MCPcopy Create free account

hub / github.com/OpenDriveLab/ReSim / functions

Functions2,541 in github.com/OpenDriveLab/ReSim

⨍ Functions 2,541◇ Types & classes 556↳ Endpoints 3
↓ 354 callersMethodappend
(self, other)
SwissArmyTransformer/sat/tokenization/glm/tokenization.py:89
↓ 317 callersFunctionprint
(*args, **kwargs)
SwissArmyTransformer/examples/mae/util/misc.py:176
↓ 241 callersMethodto
(self, device)
SwissArmyTransformer/examples/yolos/util/misc.py:288
↓ 124 callersMethodget
(self, name)
SwissArmyTransformer/sat/model/registry.py:21
↓ 111 callersMethodadd_mixin
(self, name, new_mixin, reinit=False)
SwissArmyTransformer/sat/model/base_model.py:114
↓ 96 callersMethodmax
(self)
SwissArmyTransformer/examples/mae/util/misc.py:70
↓ 93 callersMethodfrom_pretrained
(cls, name, args=None, base_cls=None, *, home_path=None, url=None, prefix='', **kwargs)
SwissArmyTransformer/examples/chatglm/chat_model.py:43
↓ 79 callersFunctionprint_rank0
(msg, level=logging.INFO, flush=True)
SwissArmyTransformer/sat/helpers.py:127
↓ 62 callersMethodstart
Start the timer.
SwissArmyTransformer/sat/training/utils.py:79
↓ 61 callersMethodstop
Stop the timer.
SwissArmyTransformer/sat/training/utils.py:86
↓ 60 callersMethodparameters
(self)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:1324
↓ 52 callersFunctionget_args
Parse all the args.
SwissArmyTransformer/sat/arguments.py:339
↓ 51 callersMethodextend
(self, other)
SwissArmyTransformer/sat/tokenization/glm/tokenization.py:102
↓ 49 callersMethodlog
Log a group of timers.
SwissArmyTransformer/sat/training/utils.py:122
↓ 42 callersFunctionappend_dims
(t, ndims: int)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:76
↓ 33 callersMethodload
(self, verbose=True)
SwissArmyTransformer/sat/ops/ops_builder/builder.py:434
↓ 32 callersFunctioninstantiate_from_config
(config, **extra_kwargs)
sat/sgm/util.py:254
↓ 31 callersMethodupdate
(self, **kwargs)
SwissArmyTransformer/examples/mae/util/misc.py:91
↓ 30 callersMethod_transpose_for_scores
Transpose a 3D tensor [b, s, np*hn] into a 4D tensor with size [b, np, s, hn].
SwissArmyTransformer/sat/model/transformer.py:99
↓ 30 callersFunctionget_model_parallel_world_size
Return world size for the model parallel group.
SwissArmyTransformer/sat/mpu/initialize.py:130
↓ 29 callersFunctionexists
(v)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:42
↓ 29 callersMethodpad
(self)
SwissArmyTransformer/sat/tokenization/icetk_glm_130B/tokenizer.py:66
↓ 28 callersFunctiontraining_main
Main training program.
SwissArmyTransformer/sat/training/deepspeed_training.py:49
↓ 23 callersFunctiondefault
(v, d)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:46
↓ 23 callersMethodexists
(model_path)
SwissArmyTransformer/sat/tokenization/glm/sp_tokenizer.py:69
↓ 22 callersFunctionSequential
(*modules)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:177
↓ 21 callersMethod__init__
(self)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:480
↓ 21 callersMethoddel_mixin
(self, name)
SwissArmyTransformer/sat/model/base_model.py:125
↓ 21 callersMethoddevice
(self)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:1308
↓ 20 callersMethoddecode
(self, input_ids, position_ids, attention_mask, encoder_outputs, ids_restore, **kw_args)
SwissArmyTransformer/sat/model/official/mae_model.py:153
↓ 20 callersFunctiondivide
Ensure that numerator is divisible by the denominator and return the division value.
SwissArmyTransformer/sat/mpu/utils.py:27
↓ 20 callersMethodinsert
(self, idx, other)
SwissArmyTransformer/sat/tokenization/glm/tokenization.py:75
↓ 20 callersMethodload
(self, path, strict=True)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:1367
↓ 18 callersMethodget_mixin
(self, name)
SwissArmyTransformer/sat/model/base_model.py:130
↓ 17 callersFunctionget_tokenizer
If you're using outer_tokenizer, call `get_tokenizer(args, outer_tokenizer)` before `training_main`.
SwissArmyTransformer/sat/tokenization/__init__.py:19
↓ 17 callersFunctionlinear
Create a linear module.
sat/sgm/modules/diffusionmodules/util.py:261
↓ 17 callersMethodstrip_empty_entries
Drop any empty strings from the list of compile and link flags
SwissArmyTransformer/sat/ops/ops_builder/builder.py:312
↓ 16 callersFunctionconv_nd
Create a 1D, 2D, or 3D convolution module.
sat/sgm/modules/diffusionmodules/util.py:248
↓ 16 callersMethodfrom_pretrained
(cls, name, args=None, base_cls=None, *, home_path=None, url=None, prefix='', **kwargs)
SwissArmyTransformer/examples/chatglm2/chat_model.py:42
↓ 16 callersMethodget_command
get command token corresponding to `name`
SwissArmyTransformer/sat/tokenization/hf_tokenizer.py:38
↓ 16 callersFunctionget_model_parallel_rank
Return my rank for the model parallel group.
SwissArmyTransformer/sat/mpu/initialize.py:140
↓ 15 callersMethodsave
(self, path, overwrite=True)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:1359
↓ 15 callersFunctionsplit_tensor_along_last_dim
Split a tensor along its last dimension. Arguments: tensor: input tensor. num_partitions: number of partitions to split the tensor
SwissArmyTransformer/sat/mpu/utils.py:34
↓ 14 callersFunctionget_model_parallel_group
Get the model parallel group the caller rank belongs to.
SwissArmyTransformer/sat/mpu/initialize.py:109
↓ 14 callersFunctionsave_checkpoint
Save a model checkpoint.
SwissArmyTransformer/sat/training/model_io.py:162
↓ 13 callersFunctionload_hf_dataset
(path, process_fn, columns=None, cache_dir='~/.cache/huggingface/datasets', offline=False, transformer_name =
SwissArmyTransformer/sat/data_utils/hf_dataset.py:21
↓ 13 callersFunctionprint
(*args, **kwargs)
SwissArmyTransformer/examples/yolos/util/misc.py:369
↓ 12 callersMethodadd_model_specific_args
(cls, parser)
SwissArmyTransformer/examples/clip/clip_finetune_model.py:30
↓ 12 callersMethodattention
(self, h_: torch.Tensor)
sat/sgm/modules/diffusionmodules/model.py:155
↓ 12 callersFunctiondefault
(val, d)
sat/sgm/modules/attention.py:64
↓ 12 callersMethodfrom_pretrained
(cls, name, args=None, *, home_path=None, url=None, prefix='', build_only=False, use_node_group=True, overwrit
SwissArmyTransformer/sat/model/base_model.py:215
↓ 12 callersFunctionnormalization
Make a standard normalization layer. :param channels: number of input channels. :return: an nn.Module for normalization.
sat/sgm/modules/diffusionmodules/util.py:228
↓ 12 callersFunctionprint_all
(msg, level=logging.INFO, flush=True)
SwissArmyTransformer/sat/helpers.py:139
↓ 11 callersMethod__init__
(self, in_channels=3, model_channels=64)
sat/sgm/modules/diffusionmodules/openaimodel.py:1225
↓ 11 callersMethoddenoise
(self, x, denoiser, alpha_cumprod_sqrt, cond, uc, timestep=None, idx=None, scale=None, scale_emb=None, **addit
sat/sgm/modules/diffusionmodules/sampling.py:529
↓ 11 callersFunctiondestroy_model_parallel
Set the groups to none.
SwissArmyTransformer/sat/mpu/initialize.py:176
↓ 11 callersFunctioninitialize_model_parallel
Initialize model data parallel groups. Arguments: model_parallel_size: number of GPUs used to parallelize model. Let's say we h
SwissArmyTransformer/sat/mpu/initialize.py:34
↓ 11 callersMethodload_state_dict
(self, *args, **kwargs)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:1356
↓ 11 callersFunctionnonlinearity
(x)
SwissArmyTransformer/sat/tokenization/cogview/vqvae/vqvae_diffusion.py:29
↓ 10 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/cait/transform_param.py:24
↓ 10 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/mixtral-8x7b/transform_param.py:47
↓ 10 callersMethodencode
(self, input_ids, position_ids, attention_mask=None, **kw_args)
SwissArmyTransformer/sat/model/official/mae_model.py:150
↓ 10 callersMethodfrom_pretrained_base
Load a pretrained checkpoint of the current model. Args: name: The identifier of the pretrained model. arg
SwissArmyTransformer/sat/model/base_model.py:187
↓ 10 callersMethodget_sigma_gen
(self, num_sigmas)
sat/sgm/modules/diffusionmodules/sampling.py:62
↓ 10 callersMethodload_state_dict
(self, state_dict: Dict[str, Any])
SwissArmyTransformer/examples/yolos/util/scheduler.py:63
↓ 10 callersFunctionpack_one
(t, pattern)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:68
↓ 10 callersMethodprepare_sampling_loop
(self, x, cond, uc=None, num_steps=None)
sat/sgm/modules/diffusionmodules/sampling.py:510
↓ 10 callersFunctionunpack_one
(t, ps, pattern)
sat/sgm/modules/autoencoding/magvit2_pytorch.py:72
↓ 9 callersFunctionNormalize
(in_channels)
SwissArmyTransformer/sat/tokenization/cogview/vqvae/vqvae_diffusion.py:34
↓ 9 callersMethod__init__
(self, *, ch, out_ch, ch_mult=(1,2,4,8), num_res_blocks, attn_resolutions, dropout=0.0, resam
SwissArmyTransformer/sat/tokenization/cogview/vqvae/vqvae_diffusion.py:196
↓ 9 callersMethodbackward
(ctx, grad_output: torch.Tensor)
SwissArmyTransformer/sat/quantization/kernels.py:61
↓ 9 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/llama/transform_param.py:54
↓ 9 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/clip/transform_param.py:25
↓ 9 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/clip/transform_param_new.py:22
↓ 9 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/bert/transform_param.py:68
↓ 9 callersMethodget_args
Get the parsed args of the current model. Args: **kwargs: will override the default args. Returns:
SwissArmyTransformer/sat/model/base_model.py:286
↓ 9 callersFunctionget_context_parallel_group
()
sat/vae_modules/utils.py:43
↓ 9 callersMethodis_rocm_pytorch
()
SwissArmyTransformer/sat/ops/ops_builder/builder.py:155
↓ 9 callersFunctionleaky_relu
(p=0.1)
sat/sgm/modules/autoencoding/losses/video_loss.py:33
↓ 9 callersFunctionmodulate
(x, shift, scale)
sat/dit_video_concat.py:438
↓ 9 callersMethodreinit
(self, parent_model=None)
SwissArmyTransformer/sat/model/base_model.py:55
↓ 9 callersFunctionrotate_half
(x)
SwissArmyTransformer/sat/model/position_embedding/rotary_embeddings.py:74
↓ 9 callersFunctionto_sigma
(neg_log_sigma)
sat/sgm/modules/diffusionmodules/sampling_utils.py:154
↓ 8 callersMethod__init__
(self, dim_in, dim_out)
sat/sgm/modules/attention.py:83
↓ 8 callersMethod__init__
( self, *, ch, out_ch, ch_mult=(1, 2, 4, 8), num_res_blocks,
sat/sgm/modules/diffusionmodules/model.py:264
↓ 8 callersMethod__init__
(self, dim_in, dim_out)
sat/vae_modules/attention.py:83
↓ 8 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/roberta/transform_param.py:68
↓ 8 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/chatglm3/transform_param.py:41
↓ 8 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/chatglm2/transform_param.py:38
↓ 8 callersFunctioncopy_layer_param
in-place copy from src to dst src and dst should be the same layer type, e.g., both are LayerNorm or both are Linear. Or at least, bo
SwissArmyTransformer/examples/chatglm2/transform_param_newmlp.py:39
↓ 8 callersFunctionfilling_sequence
seq: [2, 3, 5, ..., -1(to be generated), -1, ...] mems: [num_layers, batch_size, len_mems(index), mem_hidden_size] cache,
SwissArmyTransformer/sat/generation/autoregressive_sampling.py:52
↓ 8 callersFunctionload_checkpoint
Load a model checkpoint.
SwissArmyTransformer/sat/training/model_io.py:263
↓ 8 callersFunctiontimestep_embedding
Create sinusoidal timestep embeddings. :param timesteps: a 1-D Tensor of N indices, one per batch element. These may be
sat/sgm/modules/diffusionmodules/util.py:180
↓ 8 callersMethodwarning
(self, msg)
SwissArmyTransformer/sat/ops/ops_builder/builder.py:416
↓ 7 callersFunctionNormalize
(in_channels, num_groups=32)
sat/sgm/modules/diffusionmodules/model.py:49
↓ 7 callersMethod__init__
(self)
SwissArmyTransformer/sat/model/official/glm130B_model.py:85
↓ 7 callersMethod__init__
(self)
sat/sgm/modules/autoencoding/losses/video_loss.py:81
↓ 7 callersMethod__len__
(self)
SwissArmyTransformer/sat/data_utils/datasets.py:80
↓ 7 callersFunctioncopy_from_param
(src, dst)
SwissArmyTransformer/examples/cait/transform_param.py:37
↓ 7 callersFunctiondefault
(*args)
sat/sgm/modules/autoencoding/regularizers/lookup_free_quantization.py:33
next →1–100 of 2,541, ranked by callers