MCPcopy Create free account

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

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

⨍ Functions 1,094◇ Types & classes 182
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama.py:217
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_qformerv2_peft.py:181
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/mixtral_sparse_ens5.py:206
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_ens5p2.py:178
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/mixtral_sparse.py:200
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_ens.py:178
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_qformerv2.py:172
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/mixtral.py:182
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_peft.py:182
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_ens_light.py:179
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_adapter.py:222
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_ens_peft.py:188
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/mixtral_sparse_ens.py:204
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/internlm.py:163
↓ 1 callersMethoddestroy_kv_cache
(self)
accessory/model/LLM/llama_ens5_light.py:179
↓ 1 callersFunctiondfs_find_params_and_clean_names
( module: nn.Module, prefix: str, sharded_views_original_shape: Dict[str, torch.Size], )
accessory/util/param_group.py:169
↓ 1 callersFunctiondiscretized_gaussian_log_likelihood
Compute the log-likelihood of a Gaussian distribution discretizing to a given image. :param x: the target images. It is assumed that this
Large-DiT-ImageNet/diffusion/diffusion_utils.py:62
↓ 1 callersFunctiondiscretized_gaussian_log_likelihood
Compute the log-likelihood of a Gaussian distribution discretizing to a given image. :param x: the target images. It is assumed that this
Large-DiT-T2I/diffusion/diffusion_utils.py:62
↓ 1 callersFunctiondisplay_results
(results, show_answer_types)
accessory/eval_mm/infographicsvqa_eval.py:200
↓ 1 callersFunctiondistributed_init
(args)
Large-DiT-T2I/parallel.py:40
↓ 1 callersFunctiondownload_file
(repo_id, subfolder, filename, local_dir)
accessory/tools/download.py:19
↓ 1 callersFunctiondownload_files
(repo_id, subfolder, file_names, output_path)
accessory/tools/download.py:27
↓ 1 callersFunctiondraw_box_mask_on_image
(img: Image, l_name_box_color, predictor)
accessory/demos/multi_turn_mm_box.py:226
↓ 1 callersMethodencode_wo_prefix_space
(self, s:str)
accessory/model/tokenizer.py:71
↓ 1 callersFunctionentry_point
Evaluates the functional correctness of generated samples, and writes results to f"{sample_file}_results.jsonl.gz"
light-eval/src/eval_humaneval.py:77
↓ 1 callersFunctionestimate_pass_at_k
Estimates pass@k of each problem and returns them in an array.
light-eval/src/eval_utils/humaneval_evaluation.py:14
↓ 1 callersFunctionestimator
Calculates 1 - comb(n - c, k) / comb(n, k).
light-eval/src/eval_utils/humaneval_evaluation.py:23
↓ 1 callersMethodevaluate
(self, outputs, ds, args)
accessory/eval_mm/evaluate.py:90
↓ 1 callersMethodevaluate
(self, quesIds=None)
accessory/eval_mm/utils/vqa_eval.py:194
↓ 1 callersFunctionevaluate_functional_correctness
Evaluates the functional correctness of generated samples, and writes results to f"{sample_file}_results.jsonl.gz"
light-eval/src/eval_utils/humaneval_evaluation.py:40
↓ 1 callersFunctionevaluate_method
Method evaluate_method: evaluate method and returns the results Results. Dictionary with the following values: - method (required
accessory/eval_mm/infographicsvqa_eval.py:87
↓ 1 callersFunctionextract_and_color
Extracts tuples of text and list from a given string and also generates a Markdown string with uniquely colored substrings wrapped with <p> <
accessory/demos/multi_turn_mm_box.py:139
↓ 1 callersFunctionextract_ans
(completion, answer)
light-eval/src/eval_math.py:90
↓ 1 callersFunctionextract_ans
(ans, mode)
light-eval/src/eval_bbh.py:95
↓ 1 callersFunctionextract_ans
(ans)
light-eval/src/eval_mmlu.py:109
↓ 1 callersFunctionextract_ans
(completion)
light-eval/src/eval_gsm8k.py:95
↓ 1 callersFunctionextract_ans_by_logits
(tokenizer, logits)
light-eval/src/eval_cmmlu.py:115
↓ 1 callersFunctionextract_ans_by_logits
(tokenizer, logits)
light-eval/src/eval_ceval.py:107
↓ 1 callersFunctionextract_math_answer
(pred_str)
accessory/eval_mm/utils/math_utils.py:269
↓ 1 callersFunctionfilter_code
(completion: str)
light-eval/src/eval_humaneval.py:94
↓ 1 callersFunctionfind_free_port
()
Large-DiT-ImageNet/sample.py:116
↓ 1 callersFunctionfind_free_port
()
Large-DiT-T2I/demo.py:71
↓ 1 callersFunctionfind_free_port
Find a free port within the specified range.
accessory/util/misc.py:30
↓ 1 callersFunctionfind_sublist
(a: list, b:list)
accessory/data/conversation/dataset.py:295
↓ 1 callersFunctionfix_a_slash_b
(string)
light-eval/src/eval_utils/math_util.py:125
↓ 1 callersFunctionfix_fracs
(string)
light-eval/src/eval_utils/math_util.py:94
↓ 1 callersFunctionfix_indents
(text: str)
light-eval/src/eval_humaneval.py:99
↓ 1 callersFunctionfix_sqrt
(string)
light-eval/src/eval_utils/math_util.py:148
↓ 1 callersFunctionformat_prompt
(prompt)
light-eval/src/eval_mmvet.py:76
↓ 1 callersFunctionformat_prompt
(prompt)
light-eval/src/eval_llavabenchmark.py:84
↓ 1 callersFunctionformat_subject
(subject)
light-eval/src/eval_mmlu.py:56
↓ 1 callersMethodforward
Forward pass of DiT. x: (N, C, H, W) tensor of spatial inputs (images or latent representations of images) t: (N,)
Large-DiT-ImageNet/models.py:550
↓ 1 callersMethodforward
Forward pass of DiT. t: (N,) tensor of diffusion timesteps y: (N,) tensor of class labels
Large-DiT-T2I/models/model.py:640
↓ 1 callersFunctiongenerate_batch_completion
( model: MetaModel, prompt, batch_size )
light-eval/src/eval_humaneval.py:111
↓ 1 callersFunctiongenerate_crop_size_list
(num_patches, patch_size, max_ratio=4.0)
Large-DiT-T2I/imgproc.py:54
↓ 1 callersFunctiongenerate_few_shot_prompt
(task, dev_df, ntrain=-1)
light-eval/src/eval_cmmlu.py:103
↓ 1 callersFunctiongenerate_few_shot_prompt
(train_df, subject, k=-1)
light-eval/src/eval_mmlu.py:73
↓ 1 callersFunctiongenerate_few_shot_prompt
(subject, dev_df, ntrain=-1)
light-eval/src/eval_ceval.py:96
↓ 1 callersFunctiongenerate_output
(model, img_path, prompt)
light-eval/src/eval_mmvet.py:116
↓ 1 callersFunctiongenerate_output
(model, img_path,prompt)
light-eval/src/eval_llavabenchmark.py:129
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_math.py:20
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_bbh.py:30
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_cmmlu.py:30
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_humaneval.py:21
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_mmlu.py:23
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_ceval.py:25
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_mmvet.py:33
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_llavabenchmark.py:33
↓ 1 callersFunctionget_args_parser
()
light-eval/src/eval_gsm8k.py:20
↓ 1 callersFunctionget_args_parser
()
accessory/main_finetune.py:55
↓ 1 callersFunctionget_args_parser
()
accessory/main_pretrain.py:51
↓ 1 callersFunctionget_args_parser
()
accessory/tools/download.py:69
↓ 1 callersFunctionget_args_parser
()
accessory/tools/weight_operate.py:5
↓ 1 callersFunctionget_args_parser
()
accessory/eval_mm/inference_image_sphinx.py:304
↓ 1 callersFunctionget_args_parser
()
accessory/demos/single_turn_mm.py:24
↓ 1 callersFunctionget_args_parser
()
accessory/demos/single_turn.py:22
↓ 1 callersFunctionget_beta_schedule
This is the deprecated API for creating beta schedules. See get_named_beta_schedule() for the new library of schedules.
Large-DiT-ImageNet/diffusion/gaussian_diffusion.py:65
↓ 1 callersFunctionget_beta_schedule
This is the deprecated API for creating beta schedules. See get_named_beta_schedule() for the new library of schedules.
Large-DiT-T2I/diffusion/gaussian_diffusion.py:65
↓ 1 callersFunctionget_chunk
(lst, n, k)
light-eval/src/eval_llavabenchmark.py:96
↓ 1 callersFunctionget_eval
(content: str, max_tokens: int)
light-eval/src/eval_llavabenchmark.py:178
↓ 1 callersFunctionget_file_names
(prefix, model_size)
accessory/tools/download.py:31
↓ 1 callersMethodget_fsdp_wrap_module_list
(self)
Large-DiT-ImageNet/models.py:645
↓ 1 callersMethodget_fsdp_wrap_module_list
(self)
Large-DiT-T2I/models/model.py:735
↓ 1 callersMethodget_image_words
(self)
accessory/model/meta.py:567
↓ 1 callersFunctionget_intra_node_process_group
()
Large-DiT-T2I/parallel.py:84
↓ 1 callersMethodget_item_func
(self, index)
Large-DiT-T2I/data/dataset.py:160
↓ 1 callersMethodget_item_func
(self, index)
accessory/data/conversation/dataset.py:210
↓ 1 callersFunctionget_local_indices
(rank: int, world_size: int, dataset_len: int)
SPHINX/batch_inference.py:48
↓ 1 callersFunctionget_local_indices
(rank: int, world_size: int, dataset_len: int)
accessory/eval_mm/inference_image_sphinx.py:130
↓ 1 callersFunctionget_model_parallel_dim_dict
(model: nn.Module)
Large-DiT-ImageNet/grad_norm.py:11
↓ 1 callersFunctionget_model_parallel_dim_dict
(model: nn.Module)
Large-DiT-T2I/grad_norm.py:11
↓ 1 callersMethodget_quant_blocklist
(self)
accessory/model/meta.py:570
↓ 1 callersFunctionget_rank
(group=None)
accessory/util/misc.py:78
↓ 1 callersFunctionget_train_sampler
(dataset, rank, world_size, global_batch_size, max_steps, resume_step, seed)
Large-DiT-ImageNet/train.py:53
↓ 1 callersFunctionget_train_sampler
(dataset, rank, world_size, global_batch_size, max_steps, resume_step, seed)
Large-DiT-T2I/train.py:97
↓ 1 callersFunctionget_world_size
(group=None)
accessory/util/misc.py:72
↓ 1 callersFunctiongradio_worker
The gradio worker is responsible for displaying the WebUI and relay the requests to model workers. It should be launched only once. Args
accessory/demos/multi_turn.py:128
↓ 1 callersFunctiongradio_worker
The gradio worker is responsible for displaying the WebUI and relay the requests to model workers. It should be launched only once. Args
accessory/demos/multi_turn_mm.py:137
↓ 1 callersFunctiongradio_worker
The gradio worker is responsible for displaying the WebUI and relay the requests to model workers. It should be launched only once. Args
accessory/demos/multi_turn_mm_box.py:291
↓ 1 callersMethodgroups
(self)
Large-DiT-T2I/data/dataset.py:191
← previousnext →401–500 of 1,094, ranked by callers