Functions — github.com/LLaVA-Annonymous/LLaVA

hub / github.com/LLaVA-Annonymous/LLaVA / functions

Functions275 in github.com/LLaVA-Annonymous/LLaVA

⨍ Functions 275◇ Types & classes 38 ↳ Endpoints 9

↓ 1 callersFunctiongenerate_stream

Adapted from fastchat/serve/model_worker.py::generate_stream

llava/serve/cli.py:15

↓ 1 callersMethodgenerate_stream

(self, params)

llava/serve/model_worker.py:183

↓ 1 callersMethodgenerate_stream_gate

(self, params)

llava/serve/model_worker.py:304

↓ 1 callersFunctionget_args

()

llava/eval/eval_science_qa_gpt4.py:9

↓ 1 callersFunctionget_args

()

llava/eval/eval_science_qa.py:8

↓ 1 callersFunctionget_args

()

llava/eval/eval_science_qa_gpt4_requery.py:9

↓ 1 callersFunctionget_chunk

(lst, n, k)

llava/eval/model_vqa_science.py:25

↓ 1 callersFunctionget_chunk

(lst, n, k)

llava/eval/model_vqa.py:25

↓ 1 callersFunctionhtml_to_markdown

(val: str)

llava/data/clean_sharegpt.py:41

↓ 1 callersFunctioninit_on_device

Device initialization context manager. A context manager under which models are initialized with all parameters on the specified device.

llava/model/mpt/meta_init_context.py:37

↓ 1 callersMethodinitialize_vision_modules

(self, vision_tower, mm_vision_select_layer, pretrain_mm_mlp_adapter=None, t

llava/model/llava.py:54

↓ 1 callersMethodinitialize_vision_tokenizer

(self, mm_use_im_start_end, tokenizer, device, tune_mm_mlp_adapter=False,

llava/model/llava.py:285

↓ 1 callersMethodlist_models

(self)

llava/serve/controller.py:112

↓ 1 callersFunctionload_image

(image_file)

llava/eval/run_llava.py:25

↓ 1 callersFunctionload_model

(model_path, model_name, num_gpus)

llava/serve/model_worker.py:49

↓ 1 callersFunctionmain

(args)

llava/serve/cli.py:75

↓ 1 callersFunctionmain

()

llava/serve/test_message.py:9

↓ 1 callersFunctionmain

(args)

llava/data/alpaca-converter.py:20

↓ 1 callersFunctionmain

(args)

llava/data/clean_sharegpt.py:181

↓ 1 callersFunctionmain

(args)

llava/data/split_long_conversation.py:75

↓ 1 callersFunctionmake_delta

(base_model_path, target_model_path, delta_path, hub_repo_id)

llava/model/make_delta.py:13

↓ 1 callersFunctionmake_supervised_data_module

Make dataset and collator for supervised fine-tuning.

llava/train/train.py:517

↓ 1 callersFunctionparse_args

()

scripts/extract_mm_projector.py:8

↓ 1 callersFunctionparse_score

(review)

llava/eval/eval_gpt_review.py:37

↓ 1 callersFunctionparse_score

(review)

llava/eval/eval_gpt_review_visual.py:37

↓ 1 callersFunctionpatch_FSDP_use_orig_params

(func)

llava/train/train.py:645

↓ 1 callersFunctionpatch_config

(config)

llava/eval/model_vqa_science.py:76

↓ 1 callersFunctionpatch_config

(config)

llava/eval/model_vqa.py:36

↓ 1 callersFunctionpatch_tensor_constructor

(fn)

llava/model/mpt/meta_init_context.py:76

↓ 1 callersFunctionpost_process_code

(code)

llava/serve/gradio_web_server.py:166

↓ 1 callersFunctionpreprocess_mpt

( sources, tokenizer: transformers.PreTrainedTokenizer, )

llava/train/train.py:282

↓ 1 callersFunctionpreprocess_multimodal

( sources: Sequence[str], multimodal_cfg: dict, cur_token_len: int, )

llava/train/train.py:187

↓ 1 callersFunctionpreprocess_v1

( sources, tokenizer: transformers.PreTrainedTokenizer, )

llava/train/train.py:212

↓ 1 callersFunctionpretty_print_semaphore

(semaphore)

llava/utils.py:123

↓ 1 callersMethodreceive_heart_beat

(self, worker_name: str, queue_length: int)

llava/serve/controller.py:173

↓ 1 callersFunctionreformat_code

(val: str)

llava/data/clean_sharegpt.py:31

↓ 1 callersMethodrefresh_all_workers

(self)

llava/serve/controller.py:104

↓ 1 callersMethodremove_stable_workers_by_expiration

(self)

llava/serve/controller.py:183

↓ 1 callersFunctionreplace_llama_attn_with_flash_attn

()

llava/train/llama_flash_attn_monkey_patch.py:100

↓ 1 callersFunctionsafe_save_model_for_hf_trainer

Collects the state dict and dump to disk.

llava/train/train.py:90

↓ 1 callersFunctionskip

(conv, args)

llava/data/optional_clean.py:16

↓ 1 callersFunctionsmall_param_init_fn_

(module: nn.Module, n_layers: int, d_model: int, init_div_is_residual: Union[int, float, str, bool]=True, emb_

llava/model/mpt/param_init_fns.py:137

↓ 1 callersFunctionsmart_tokenizer_and_embedding_resize

Resize tokenizer and embedding. Note: This is the unoptimized version that may make your embedding size not be divisible by 64.

llava/train/train.py:103

↓ 1 callersFunctionsplit_contents

Keep the maximum round of conversations within the max token length constraint

llava/data/split_long_conversation.py:33

↓ 1 callersFunctionsplit_list

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

llava/eval/model_vqa_science.py:19

↓ 1 callersFunctionsplit_list

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

llava/eval/model_vqa.py:19

↓ 1 callersFunctionunwrap_model

Recursively unwraps a model from potential containers (as used in distributed training). Args: model (`torch.nn.Module`): The model

llava/train/llava_trainer.py:9

↓ 1 callersFunctionviolates_moderation

Check whether the text violates OpenAI moderation API.

llava/utils.py:102

↓ 1 callersMethodworker_api_generate_stream

(self, params)

llava/serve/controller.py:193

↓ 1 callersMethodworker_api_get_status

(self)

llava/serve/controller.py:220

Method__call__

(self, output_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs)

llava/model/utils.py:35

Method__call__

(self, output_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs)

llava/eval/model_vqa_science.py:99

Method__call__

(self, output_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs)

llava/eval/model_qa.py:21

Method__call__

(self, output_ids: torch.LongTensor, scores: torch.FloatTensor, **kwargs)

llava/eval/model_vqa.py:138

Method__call__

(self, instances: Sequence[Dict])

llava/train/train.py:491

llava/train/train.py:407

Method__getitem__

(self, i)

llava/train/train.py:429

Method__init__

(self, logger, log_level=logging.INFO)

llava/utils.py:64

Method__init__

(self, controller_addr, worker_addr, worker_id, no_register, model_path, mod

llava/serve/model_worker.py:104

Method__init__

Parameters: value: Default value to show in chatbot. If callable, the function will be called whenever the app loads to set the i

llava/serve/gradio_patch.py:26

Method__init__

(self, dispatch_method: str)

llava/serve/controller.py:58

Method__init__

(self, config)

llava/model/llava.py:191

Method__init__

(self, keywords, tokenizer, input_ids)

llava/model/utils.py:27

Method__init__

(self, config)

llava/model/llava_mpt.py:177

Method__init__

(self, d_model: int, n_heads: int, expansion_ratio: int, attn_config: Dict={'attn_type': 'multihead_attention'

llava/model/mpt/blocks.py:22

Method__init__

(self, config: MPTConfig)

llava/model/mpt/modeling_mpt.py:222

Method__init__

The MPT configuration class. Args: d_model (int): The size of the embedding dimension of the model. n_heads (int): Th

llava/model/mpt/configuration_mpt.py:10

Method__init__

(self, normalized_shape, eps=1e-05, elementwise_affine=True, device=None, dtype=None)

llava/model/mpt/norm.py:16

Method__init__

(self, normalized_shape, eps=1e-05, weight=True, dtype=None, device=None)

llava/model/mpt/norm.py:48

Method__init__

(self, d_model: int, n_heads: int, attn_impl: str='triton', clip_qkv: Optional[float]=None, qk_ln: bool=False,

llava/model/mpt/attention.py:181

Method__init__

(self, keywords, tokenizer, input_ids)

llava/eval/model_vqa_science.py:93

Method__init__

(self, keywords, tokenizer, input_ids)

llava/eval/model_qa.py:15

Method__init__

(self, keywords, tokenizer, input_ids)

llava/eval/model_vqa.py:132

Method__init__

(self, data_path: str, tokenizer: transformers.PreTrainedTokenizer, multimod

llava/train/train.py:414

Method__len__

(self)

llava/train/train.py:404

Method__len__

(self)

llava/train/train.py:426

Function_build_alibi_tensor

(self: BloomModel, batch_size: int, query_length: int, key_length: int, dtype: torch.dtype, device: torch.devi

llava/model/mpt/hf_prefixlm_converter.py:140

Function_prepare_attn_mask

(self: BloomModel, attention_mask: torch.Tensor, bidirectional_mask: Optional[torch.Tensor], input_shape: Tupl

llava/model/mpt/hf_prefixlm_converter.py:126

Function_prepare_decoder_attention_mask

(self, attention_mask, input_shape, inputs_embeds, past_key_values_length)

llava/model/mpt/hf_prefixlm_converter.py:285

Function_prepare_decoder_attention_mask

(self, attention_mask, input_shape, inputs_embeds, past_key_values_length)

llava/train/llama_flash_attn_monkey_patch.py:94

Method_reorder_cache

Used by HuggingFace generate when using beam search with kv-caching. See https://github.com/huggingface/transformers/blob/3ec7a47664ebe40c40f

llava/model/mpt/modeling_mpt.py:302

Method_set_gradient_checkpointing

(self, module, value=False)

llava/model/llava_mpt.py:196

Methodactivation_checkpointing_fn

(self, module)

llava/model/mpt/modeling_mpt.py:217

Methodactivation_checkpointing_fn

(self, module)

llava/model/mpt/modeling_mpt.py:278

Functionadd_bidirectional_mask_if_missing

Attempts to add bidirectional_mask to batch if missing. Raises: KeyError if bidirectional_mask is missing and can't be inferred

llava/model/mpt/hf_prefixlm_converter.py:401

Functionadd_text

(state, text, image, image_process_mode, request: gr.Request)

llava/serve/gradio_web_server.py:141

Functionbaseline_param_init_fn_

(module: nn.Module, init_std: float, n_layers: int, d_model: Optional[int]=None, init_div_is_residual: Union[i

llava/model/mpt/param_init_fns.py:131

Functionbuild_prompt

(problems, shot_qids, test_qid, args)

scripts/convert_sqa_to_llava_base_prompt.py:244

Functionbuild_prompt_gpt4

(problems, shot_qids, test_qid, args)

scripts/convert_sqa_to_llava_base_prompt.py:291

Functionclean_html_one_sample

(sample)

llava/data/clean_sharegpt.py:74

Functionclear_history

(request: gr.Request)

llava/serve/gradio_web_server.py:135

Functionconvert_caps

(results)

llava/eval/eval_science_qa_gpt4.py:19

Functionconvert_caps

(results)

llava/eval/eval_science_qa.py:19

Functionconvert_caps

(results)

llava/eval/eval_science_qa_gpt4_requery.py:21

Functionconvert_hf_causal_lm_to_prefix_lm

Converts a HuggingFace Causal LM to a Prefix LM. Supported HuggingFace model classes: - `GPT2LMHeadModel` - `GPTNeoForCausalLM`

llava/model/mpt/hf_prefixlm_converter.py:335

Functionconvert_to_jsonl

(base_dir, split, prompt_format="QCM-LEPA")

scripts/convert_sqa_to_llava.py:49

Functionconvert_to_llava

(base_dir, split, prompt_format="QCM-LEPA")

scripts/convert_sqa_to_llava.py:8

Functioncustom_forward

(*inputs)

llava/model/mpt/hf_prefixlm_converter.py:209

Functiondownvote_last_response

(state, model_selector, request: gr.Request)

llava/serve/gradio_web_server.py:113

← previous next →101–200 of 275, ranked by callers