_modality_trunks( method _create_modality_heads (line 366) | def _create_modality_heads( method _create_modality_postprocessors (line 418) | def _create_modality_postprocessors(self, out_embed_dim): method forward (line 444) | def forward(self, inputs): function imagebind_huge (line 479) | def imagebind_huge(pretrained=False): FILE: imagebind/models/multimodal_preprocessors.py function get_sinusoid_encoding_table (line 26) | def get_sinusoid_encoding_table(n_position, d_hid): function interpolate_pos_encoding_2d (line 45) | def interpolate_pos_encoding_2d(target_spatial_size, pos_embed): function interpolate_pos_encoding (line 65) | def interpolate_pos_encoding( function _get_pos_embedding (line 104) | def _get_pos_embedding( class PatchEmbedGeneric (line 121) | class PatchEmbedGeneric(nn.Module): method __init__ (line 126) | def __init__(self, proj_stem, norm_layer: Optional[nn.Module] = None): method get_patch_layout (line 137) | def get_patch_layout(self, img_size): method forward (line 151) | def forward(self, x): class SpatioTemporalPosEmbeddingHelper (line 160) | class SpatioTemporalPosEmbeddingHelper(VerboseNNModule): method __init__ (line 161) | def __init__( method get_pos_embedding (line 183) | def get_pos_embedding(self, vision_input, all_vision_tokens): class RGBDTPreprocessor (line 195) | class RGBDTPreprocessor(VerboseNNModule): method __init__ (line 196) | def __init__( method init_parameters (line 236) | def init_parameters(self, init_param_style): method tokenize_input_and_cls_pos (line 255) | def tokenize_input_and_cls_pos(self, input, stem, mask): method forward (line 273) | def forward(self, vision=None, depth=None, patch_mask=None): class AudioPreprocessor (line 301) | class AudioPreprocessor(RGBDTPreprocessor): method __init__ (line 302) | def __init__(self, audio_stem: PatchEmbedGeneric, **kwargs) -> None: method forward (line 305) | def forward(self, audio=None): class ThermalPreprocessor (line 309) | class ThermalPreprocessor(RGBDTPreprocessor): method __init__ (line 310) | def __init__(self, thermal_stem: PatchEmbedGeneric, **kwargs) -> None: method forward (line 313) | def forward(self, thermal=None): function build_causal_attention_mask (line 317) | def build_causal_attention_mask(context_length): class TextPreprocessor (line 326) | class TextPreprocessor(VerboseNNModule): method __init__ (line 327) | def __init__( method init_parameters (line 362) | def init_parameters(self, init_param_style="openclip"): method forward (line 378) | def forward(self, text): class Im2Video (line 406) | class Im2Video(nn.Module): method __init__ (line 409) | def __init__(self, time_dim=2): method forward (line 413) | def forward(self, x): class PadIm2Video (line 423) | class PadIm2Video(Im2Video): method __init__ (line 424) | def __init__(self, ntimes, pad_type, time_dim=2): method forward (line 431) | def forward(self, x): function bytes_to_unicode (line 447) | def bytes_to_unicode(): function get_pairs (line 473) | def get_pairs(word): function basic_clean (line 485) | def basic_clean(text): function whitespace_clean (line 491) | def whitespace_clean(text): class SimpleTokenizer (line 497) | class SimpleTokenizer(object): method __init__ (line 498) | def __init__(self, bpe_path: str, context_length=77): method bpe (line 525) | def bpe(self, token): method encode (line 566) | def encode(self, text): method decode (line 576) | def decode(self, tokens): method __call__ (line 585) | def __call__(self, texts, context_length=None): class IMUPreprocessor (line 606) | class IMUPreprocessor(VerboseNNModule): method __init__ (line 607) | def __init__( method init_parameters (line 635) | def init_parameters(self, init_param_style): method tokenize_input_and_cls_pos (line 650) | def tokenize_input_and_cls_pos(self, input, stem): method forward (line 665) | def forward(self, imu): FILE: imagebind/models/transformer.py class Attention (line 23) | class Attention(nn.Module): method __init__ (line 24) | def __init__( method forward (line 45) | def forward(self, x): class Mlp (line 68) | class Mlp(nn.Module): method __init__ (line 69) | def __init__( method forward (line 85) | def forward(self, x): class MultiheadAttention (line 94) | class MultiheadAttention(nn.MultiheadAttention): method forward (line 95) | def forward(self, x: torch.Tensor, attn_mask: torch.Tensor): class ViTAttention (line 99) | class ViTAttention(Attention): method forward (line 100) | def forward(self, x: torch.Tensor, attn_mask: torch.Tensor): class BlockWithMasking (line 105) | class BlockWithMasking(nn.Module): method __init__ (line 106) | def __init__( method forward (line 159) | def forward(self, x: torch.Tensor, attn_mask: torch.Tensor): class SimpleTransformer (line 176) | class SimpleTransformer(nn.Module): method __init__ (line 177) | def __init__( method _init_weights (line 230) | def _init_weights(self, m): method forward (line 245) | def forward(