Functions — github.com/Agent-RL/ReCall

hub / github.com/Agent-RL/ReCall / functions

Functions1,539 in github.com/Agent-RL/ReCall

⨍ Functions 1,539◇ Types & classes 237↳ Endpoints 28

(self, method_name, dispatch_fn, collect_fn, execute_fn, blocking)

src/verl/single_controller/ray/base.py:37

Gather the label from data. The value in label should be [0, vocab_size) Args: data: (..., vocab_size) label (torch.IntTensor) :

src/verl/utils/torch_functional.py:33

↓ 1 callersFunctiongenerate_function

(name)

src/verl/single_controller/ray/base.py:435

↓ 1 callersMethodgenerate_tree_of_thoughts_batch

(self, initial_prompts_batch: List[str])

src/flashrag/pipeline/active_pipeline.py:1243

↓ 1 callersMethodget_all_params

Get all the parameters of the models in all pp ranks Returns: params: List[List[Dict[str, Tensor]]]: a list of parameters in all

src/verl/workers/sharding_manager/megatron_vllm.py:185

↓ 1 callersMethodget_answer_token_indices

Obtain the index of token corresponsding to the option

src/flashrag/refiner/kg_refiner.py:359

↓ 1 callersMethodget_availale_master_addr_port

(self)

src/verl/single_controller/base/worker.py:63

↓ 1 callersFunctionget_background_loop

()

src/flashrag/generator/openai_generator.py:14

↓ 1 callersFunctionget_checkpoint_tracker_filename

Tracker file rescords the latest chckpoint during training to restart from.

src/verl/utils/checkpoint/checkpoint_manager.py:144

↓ 1 callersMethodget_compressed_input

( self, loss, input_ids, attention_mask, end=200, iterative_si

src/flashrag/refiner/llmlingua_compressor.py:1437

↓ 1 callersFunctionget_custom_reward_fn

(config)

src/verl/trainer/main_ppo.py:24

↓ 1 callersFunctionget_custom_reward_fn

(config)

src/verl/trainer/main_eval.py:29

↓ 1 callersFunctionget_default_model_parallel_config

()

src/verl/utils/megatron/tensor_parallel.py:43

↓ 1 callersFunctionget_device_flops

(unit="T")

src/verl/utils/flops_counter.py:21

↓ 1 callersMethodget_dynamic_compression_ratio

( self, context: list, target_token: float, iterative_size: int, dynam

src/flashrag/refiner/llmlingua_compressor.py:1094

↓ 1 callersMethodget_engine

(model_path, device='cpu', **kwargs)

src/flashrag/generator/multimodal_generator.py:370

↓ 1 callersMethodget_examplars_for_reasoning_chain

(self, all_query)

src/flashrag/refiner/kg_refiner.py:167

↓ 1 callersMethodget_examplars_for_triple

(self, doc_list, batch_size=64)

src/flashrag/refiner/kg_refiner.py:134

↓ 1 callersFunctionget_f1_score

(prediction: str, ground_truths: Union[str, List[str]])

src/verl/utils/reward_score/re_call.py:64

↓ 1 callersFunctionget_host_ip

()

scripts/train/get_host_ip.py:4

↓ 1 callersFunctionget_layer_index

(layer_name: str)

src/verl/workers/sharding_manager/patch/fsdp_vllm_patch.py:36

↓ 1 callersFunctionget_local_temp_path

Return a local temp path that joins cache_dir and basename of hdfs_path Args: hdfs_path: cache_dir: Returns:

src/verl/utils/fs.py:40

↓ 1 callersMethodget_megatron_global_info

(self)

src/verl/single_controller/base/megatron/worker.py:23

↓ 1 callersFunctionget_micro_data_parallel_rank

()

src/verl/workers/sharding_manager/megatron_vllm.py:485

↓ 1 callersFunctionget_model_size

(model: nn.Module, scale='auto')

src/verl/utils/model.py:119

↓ 1 callersMethodget_n_gpus

Get the number of gpus in this cluster.

src/verl/trainer/ppo/ray_trainer.py:105

↓ 1 callersFunctionget_nccl_id_store_by_name

(name)

src/verl/utils/rendezvous/ray_backend.py:34

↓ 1 callersFunctionget_parallel_model_from_config

(config, megatron_config, pre_process=No

src/verl/utils/model.py:248

↓ 1 callersMethodget_partitions

(self)

src/verl/utils/seqlen_balancing.py:63

↓ 1 callersFunctionget_predefined_dispatch_fn

(dispatch_mode)

src/verl/single_controller/base/decorator.py:305

↓ 1 callersFunctionget_predefined_execute_fn

Note that here we only asks execute_all and execute_rank_zero to be implemented Leave the choice of how these two functions handle argument '

src/verl/single_controller/base/decorator.py:355

↓ 1 callersMethodget_prefix_length

(self, prefix: str, text: str)

src/flashrag/refiner/llmlingua_compressor.py:1060

↓ 1 callersFunctionget_random_string

(length: int)

src/verl/single_controller/ray/base.py:30

↓ 1 callersMethodget_reasoning_chain

(self, all_query, all_doc_triples, triple_to_doc_ids)

src/flashrag/refiner/kg_refiner.py:379

↓ 1 callersFunctionget_refiner

(config, retriever=None, generator=None)

src/flashrag/utils/utils.py:124

↓ 1 callersMethodget_rerank_scores

Return flatten list of scores for each (query,doc) pair Args: query_list: List of N queries doc_list: Nested list of

src/flashrag/retriever/reranker.py:22

↓ 1 callersMethodget_rng_state

collect rng state across data parallel ranks

src/verl/utils/checkpoint/megatron_checkpoint_manager.py:93

↓ 1 callersFunctionget_rope_index

Gets the position ids for Qwen2-VL, it should be generated before sharding the sequence. The batch dim has been removed and the input_ids sho

src/verl/models/transformers/qwen2_vl.py:32

↓ 1 callersMethodget_self_information

(self, text: str)

src/flashrag/refiner/selective_context_compressor.py:101

↓ 1 callersFunctionget_spec_layer_idx_from_weight_name

(config: PretrainedConfig, weight_name: str)

src/verl/workers/sharding_manager/patch/fsdp_vllm_patch.py:26

↓ 1 callersMethodget_structured_dynamic_compression_ratio

( self, context: list, iterative_size: int, dynamic_ratio: list, start

src/flashrag/refiner/llmlingua_compressor.py:1130

↓ 1 callersMethodget_supported_archs

()

src/verl/models/registry.py:52

↓ 1 callersFunctionget_ulysses_sequence_parallel_rank

Get ulysses sequence parallel rank.

src/verl/utils/ulysses.py:53

↓ 1 callersFunctionget_weight_loader

(arch: str)

src/verl/models/weight_loader_registry.py:16

↓ 1 callersFunctionget_weight_saver

(arch: str)

src/verl/models/weight_loader_registry.py:31

↓ 1 callersFunctionhandle_base

(x)

src/verl/utils/reward_score/prime_math/grader.py:138

↓ 1 callersFunctionhandle_pi

(string, pi)

src/verl/utils/reward_score/prime_math/grader.py:147

↓ 1 callersFunctionhash_object

Returns a character hash code of arbitrary Python objects.

src/flashrag/utils/utils.py:169

↓ 1 callersMethodinit_llmlingua2

( self, max_batch_size: int = 50, max_force_token: int = 100, )

src/flashrag/refiner/llmlingua_compressor.py:257

↓ 1 callersFunctioninit_logger

(log_path: str, log_name: str)

scripts/evaluation/utils.py:44

↓ 1 callersMethodinit_megatron

(self, default_megatron_kwargs: Optional[Dict] = None)

src/verl/single_controller/ray/megatron.py:58

↓ 1 callersMethodinit_prompt

(self, func_schemas, question)

src/re_call/inference/re_call.py:47

↓ 1 callersFunctioninit_retriever

(args)

scripts/serving/retriever_serving.py:19

↓ 1 callersMethodinit_workers

Init resource pool and worker group

src/verl/trainer/ppo/ray_trainer.py:610

↓ 1 callersFunctioninitialize_global_process_group

(timeout_second=36000)

src/verl/utils/distributed.py:18

↓ 1 callersFunctionis_begin_of_new_word

(token, model_name, force_tokens, token_map)

src/flashrag/refiner/llmlingua_compressor.py:81

↓ 1 callersFunctionis_correct_minerva

Check if the solution is correct according to Minerva criteria. Args: solution_str: The solution string to check gt: The grou

src/verl/utils/reward_score/math_dapo.py:183

↓ 1 callersFunctionis_correct_strict_box

Check if the prediction is correct using strict boxed answer criteria. Args: pred: The prediction string gt: The ground truth

src/verl/utils/reward_score/math_dapo.py:212

↓ 1 callersFunctionis_equiv

(str1, str2, verbose=False)

src/verl/utils/reward_score/math.py:32

↓ 1 callersFunctionis_megatron_core_available

()

src/verl/utils/import_utils.py:25

↓ 1 callersMethoditerative_compress_prompt

( self, context: List[str], target_token: float, iterative_size: int = 200,

src/flashrag/refiner/llmlingua_compressor.py:1540

↓ 1 callersMethodjudge_sent_confidence

(self, sent, sent_score)

src/flashrag/pipeline/active_pipeline.py:846

↓ 1 callersFunctionkarmarkar_karp

(seqlen_list: List[int], k_partitions: int, equal_size: bool)

src/verl/utils/seqlen_balancing.py:25

↓ 1 callersFunctionlast_boxed_only_string

(string)

src/verl/utils/reward_score/math.py:63

↓ 1 callersMethodload_all_retriever

(self, config)

src/flashrag/retriever/retriever.py:589

↓ 1 callersMethodload_checkpoint

(self, checkpoint_path, hdfs_path=None, del_local_after_load=True)

src/verl/workers/megatron_workers.py:678

↓ 1 callersMethodload_clip_model

(self)

src/flashrag/retriever/encoder.py:154

↓ 1 callersMethodload_corpus

(self, corpus)

src/flashrag/retriever/retriever.py:344

↓ 1 callersMethodload_image

(self, image, input_size=448, max_num=12)

src/flashrag/generator/multimodal_generator.py:228

↓ 1 callersMethodload_index

(self)

src/flashrag/retriever/retriever.py:350

↓ 1 callersFunctionload_megatron_model_weights

(config, model_config, parallel_model,

src/verl/utils/model.py:277

↓ 1 callersFunctionload_megatron_param_and_grad

(module_list: nn.ModuleList, device_id, load_grad=False, hybrid_engine=None)

src/verl/utils/megatron_utils.py:248

↓ 1 callersMethodload_model

(self)

src/flashrag/retriever/retriever.py:372

↓ 1 callersMethodload_model

(self, model_name: str, device_map: str = "cuda", model_config: dict = {})

src/flashrag/refiner/llmlingua_compressor.py:273

↓ 1 callersMethodload_model_cls

(model_arch: str, value=False)

src/verl/models/registry.py:36

↓ 1 callersMethodload_model_corpus

(self, corpus)

src/flashrag/retriever/retriever.py:228

↓ 1 callersMethodload_optimizer

(self, ckpt_path)

src/verl/utils/checkpoint/megatron_checkpoint_manager.py:169

↓ 1 callersMethodload_params_to_cuda

load all model params to cuda

src/verl/workers/sharding_manager/megatron_vllm.py:133

↓ 1 callersMethodload_params_to_cuda

(self)

src/verl/workers/reward_model/megatron/reward_model.py:262

↓ 1 callersMethodload_prompts

(self)

src/flashrag/pipeline/branching_pipeline.py:77

↓ 1 callersFunctionload_replug_model

(name_or_path)

src/flashrag/pipeline/replug_utils.py:225

↓ 1 callersMethodload_rng_state

(rng_state)

src/verl/utils/checkpoint/checkpoint_manager.py:117

↓ 1 callersMethodload_rng_states

(self, ckpt_path, data_parallel_random_init=False, use_dist_ckpt=False)

src/verl/utils/checkpoint/megatron_checkpoint_manager.py:175

↓ 1 callersMethodload_special_tokens

(self, tokenizer, use_grounding, use_utility)

src/flashrag/pipeline/active_pipeline.py:265

↓ 1 callersFunctionload_state_dict_to_megatron_gptmodel

Load merged state_dict to sharded Megatron module in training.

src/verl/models/mcore/loader.py:52

↓ 1 callersMethodlog_generations_to_mlflow

Log validation generation to mlflow as artifacts

src/verl/utils/tracking.py:240

↓ 1 callersMethodlog_generations_to_swanlab

Log samples to swanlab as text

src/verl/utils/tracking.py:218

↓ 1 callersMethodlog_generations_to_wandb

Log samples to wandb as a table

src/verl/utils/tracking.py:191

↓ 1 callersFunctionlog_seqlen_unbalance

(seqlen_list: List[int], partitions: List[List[int]], prefix)

src/verl/utils/seqlen_balancing.py:186

↓ 1 callersFunctionlogprobs_from_logits_flash_attn

(logits, labels)

src/verl/utils/torch_functional.py:64

↓ 1 callersFunctionlogprobs_from_logits_v2

A memory efficient implementation of logprobs_from_logits

src/verl/utils/torch_functional.py:77

↓ 1 callersFunctionlower

(text)

src/flashrag/evaluator/utils.py:16

↓ 1 callersFunctionlower

(text)

src/verl/utils/reward_score/re_call.py:59

↓ 1 callersFunctionmain

()

src/flashrag/retriever/index_builder.py:364

↓ 1 callersFunctionmain

(config)

src/verl/trainer/main_ppo.py:58

↓ 1 callersFunctionmain

(config)

src/verl/trainer/main_generation.py:40

↓ 1 callersFunctionmain

(config)

src/verl/trainer/main_eval.py:70

↓ 1 callersFunctionmain

(config)

src/verl/trainer/fsdp_sft_trainer.py:528

↓ 1 callersMethodmask_a_sent

(self, sent, level)

src/flashrag/refiner/selective_context_compressor.py:260

↓ 1 callersFunctionmasked_var

Compute variance of tensor with masked values.

src/verl/utils/torch_functional.py:123

← previous next →601–700 of 1,539, ranked by callers