Functions — github.com/Alpha-VLLM/LLaMA2-Accessory

hub / github.com/Alpha-VLLM/LLaMA2-Accessory / functions

Functions1,094 in github.com/Alpha-VLLM/LLaMA2-Accessory

⨍ Functions 1,094◇ Types & classes 182

↓ 1 callersFunctionsetup_mixed_precision

(args)

Large-DiT-T2I/train.py:255

↓ 1 callersFunctionshow_mask

(img: Image, mask: torch.Tensor, color)

accessory/demos/multi_turn_mm_box.py:218

↓ 1 callersFunctionspace_timesteps

Create a list of timesteps to use from an original diffusion process, given the number of timesteps we want to take from equally-sized portio

Large-DiT-ImageNet/diffusion/respace.py:12

↓ 1 callersFunctionspace_timesteps

Create a list of timesteps to use from an original diffusion process, given the number of timesteps we want to take from equally-sized portio

Large-DiT-T2I/diffusion/respace.py:12

↓ 1 callersMethodsparse_transpose

( self, size, row_indices, column_indices)

accessory/model/LLM/mixtral_sparse_ens5.py:305

↓ 1 callersMethodsparse_transpose

( self, size, row_indices, column_indices)

accessory/model/LLM/mixtral_sparse.py:299

↓ 1 callersMethodsparse_transpose

( self, size, row_indices, column_indices)

accessory/model/LLM/mixtral_sparse_ens.py:303

↓ 1 callersFunctionsplit_list

Split a list into n (roughly) equal-sized chunks

light-eval/src/eval_mmvet.py:80

↓ 1 callersFunctionsplit_list

Split a list into n (roughly) equal-sized chunks

light-eval/src/eval_llavabenchmark.py:90

↓ 1 callersFunctiontensor_load_shard

r"""A helper function to partially load a tensor. This can save memory sometimes as this allows tensor parallel shards to stream into memory (

accessory/util/tensor_parallel.py:488

↓ 1 callersMethodtimestep_embedding

Create sinusoidal timestep embeddings. :param t: a 1-D Tensor of N indices, one per batch element. These ma

Large-DiT-ImageNet/models.py:57

↓ 1 callersMethodtimestep_embedding

Create sinusoidal timestep embeddings. :param t: a 1-D Tensor of N indices, one per batch element. These ma

Large-DiT-T2I/models/model.py:57

↓ 1 callersMethodtoken_drop

Drops labels to enable classifier-free guidance.

Large-DiT-ImageNet/models.py:101

↓ 1 callersMethodtoken_drop

Drops labels to enable classifier-free guidance.

Large-DiT-T2I/models/model.py:101

↓ 1 callersMethodtopology

(self, x: torch.Tensor, padded_bins: torch.Tensor)

accessory/model/LLM/mixtral_sparse_ens5.py:335

↓ 1 callersMethodtopology

(self, x: torch.Tensor, padded_bins: torch.Tensor)

accessory/model/LLM/mixtral_sparse.py:329

↓ 1 callersMethodtopology

(self, x: torch.Tensor, padded_bins: torch.Tensor)

accessory/model/LLM/mixtral_sparse_ens.py:333

↓ 1 callersFunctiontrain_one_epoch

(model: torch.nn.Module, data_loader, optimizer: torch.optim.Optimizer,

accessory/engine_finetune.py:13

↓ 1 callersFunctiontrain_one_epoch

(model: torch.nn.Module, data_loader, val_loader, optimizer: torch.optim.Optimizer,

accessory/engine_pretrain.py:14

↓ 1 callersMethodtraining_losses

( self, model, *args, **kwargs )

Large-DiT-ImageNet/diffusion/respace.py:94

↓ 1 callersMethodtraining_losses

( self, model, *args, **kwargs )

Large-DiT-T2I/diffusion/respace.py:94

↓ 1 callersMethodunpatchify

x: (N, T, patch_size**2 * C) imgs: (N, H, W, C)

Large-DiT-ImageNet/models.py:527

↓ 1 callersMethodunpatchify

x: (N, T, patch_size**2 * C) imgs: (N, H, W, C)

Large-DiT-T2I/models/model.py:575

↓ 1 callersMethodupdateProgress

(self, progress)

accessory/eval_mm/utils/vqa_eval.py:320

↓ 1 callersFunctionupdate_ema

Step the EMA model towards the current model.

Large-DiT-ImageNet/train.py:77

↓ 1 callersFunctionupdate_ema

Step the EMA model towards the current model.

Large-DiT-T2I/train.py:121

↓ 1 callersMethodupdate_with_all_losses

Update the reweighting using losses from a model. Sub-classes should override this method to update the reweighting using los

Large-DiT-ImageNet/diffusion/timestep_sampler.py:106

↓ 1 callersMethodupdate_with_all_losses

Update the reweighting using losses from a model. Sub-classes should override this method to update the reweighting using los

Large-DiT-T2I/diffusion/timestep_sampler.py:106

↓ 1 callersFunctionval_one_epoch

(model: torch.nn.Module, data_loader: Iterable, epoch: int, args=None)

accessory/engine_pretrain.py:119

↓ 1 callersFunctionvalidate_data

Method validate_data: validates that all files in the results folder are correct (have the correct name contents). Va

accessory/eval_mm/infographicsvqa_eval.py:36

↓ 1 callersMethodweights

Get a numpy array of weights, one per diffusion step. The weights needn't be normalized, but must be positive.

Large-DiT-ImageNet/diffusion/timestep_sampler.py:38

↓ 1 callersMethodweights

Get a numpy array of weights, one per diffusion step. The weights needn't be normalized, but must be positive.

Large-DiT-T2I/diffusion/timestep_sampler.py:38

↓ 1 callersFunctionwithin_eps

(pred: float, gt: float)

accessory/eval_mm/utils/math_utils.py:460

↓ 1 callersFunctionworker_func

()

accessory/demos/single_turn_mm.py:155

↓ 1 callersFunctionworker_func

()

accessory/demos/single_turn.py:143

↓ 1 callersFunctionwrite_configs

( params: Dict[str, Any], dtype: torch.dtype, dest_dir: str, vocab_size: int )

accessory/tools/convert_weights_to_hf.py:284

↓ 1 callersFunctionwrite_hf_ckpt

( hf_state_dict: List[Dict[str, torch.Tensor]], dest_dir: str, tokenizer_path: str, params: Dict[str,

accessory/tools/convert_weights_to_hf.py:345

↓ 1 callersFunctionwrite_model_weights

( hf_state_dict: List[Dict[str, torch.Tensor]], dest_dir: str )

accessory/tools/convert_weights_to_hf.py:240

↓ 1 callersFunctionwrite_tokenizer

(tokenizer_path: str, dest_dir: str)

accessory/tools/convert_weights_to_hf.py:266

FunctionDiT_Llama_3B_patch2

(**kwargs)

Large-DiT-ImageNet/models.py:673

FunctionDiT_Llama_3B_patch2

(**kwargs)

Large-DiT-T2I/models/model.py:750

FunctionDiT_Llama_600M_patch2

(**kwargs)

Large-DiT-ImageNet/models.py:666

FunctionDiT_Llama_600M_patch2

(**kwargs)

Large-DiT-T2I/models/model.py:744

FunctionDiT_Llama_7B_patch2

(**kwargs)

Large-DiT-ImageNet/models.py:680

FunctionDiT_Llama_7B_patch2

(**kwargs)

Large-DiT-T2I/models/model.py:756

Method__call__

(self, x, ts, **kwargs)

Large-DiT-ImageNet/diffusion/respace.py:124

Method__call__

(self, x, ts, **kwargs)

Large-DiT-T2I/diffusion/respace.py:124

Method__call__

(self, img: Image.Image)

accessory/eval_mm/inference_image_sphinx.py:53

Method__call__

(self, img: Image.Image)

accessory/data/transform.py:19

Method__call__

(self, loss, optimizer, model: FSDP, clip_grad=None, parameters=None, create_graph=False, update_grad=True)

accessory/util/misc.py:301

Method__enter__

(self)

accessory/util/tensor_type.py:20

Method__eq__

(self, other)

light-eval/src/eval_utils/math_util.py:246

Method__exit__

( self, exc_type: Optional[type[BaseException]], exc_val: Optional[BaseException],

accessory/util/tensor_type.py:35

Method__getattr__

(self, attr)

accessory/util/misc.py:225

Method__getitem__

(self, idx: int)

SPHINX/batch_inference.py:41

Method__getitem__

(self, index)

Large-DiT-T2I/data/dataset.py:169

Method__getitem__

(self, idx)

accessory/eval_mm/inference_image_sphinx.py:92

Method__getitem__

(self, idx)

accessory/data/falcon.py:187

Method__getitem__

(self, idx)

accessory/data/falcon_packed.py:183

Method__getitem__

(self, index)

accessory/data/alpaca.py:164

Method__getitem__

(self, index)

accessory/data/conversation/dataset.py:275

Method__init__

(self, hidden_size, frequency_embedding_size=256)

Large-DiT-ImageNet/models.py:40

Method__init__

(self, num_classes, hidden_size, dropout_prob)

Large-DiT-ImageNet/models.py:91

Method__init__

Initialize the FeedForward module. Args: dim (int): Input dimension. hidden_dim (int): Hidden dimension of t

Large-DiT-ImageNet/models.py:299

Method__init__

Initialize a TransformerBlock. Args: layer_id (int): Identifier for the layer. dim (int): Embedding dimensio

Large-DiT-ImageNet/models.py:356

Method__init__

(self, hidden_size, patch_size, out_channels)

Large-DiT-ImageNet/models.py:453

Method__init__

( self, input_size: int = 32, patch_size: int = 2, in_channels: int = 4,

Large-DiT-ImageNet/models.py:481

Method__init__

(self, diffusion)

Large-DiT-ImageNet/diffusion/timestep_sampler.py:63

Method__init__

(self, diffusion, history_per_term=10, uniform_prob=0.001)

Large-DiT-ImageNet/diffusion/timestep_sampler.py:121

Method__init__

( self, *, betas, model_mean_type, model_var_type, loss_type

Large-DiT-ImageNet/diffusion/gaussian_diffusion.py:153

Method__init__

(self, use_timesteps, **kwargs)

Large-DiT-ImageNet/diffusion/respace.py:73

Method__init__

(self, image_size: int, list_path: str)

SPHINX/batch_inference.py:26

Method__init__

(self, transform, tokenizer_path, text_dropout_prob, max_words=128)

Large-DiT-T2I/train.py:55

Method__init__

(self, diffusion)

Large-DiT-T2I/diffusion/timestep_sampler.py:63

Method__init__

(self, diffusion, history_per_term=10, uniform_prob=0.001)

Large-DiT-T2I/diffusion/timestep_sampler.py:121

Method__init__

( self, *, betas, model_mean_type, model_var_type, loss_type

Large-DiT-T2I/diffusion/gaussian_diffusion.py:153

Method__init__

(self, use_timesteps, **kwargs)

Large-DiT-T2I/diffusion/respace.py:73

Method__init__

(self, message=None)

Large-DiT-T2I/data/dataset.py:22

Method__init__

(self, config_path, item_processor: ItemProcessor, cache_on

Large-DiT-T2I/data/dataset.py:36

Method__init__

(self, hidden_size, frequency_embedding_size=256)

Large-DiT-T2I/models/model.py:40

Method__init__

(self, num_classes, hidden_size, dropout_prob)

Large-DiT-T2I/models/model.py:91

Method__init__

Initialize the FeedForward module. Args: dim (int): Input dimension. hidden_dim (int): Hidden dimension of t

Large-DiT-T2I/models/model.py:333

Method__init__

Initialize a TransformerBlock. Args: layer_id (int): Identifier for the layer. dim (int): Embedding dimensio

Large-DiT-T2I/models/model.py:390

Method__init__

(self, hidden_size, patch_size, out_channels)

Large-DiT-T2I/models/model.py:490

Method__init__

( self, patch_size: int = 2, max_seq_len: int = 288, in_channels: int = 4,

Large-DiT-T2I/models/model.py:518

Method__init__

pad an image to squre (borrowed from LLAVA, thx) :param background_color: rgb values for padded pixels, normalized to [0, 1]

accessory/eval_mm/inference_image_sphinx.py:46

Method__init__

(self, anno_path, prompt, img_root, img_size=224)

accessory/eval_mm/inference_image_sphinx.py:80

Method__init__

(self, config, global_config, prompt)

accessory/eval_mm/evaluate.py:85

Method__init__

(self)

accessory/eval_mm/utils/metric.py:54

Method__init__

(self)

accessory/eval_mm/utils/metric.py:104

Method__init__

Constructor of VQA helper class for reading and visualizing questions and answers. :param annotation_file (str): location of VQA anno

accessory/eval_mm/utils/vqa.py:33

Method__init__

(self, vqa=None, vqaRes=None, n=2)

accessory/eval_mm/utils/vqa_eval.py:20

Method__init__

(self, code, timeout, use_process: bool)

accessory/eval_mm/utils/math_utils.py:24

Method__init__

( self, llama_type: str, llama_config: str|List[str], tokenizer_path: str, with_visual: bool =

accessory/model/meta.py:20

Method__init__

Create a tokenizer, with inner implementation either spm or HF transformers tokenzier :param model_path: - when using spm

accessory/model/tokenizer.py:16

Method__init__

(self, *from_pretrained_args, gpus:Optional[int]=None, gpu_ids:Optional[List

accessory/model/multi_gpu_wrapper.py:172

Method__init__

( self, in_features: int, out_features: int, bias: bool = True, gather

accessory/model/peft.py:79

Method__init__

( self, in_features: int, out_features: int, bias: bool = True, input_

accessory/model/peft.py:189

Method__init__

Initialize the RMSNorm normalization layer. Args: dim (int): The dimension of the input tensor.

accessory/model/components.py:11

Method__init__

(self, args: ModelArgs)

accessory/model/LLM/mixtral_peft.py:63

← previous next →601–700 of 1,094, ranked by callers