MCPcopy Create free account

hub / github.com/CarlanLark/Lp-Reg-dev / functions

Functions1,299 in github.com/CarlanLark/Lp-Reg-dev

⨍ Functions 1,299◇ Types & classes 184↳ Endpoints 27
↓ 1 callersFunction_fetch_tp_shard_tensor_qkv
fetch tensor in tp shards across mp_group
verl/models/llama/megatron/checkpoint_utils/llama_loader.py:154
↓ 1 callersFunction_fetch_tp_shard_tensor_vocab
fetch tensor in tp shards
verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py:98
↓ 1 callersFunction_fetch_tp_shard_tensor_vocab
fetch tensor in tp shards
verl/models/llama/megatron/checkpoint_utils/llama_loader.py:100
↓ 1 callersFunction_fix_a_slash_b
(string)
verl/utils/reward_score/prime_math/__init__.py:250
↓ 1 callersFunction_fix_a_slash_b
(string)
verl/utils/reward_score/prime_math/math_normalize.py:90
↓ 1 callersFunction_fix_fracs
(string)
verl/utils/reward_score/prime_math/__init__.py:219
↓ 1 callersFunction_fix_fracs
(string)
verl/utils/reward_score/prime_math/math_normalize.py:58
↓ 1 callersFunction_fix_sqrt
(string)
verl/utils/reward_score/prime_math/__init__.py:273
↓ 1 callersFunction_fix_sqrt
(string)
verl/utils/reward_score/prime_math/math_normalize.py:115
↓ 1 callersFunction_flatten_dict
(raw: Dict[str, Any], *, sep: str)
verl/utils/tracking.py:184
↓ 1 callersMethod_forward_head
(self, hidden_states)
verl/models/qwen2/megatron/modeling_qwen2_megatron.py:304
↓ 1 callersMethod_forward_head
(self, hidden_states)
verl/models/qwen2/megatron/modeling_qwen2_megatron.py:608
↓ 1 callersMethod_forward_head
(self, hidden_states)
verl/models/llama/megatron/modeling_llama_megatron.py:304
↓ 1 callersMethod_forward_head
(self, hidden_states)
verl/models/llama/megatron/modeling_llama_megatron.py:561
↓ 1 callersMethod_forward_micro_batch
(self, micro_batch)
verl/workers/fsdp_workers.py:1261
↓ 1 callersFunction_fused_linear_for_ppo_bwd
( dlog_probs: Optional[torch.FloatTensor], dentropy: Optional[torch.FloatTensor], hidden_states: t
verl/utils/experimental/torch_functional.py:39
↓ 1 callersFunction_fused_linear_for_ppo_fwd
( hidden_states: torch.FloatTensor, vocab_weights: torch.FloatTensor, input_ids: torch.LongTensor,
verl/utils/experimental/torch_functional.py:19
↓ 1 callersMethod_generate_minibatch
(self, prompts: DataProto)
verl/workers/rollout/hf_rollout.py:53
↓ 1 callersFunction_get_free_port
()
verl/workers/rollout/async_server.py:44
↓ 1 callersMethod_get_free_port
(self)
verl/single_controller/base/worker.py:62
↓ 1 callersMethod_get_node_ip
(self)
verl/single_controller/base/worker.py:45
↓ 1 callersFunction_get_parallel_model_architecture_from_config
(config: PretrainedConfig, value=False)
verl/utils/model.py:274
↓ 1 callersFunction_get_qualified_name
Get full qualified name including module and class (if any).
verl/utils/import_utils.py:86
↓ 1 callersMethod_init_head
(self, config: Qwen2Config)
verl/models/qwen2/megatron/modeling_qwen2_megatron.py:290
↓ 1 callersMethod_init_head
(self, config)
verl/models/qwen2/megatron/modeling_qwen2_megatron.py:547
↓ 1 callersMethod_init_head
(self, config)
verl/models/llama/megatron/modeling_llama_megatron.py:290
↓ 1 callersMethod_init_head
(self, config)
verl/models/llama/megatron/modeling_llama_megatron.py:547
↓ 1 callersMethod_init_rope
(self)
verl/models/qwen2/megatron/layers/parallel_attention.py:211
↓ 1 callersMethod_init_rope
(self)
verl/models/llama/megatron/layers/parallel_attention.py:232
↓ 1 callersMethod_init_with_detached_workers
(self, worker_names, worker_handles)
verl/single_controller/ray/base.py:224
↓ 1 callersMethod_init_with_resource_pool
(self, resource_pool, ray_cls_with_init, bin_pack, detached)
verl/single_controller/ray/base.py:233
↓ 1 callersFunction_inject_implicit_mixed_number
Automatically make a mixed number evalable e.g. 7 3/4 => 7+3/4
verl/utils/reward_score/prime_math/__init__origin.py:105
↓ 1 callersFunction_inject_implicit_mixed_number
Automatically make a mixed number evalable e.g. 7 3/4 => 7+3/4
verl/utils/reward_score/prime_math/__init__.py:767
↓ 1 callersFunction_inner_call_method
(_method)
verl/utils/reward_score/prime_code/testing_util.py:545
↓ 1 callersFunction_is_float
(num: str)
verl/utils/reward_score/prime_math/__init__origin.py:71
↓ 1 callersFunction_is_float
(num: str)
verl/utils/reward_score/prime_math/__init__.py:733
↓ 1 callersFunction_is_int
(x: float)
verl/utils/reward_score/prime_math/__init__origin.py:79
↓ 1 callersFunction_is_int
(x: float)
verl/utils/reward_score/prime_math/__init__.py:741
↓ 1 callersMethod_is_worker_alive
(self, worker)
verl/single_controller/base/worker_group.py:118
↓ 1 callersFunction_last_boxed_only_string
(string)
verl/utils/reward_score/prime_math/__init__origin.py:307
↓ 1 callersMethod_load_params_to_cuda
(self, pp_rank, to_empty=False)
verl/workers/sharding_manager/megatron_vllm.py:132
↓ 1 callersFunction_make_causal_mask
Make causal mask used for bi-directional self-attention.
verl/utils/torch_functional.py:501
↓ 1 callersFunction_make_causal_mask
Make causal mask used for bi-directional self-attention.
verl/models/qwen2/megatron/modeling_qwen2_megatron.py:47
↓ 1 callersFunction_make_causal_mask
Make causal mask used for bi-directional self-attention.
verl/models/llama/megatron/modeling_llama_megatron.py:47
↓ 1 callersFunction_map_each_response
(resp)
verl/workers/rollout/sglang_rollout/sglang_rollout.py:71
↓ 1 callersMethod_maybe_log_val_generations
Log a table of validation samples to the configured logger (wandb or swanlab)
verl/trainer/ppo/ray_trainer.py:532
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader_depracated.py:21
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader.py:21
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/qwen2/megatron/checkpoint_utils/qwen2_saver.py:38
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/mcore/saver.py:47
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/mcore/loader.py:24
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/llama/megatron/checkpoint_utils/llama_saver.py:38
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/llama/megatron/checkpoint_utils/llama_loader.py:21
↓ 1 callersFunction_megatron_calc_layer_map
Calculate the mapping of global layer_idx to local layer_idx Returns: layer_map (Dict: int -> tuple(int, int, int)): mapping f
verl/models/llama/megatron/checkpoint_utils/llama_loader_depracated.py:21
↓ 1 callersFunction_mkdir
hdfs mkdir
verl/utils/hdfs_io.py:75
↓ 1 callersMethod_normalize_config_bsz
(self)
verl/trainer/fsdp_sft_trainer.py:112
↓ 1 callersMethod_optimizer_step
(self)
verl/workers/critic/dp_critic.py:113
↓ 1 callersMethod_optimizer_step
(self)
verl/workers/actor/dp_actor.py:347
↓ 1 callersFunction_parse_latex
Attempts to parse latex to an expression sympy can read.
verl/utils/reward_score/prime_math/__init__origin.py:53
↓ 1 callersFunction_parse_latex
Attempts to parse latex to an expression sympy can read.
verl/utils/reward_score/prime_math/__init__.py:715
↓ 1 callersMethod_postprocess
(self, batch: DataProto, batch_conversations: List[List[List[Dict[str, str]]]], n: int)
verl/trainer/naive_chat_scheduler.py:92
↓ 1 callersFunction_pre_process_inputs
(pad_token_id, prompt_token_ids: torch.Tensor)
verl/workers/rollout/vllm_rollout/fire_vllm_rollout.py:49
↓ 1 callersFunction_pre_process_inputs
(pad_token_id, prompt_token_ids: torch.Tensor)
verl/workers/rollout/vllm_rollout/vllm_rollout.py:58
↓ 1 callersFunction_pre_process_inputs
(pad_token_id, prompt_token_ids: torch.Tensor)
verl/workers/rollout/vllm_rollout/vllm_rollout_spmd.py:59
↓ 1 callersMethod_prepare_decoder_attention_mask
(self, attention_mask, input_shape, inputs_embeds)
verl/models/qwen2/megatron/modeling_qwen2_megatron.py:97
↓ 1 callersMethod_prepare_decoder_attention_mask
(self, attention_mask, input_shape, inputs_embeds)
verl/models/llama/megatron/modeling_llama_megatron.py:97
↓ 1 callersMethod_preprocess_prompt_to_async_rollout_requests
(self, prompts: DataProto, n: int)
verl/workers/rollout/sglang_rollout/async_sglang_rollout.py:688
↓ 1 callersFunction_preprocess_tensor_for_update_weights
(tensor: torch.Tensor)
verl/workers/sharding_manager/fsdp_sglang.py:58
↓ 1 callersMethod_read_files_and_process
(self)
verl/utils/dataset/multiturn_sft_dataset.py:60
↓ 1 callersMethod_read_files_and_tokenize
(self)
verl/utils/dataset/rm_dataset.py:85
↓ 1 callersMethod_read_files_and_tokenize
(self)
verl/utils/dataset/sft_dataset.py:74
↓ 1 callersFunction_remove_right_units
(string)
verl/utils/reward_score/prime_math/__init__.py:264
↓ 1 callersFunction_remove_right_units
(string)
verl/utils/reward_score/prime_math/math_normalize.py:105
↓ 1 callersMethod_save_evaluation_results
(self, data, save_tensor=False)
verl/trainer/ppo/ray_trainer.py:556
↓ 1 callersFunction_split_args_kwargs_data_proto_with_auto_padding
(chunks, *args, **kwargs)
verl/single_controller/base/decorator.py:80
↓ 1 callersMethod_start_fastapi_server
(self)
verl/workers/rollout/async_server.py:59
↓ 1 callersFunction_str_to_int
(x: str)
verl/utils/reward_score/prime_math/__init__origin.py:99
↓ 1 callersFunction_str_to_int
(x: str)
verl/utils/reward_score/prime_math/__init__.py:761
↓ 1 callersFunction_strip_string
(string)
verl/utils/reward_score/prime_math/__init__.py:218
↓ 1 callersFunction_strip_string
(string)
verl/utils/reward_score/prime_math/math_normalize.py:130
↓ 1 callersMethod_switch_chat_template
(self, data: DataProto)
verl/workers/fsdp_workers.py:1319
↓ 1 callersFunction_sympy_parse
Parses an expression with sympy.
verl/utils/reward_score/prime_math/__init__origin.py:44
↓ 1 callersFunction_sympy_parse
Parses an expression with sympy.
verl/utils/reward_score/prime_math/__init__.py:703
↓ 1 callersFunction_transform_params_to_json_serializable
(x, convert_list_to_dict: bool)
verl/utils/tracking.py:164
↓ 1 callersMethod_validate_config
(self)
verl/trainer/ppo/ray_trainer.py:330
↓ 1 callersMethod_validate_config
Validate config options not implemented for Megatron backend
verl/workers/critic/megatron_critic.py:79
↓ 1 callersMethod_validate_config
Validate config options not implemented for Megatron backend
verl/workers/actor/megatron_actor.py:137
↓ 1 callersFunctionacc_reward
(predict_str: str, ground_truth: str)
verl/utils/reward_score/geo3k.py:26
↓ 1 callersMethodadd_tool_response_message
Currently, we only support chatml format.
verl/workers/rollout/schemas.py:160
↓ 1 callersFunctionall_gather_tensor
(local_tensor: Tensor, group: Optional[dist.ProcessGroup] = None, async_op: bool = False)
verl/utils/ulysses.py:155
↓ 1 callersFunctionallgather_dict_tensors
TODO: optimize this. - We can use async ops - We can use only one allgather Args: tensors: size: group:
verl/utils/torch_functional.py:195
↓ 1 callersFunctionapply_rotary_pos_emb
(q, k, cos, sin, position_ids)
verl/models/llama/megatron/layers/parallel_attention.py:149
↓ 1 callersFunctionapply_rotary_pos_emb_rmpad_flash
(q, k, cos, sin, cu_seqlens, max_seqlen)
verl/models/qwen2/megatron/layers/parallel_attention.py:290
↓ 1 callersFunctionapply_rotary_pos_emb_rmpad_flash
(q, k, cos, sin, cu_seqlens, max_seqlen)
verl/models/llama/megatron/layers/parallel_attention.py:344
↓ 1 callersFunctionare_equal_under_sympy
(ground_truth_normalized: str, given_normalized: str)
verl/utils/reward_score/prime_math/__init__origin.py:213
↓ 1 callersFunctionare_equal_under_sympy
(ground_truth_normalized: str, given_normalized: str)
verl/utils/reward_score/prime_math/__init__.py:884
↓ 1 callersFunctionasync_server_class
Get async server class. Args: rollout_backend: str, rollout backend, should be "vllm" or "sglang". Returns: Type[AsyncServer
verl/workers/rollout/async_server.py:341
↓ 1 callersFunctionbroadcast_params
(module)
verl/models/qwen2/megatron/checkpoint_utils/qwen2_loader_depracated.py:63
↓ 1 callersFunctionbroadcast_params
(module)
verl/models/mcore/loader.py:66
↓ 1 callersFunctionbroadcast_params
(module)
verl/models/llama/megatron/checkpoint_utils/llama_loader_depracated.py:65
← previousnext →401–500 of 1,299, ranked by callers