Functions — github.com/NVlabs/DoRA

hub / github.com/NVlabs/DoRA / functions

Functions2,507 in github.com/NVlabs/DoRA

⨍ Functions 2,507◇ Types & classes 532↳ Endpoints 73

↓ 645 callersMethodto

(self, *args, **kwargs)

image_video_text_understanding/CLIP-ViL/src/tools/vision_helpers.py:92

↓ 168 callersMethodget

(cls, config_name: str)

image_video_text_understanding/VL-T5/src/adapters/adapter_configuration.py:64

↓ 156 callersMethodload

(cls, path)

image_video_text_understanding/VL-T5/src/param.py:324

↓ 114 callersMethodupdate

(self, new_val)

image_video_text_understanding/VL-T5/src/utils.py:48

↓ 106 callersMethodfrom_pretrained

(cls, pretrained_model_name_or_path: str, **kwargs)

image_video_text_understanding/VL-T5/inference/utils.py:178

↓ 80 callersMethodsave

(self, path)

image_video_text_understanding/VL-T5/src/param.py:319

↓ 77 callersMethodfrom_pretrained

r""" Instantiate a [`LoraModel`] from a pretrained Lora configuration and weights. Args: model ([`~transformers.PreTraine

visual_instruction_tuning/peft/src/peft/peft_model.py:207

↓ 73 callersMethodencode

(self, text)

image_video_text_understanding/VL-T5/src/clip/simple_tokenizer.py:121

↓ 69 callersMethodtrain_step

(self, batch)

image_video_text_understanding/VL-T5/src/gqa_model.py:12

↓ 65 callersMethodcopy

(self)

visual_instruction_tuning/llava/conversation.py:190

↓ 64 callersMethodbackward

(ctx, grad)

image_video_text_understanding/VL-T5/inference/modeling_frcnn.py:411

↓ 64 callersMethodeval

(self)

image_video_text_understanding/VL-T5/src/lora/layers.py:138

↓ 58 callersMethodstride

(self)

image_video_text_understanding/VL-T5/inference/modeling_frcnn.py:804

↓ 55 callersMethodstep

(self)

image_video_text_understanding/VL-T5/src/utils.py:152

↓ 54 callersMethodupdate

(self, adapter_name)

visual_instruction_tuning/peft/src/peft/utils/other.py:146

↓ 53 callersFunctionget_peft_model

Returns a Peft model object from a model and a config. Args: model ([`transformers.PreTrainedModel`]): Model to be wrapped.

visual_instruction_tuning/peft/src/peft/mapping.py:87

↓ 47 callersMethodfrom_pretrained

See `AutoTokenizer.from_pretrained` docstring.

visual_instruction_tuning/llava/model/language_model/mpt/adapt_tokenizer.py:37

↓ 44 callersMethodtest_step

(self, batch, **kwargs)

image_video_text_understanding/VL-T5/src/gqa_model.py:49

↓ 44 callersMethodvis_forward

(self, batch, device)

image_video_text_understanding/VL-T5/src/modeling_t5.py:715

↓ 43 callersMethodfrom_pretrained

r""" Args: Instantiate a `LoraModel` from a pretrained Lora configuration and weights. model (`transformers.PreTrainedMode

commonsense_reasoning/peft/src/peft/peft_model.py:133

↓ 42 callersMethodresize_token_embeddings

(self, new_num_tokens: int)

image_video_text_understanding/VL-T5/src/my_transformers/modeling_bart.py:1433

↓ 38 callersFunctiontokenizer_image_token

(prompt, tokenizer, image_token_index=IMAGE_TOKEN_INDEX, return_tensors=None)

visual_instruction_tuning/llava/mm_utils.py:43

↓ 35 callersMethodinit_weights

(self)

image_video_text_understanding/VL-T5/src/trainer_base.py:688

↓ 35 callersMethodsave_pretrained

r""" This function saves the adapter model and the adapter configuration files to a directory, so that it can be reloaded using the [`

visual_instruction_tuning/peft/src/peft/peft_model.py:124

↓ 34 callersMethodgenerate

(self, **kwargs)

commonsense_reasoning/peft/src/peft/peft_model.py:592

↓ 34 callersMethodget_lr

(self)

image_video_text_understanding/CLIP-ViL/src/lxrt/optimization.py:86

↓ 32 callersMethodconvert_tokens_to_ids

Converts a sequence of tokens into ids using the vocab.

image_video_text_understanding/CLIP-ViL/src/lxrt/tokenization.py:115

↓ 29 callersMethodappend_message

(self, role, message)

visual_instruction_tuning/llava/conversation.py:106

↓ 29 callersFunctiontranspose

(weight, fan_in_fan_out)

visual_instruction_tuning/peft/src/peft/utils/other.py:265

↓ 29 callersMethodwrite

(self, buf)

visual_instruction_tuning/llava/utils.py:73

↓ 27 callersFunctionget_vis_encoder

(backbone='openai/clip-vit-base-patch32', **kwargs)

image_video_text_understanding/VL-T5/src/vis_encoder.py:88

↓ 25 callersMethodgenerate

(self, **kwargs)

visual_instruction_tuning/peft/src/peft/peft_model.py:974

↓ 24 callersFunctionparse_args

(parse=True, **optional_kwargs)

image_video_text_understanding/VL-T5/src/param.py:67

↓ 22 callersMethod__init__

(self)

image_video_text_understanding/CLIP-ViL/src/lxrt/modeling.py:133

↓ 20 callersMethodto_dict

(self)

visual_instruction_tuning/peft/src/peft/utils/config.py:71

↓ 19 callersMethodcreate_model

(self, model_class, config=None, **kwargs)

image_video_text_understanding/VL-T5/src/trainer_base.py:313

↓ 18 callersMethoddim

(self)

image_video_text_understanding/CLIP-ViL/src/lxrt/entry.py:126

↓ 18 callersMethodget_prompt

(self)

visual_instruction_tuning/llava/conversation.py:29

↓ 18 callersMethodload

(self, path)

image_video_text_understanding/CLIP-ViL/src/tasks/gqa.py:409

↓ 18 callersMethodload_checkpoint

(self, ckpt_path)

image_video_text_understanding/VL-T5/src/trainer_base.py:680

↓ 18 callersMethodset_epoch

(self, epoch)

image_video_text_understanding/VL-T5/src/multitask_data.py:33

↓ 18 callersFunctionset_global_logging_level

Override logging levels of different modules based on their name as a prefix. It needs to be invoked after the modules have been loaded so th

image_video_text_understanding/VL-T5/src/utils.py:74

↓ 18 callersFunctiontranspose

(weight, fan_in_fan_out)

commonsense_reasoning/peft/src/peft/utils/other.py:158

↓ 17 callersMethodcreate_config

(self)

image_video_text_understanding/VL-T5/src/trainer_base.py:183

↓ 17 callersMethodcreate_optimizer_and_scheduler

(self)

image_video_text_understanding/VL-T5/src/trainer_base.py:573

↓ 17 callersMethodcreate_tokenizer

(self, **kwargs)

image_video_text_understanding/VL-T5/src/trainer_base.py:547

↓ 17 callersMethodget_prompt

Returns the virtual prompts to use for Peft. Only applicable when `peft_config.peft_type != PeftType.LORA`.

visual_instruction_tuning/peft/src/peft/peft_model.py:345

↓ 15 callersFunction_get_submodules

(model, key)

visual_instruction_tuning/peft/src/peft/utils/other.py:155

↓ 15 callersMethodencode

(self, text)

image_video_text_understanding/CLIP-ViL/clip/simple_tokenizer.py:121

↓ 15 callersMethodget_input_embeddings

(self)

visual_instruction_tuning/llava/model/language_model/mpt/modeling_mpt.py:81

↓ 15 callersMethodtrain

(self)

image_video_text_understanding/VL-T5/src/tvc.py:159

↓ 14 callersMethod__init__

(self, cfg, input_shape: Dict[str, ShapeSpec])

image_video_text_understanding/VL-T5/inference/modeling_frcnn.py:1503

↓ 14 callersMethod_shape

(self, tensor: torch.Tensor, seq_len: int, bsz: int)

image_video_text_understanding/VL-T5/src/my_transformers/modeling_bart.py:168

↓ 14 callersFunctionis_bnb_available

()

visual_instruction_tuning/peft/src/peft/import_utils.py:18

↓ 13 callersMethoddevice

(self)

visual_instruction_tuning/llava/model/multimodal_encoder/clip_encoder.py:62

↓ 13 callersMethodfreeze_whole_model

(self)

image_video_text_understanding/VL-T5/src/trainer_base.py:340

↓ 13 callersMethodget_model

(self)

visual_instruction_tuning/llava/model/llava_arch.py:88

↓ 13 callersMethodpartial_eval

(self)

image_video_text_understanding/VL-T5/src/trainer_base.py:344

↓ 13 callersMethodprint_trainable_params_percentage

(self, model)

image_video_text_understanding/VL-T5/src/trainer_base.py:325

↓ 13 callersMethodunfreeze_parameters

(self)

image_video_text_understanding/VL-T5/src/trainer_base.py:380

↓ 12 callersMethodprepare_inputs_for_testing

(self)

visual_instruction_tuning/peft/tests/test_decoder_models.py:57

↓ 12 callersMethodto_gradio_chatbot

(self)

visual_instruction_tuning/llava/conversation.py:159

↓ 11 callersMethod__init__

(self, config: BartConfig)

image_video_text_understanding/VL-T5/src/modeling_bart.py:1320

↓ 11 callersMethod__init__

(self, config: BartConfig)

image_video_text_understanding/VL-T5/src/my_transformers/modeling_bart.py:1294

↓ 11 callersMethod__init__

(self, config, has_relative_attention_bias=False)

image_video_text_understanding/VL-T5/src/my_transformers/modeling_t5.py:645

↓ 11 callersMethod_create_test_llama_config

Create a test config for a small Llama model for testing.

visual_instruction_tuning/peft/tests/test_adaption_prompt.py:61

↓ 11 callersMethodget_prompt

Returns the virtual prompts to use for Peft. Only applicable when `peft_config.peft_type != PeftType.LORA`.

commonsense_reasoning/peft/src/peft/peft_model.py:246

↓ 10 callersMethodembed_tokens

(self, x)

visual_instruction_tuning/llava/model/language_model/llava_mpt.py:41

↓ 10 callersMethodevaluate

(self, loader, dump_path=None)

image_video_text_understanding/VL-T5/src/tvc.py:393

↓ 10 callersFunctionload_pretrained_model

(model_path, model_base, model_name, load_8bit=False, load_4bit=False, device_map="auto", device="cuda", **kwa

visual_instruction_tuning/llava/model/builder.py:30

↓ 10 callersFunctionreduce_dict

Reduce the values in the dictionary from all processes so that process with rank 0 has the reduced results. Args: input_dict (dic

image_video_text_understanding/VL-T5/src/dist_utils.py:267

↓ 9 callersMethod__init__

( self, adapter_name: str, in_features: int, out_features: int, r: int

visual_instruction_tuning/peft/src/peft/tuners/lora.py:766

↓ 9 callersFunctiondisable_torch_init

Disable the redundant torch default initialization to accelerate model creation.

visual_instruction_tuning/llava/utils.py:93

↓ 9 callersMethodget

(cls, config_name: str)

image_video_text_understanding/CLIP-ViL/src/lxrt/adapters/adapter_configuration.py:64

↓ 9 callersFunctionget_peft_model_state_dict

Get the state dict of the Peft model. Args: model ([`PeftModel`]): The Peft model. When using torch.nn.DistributedDataParallel, Deep

visual_instruction_tuning/peft/src/peft/utils/save_and_load.py:26

↓ 9 callersFunctionresize_pos_embed

(posemb, posemb_new)

image_video_text_understanding/CLIP-ViL/src/lxrt/visual_transformers.py:17

↓ 9 callersFunctionto_image_list

tensors can be an ImageList, a torch.Tensor or an iterable of Tensors. It can't be a numpy array. When tensors is an iterable of Tensors,

image_video_text_understanding/CLIP-ViL/src/tasks/vision_helpers.py:97

↓ 9 callersMethodtokenize

(self, text)

image_video_text_understanding/CLIP-ViL/src/lxrt/tokenization.py:105

↓ 8 callersFunction_expand_mask

Expands attention_mask from `[bsz, seq_len]` to `[bsz, 1, tgt_seq_len, src_seq_len]`.

image_video_text_understanding/VL-T5/src/my_transformers/modeling_bart.py:106

↓ 8 callersMethodadd_adapter

(self, adapter_name, config=None)

visual_instruction_tuning/peft/src/peft/tuners/ia3.py:135

↓ 8 callersMethodbackward

(ctx, do)

visual_instruction_tuning/llava/model/language_model/mpt/flash_attn_triton.py:475

↓ 8 callersMethoddecode

(self, tokens)

image_video_text_understanding/CLIP-ViL/clip/simple_tokenizer.py:129

↓ 8 callersMethoddisable_adapter

Disables the adapter module.

visual_instruction_tuning/peft/src/peft/peft_model.py:430

↓ 8 callersMethoddump_result

Dump the result to a GQA-challenge submittable json file. GQA json file submission requirement: results = [result]

image_video_text_understanding/VL-T5/src/gqa_data.py:434

↓ 8 callersMethodeval

(self)

commonsense_reasoning/peft/src/peft/tuners/dora.py:403

↓ 8 callersMethodevaluate_raw

https://github.com/GT-Vision-Lab/VQA/blob/master/PythonEvaluationTools/vqaEvaluation/vqaEval.py

image_video_text_understanding/VL-T5/src/vqa_data.py:569

↓ 8 callersFunctionget_model_name_from_path

(model_path)

visual_instruction_tuning/llava/mm_utils.py:65

↓ 8 callersFunctionlinear_layer

Generates a linear module and initializes it.

image_video_text_understanding/CLIP-ViL/src/lxrt/adapters/adapter_utils.py:22

↓ 8 callersFunctionlinear_layer

Generates a linear module and initializes it.

image_video_text_understanding/VL-T5/src/adapters/adapter_utils.py:22

↓ 8 callersFunctionpad_list_tensors

location will always be cpu for np tensors

image_video_text_understanding/VL-T5/inference/modeling_frcnn.py:46

↓ 8 callersMethodreset_parameters

(self)

commonsense_reasoning/peft/src/peft/tuners/dora.py:374

↓ 8 callersMethodstep

Performs a single optimization step. Arguments: closure (callable, optional): A closure that reevaluates the model

image_video_text_understanding/CLIP-ViL/src/lxrt/optimization.py:101

↓ 8 callersMethodtest_step

(self, batch, **kwargs)

image_video_text_understanding/VL-T5/src/multitask_model.py:55

↓ 8 callersMethodto_dict

(self)

commonsense_reasoning/peft/src/peft/utils/config.py:61

↓ 8 callersMethodtrain_step

(self, batch, **kwargs)

image_video_text_understanding/VL-T5/src/multitask_model.py:17

↓ 8 callersMethodvalid_step

(self, batch, **kwargs)

image_video_text_understanding/VL-T5/src/multitask_model.py:36

↓ 8 callersMethodvalid_step

(self, batch)

image_video_text_understanding/VL-T5/src/vcr_model.py:101

↓ 7 callersMethod__init__

(self, embed_dim: int, # vision image_resolution: int,

image_video_text_understanding/CLIP-ViL/clip/model.py:325

↓ 7 callersMethod__init__

(self, embed_dim: int, # vision image_resolution: int,

image_video_text_understanding/VL-T5/src/clip/model.py:383

↓ 7 callersMethod__init__

( self, in_features: int, out_features: int, r: int = 0, lora_alpha: i

commonsense_reasoning/peft/src/peft/tuners/lora.py:290

next →1–100 of 2,507, ranked by callers