MCPcopy Create free account

hub / github.com/PRIME-RL/PRIME / functions

Functions2,604 in github.com/PRIME-RL/PRIME

Method__init__
(self, module: nn.Module, config)
training/verl/workers/rollout/hf_rollout.py:37
Method__init__
A naive rollout. It requires the module to be compatible with huggingface APIs. That is: The module should define __call__ to receive input_id
training/verl/workers/rollout/naive/naive_rollout.py:38
Method__init__
A vLLM rollout. It requires the module is supported by the vllm. Args: module: module here follows huggingface APIs c
training/verl/workers/rollout/vllm_rollout/vllm_rollout.py:59
Method__init__
MeagtronPPOActor class. This class implements the simple PPO logics when the model is built with Megatron. Args: config (OmegaCon
training/verl/workers/actor/megatron_actor.py:50
Method__init__
The base class for PPO actor Args: config (DictConfig): a config passed to the PPOActor. We expect the type to be
training/verl/workers/actor/base.py:28
Method__init__
When optimizer is None, it is Reference Policy
training/verl/workers/actor/dp_prime.py:39
Method__init__
When optimizer is None, it is Reference Policy
training/verl/workers/actor/dp_actor.py:35
Method__init__
(self, module: FSDP, inference_engine: LLM, model_config, full_params: bool = False)
training/verl/workers/hybrid_engine/fsdp_vllm.py:35
Method__init__
(self, model_provider)
training/verl/workers/hybrid_engine/megatron_vllm.py:37
Method__init__
(self, module: AllGatherPPModel, inference_engine: LLM, model_config, layer_name_mapping)
training/verl/workers/hybrid_engine/megatron_vllm.py:240
Method__init__
(self, config)
training/verl/workers/reward_model/base.py:25
Method__init__
(self, config, model_config, reward_model_module: torch.nn.
training/verl/workers/reward_model/megatron/reward_model.py:39
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig)
training/verl/models/llama/megatron/modeling_llama_megatron.py:158
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig)
training/verl/models/llama/megatron/modeling_llama_megatron.py:224
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig)
training/verl/models/llama/megatron/modeling_llama_megatron.py:282
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig, pre_process, post_process)
training/verl/models/llama/megatron/modeling_llama_megatron.py:411
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig, pre_process, post_process)
training/verl/models/llama/megatron/modeling_llama_megatron.py:521
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig)
training/verl/models/llama/megatron/layers/parallel_decoder.py:101
Method__init__
(self, config, megatron_config: ModelParallelConfig = None)
training/verl/models/llama/megatron/layers/parallel_mlp.py:33
Method__init__
(self, dim, max_position_embeddings=2048, base=10000, device=None, scaling_factor=1.0)
training/verl/models/llama/megatron/layers/parallel_attention.py:75
Method__init__
(self, dim, max_position_embeddings=2048, base=10000, device=None, scaling_factor=1.0)
training/verl/models/llama/megatron/layers/parallel_attention.py:94
Method__init__
(self, config: LlamaConfig, megatron_config: ModelParallelConfig)
training/verl/models/llama/megatron/layers/parallel_attention.py:146
Method__init__
(self, input_size, gate_ouput_size, up_output_size,
training/verl/models/llama/megatron/layers/parallel_linear.py:54
Method__init__
LlamaRMSNorm is equivalent to T5LayerNorm
training/verl/models/llama/megatron/layers/parallel_rmsnorm.py:27
Method__init__
( self, hf_config: PretrainedConfig, dtype: str, seed: int, load_forma
training/verl/third_party/vllm/vllm_v_0_3_1/config.py:75
Method__init__
( self, block_size: int, gpu_memory_utilization: float, swap_space: int,
training/verl/third_party/vllm/vllm_v_0_3_1/config.py:251
Method__init__
( self, pipeline_parallel_size: int, tensor_parallel_size: int, worker_use_ray
training/verl/third_party/vllm/vllm_v_0_3_1/config.py:329
Method__init__
( self, max_num_batched_tokens: Optional[int], max_num_seqs: int, max_model_le
training/verl/third_party/vllm/vllm_v_0_3_1/config.py:383
Method__init__
(self, device: str = "cuda")
training/verl/third_party/vllm/vllm_v_0_3_1/config.py:417
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict tokenizer:
training/verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py:70
Method__init__
(self, tokenizer: PreTrainedTokenizer, enable_lora: bool, max_num_seqs: int, max_input_length
training/verl/third_party/vllm/vllm_v_0_3_1/tokenizer.py:28
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict model_conf
training/verl/third_party/vllm/vllm_v_0_3_1/model_runner.py:48
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict model_conf
training/verl/third_party/vllm/vllm_v_0_3_1/worker.py:47
Method__init__
(self, hf_config: PretrainedConfig, *args, **kwargs)
training/verl/third_party/vllm/vllm_v_0_6_3/config.py:46
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict tokenizer
training/verl/third_party/vllm/vllm_v_0_6_3/llm.py:85
Method__init__
( self, # NOTE(sgm): first two arguments are added for verl model: Union[nn.Module, Di
training/verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py:95
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py:107
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py:138
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py:238
Method__init__
(self, tokenizer: PreTrainedTokenizer, enable_lora: bool, max_num_seqs: int, max_input_length
training/verl/third_party/vllm/vllm_v_0_6_3/tokenizer.py:26
Method__init__
( self, model: Union[nn.Module, Dict], # [verl] model itself or its parameter dict mo
training/verl/third_party/vllm/vllm_v_0_6_3/model_runner.py:62
Method__init__
( self, model, # pytorch model itself or its parameter dict model_config: ModelConfig
training/verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py:45
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict model_con
training/verl/third_party/vllm/vllm_v_0_6_3/worker.py:61
Method__init__
( self, hf_config: PretrainedConfig, tokenizer_mode: str, trust_remote_code: b
training/verl/third_party/vllm/vllm_v_0_5_4/config.py:99
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict tokenizer:
training/verl/third_party/vllm/vllm_v_0_5_4/llm.py:97
Method__init__
( self, # NOTE(sgm): first two arguments are added for verl model: Union[nn.Module, Di
training/verl/third_party/vllm/vllm_v_0_5_4/llm_engine_sp.py:77
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py:106
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py:128
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py:215
Method__init__
(self, tokenizer: PreTrainedTokenizer, enable_lora: bool, max_num_seqs: int, max_input_length
training/verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py:28
Method__init__
( self, model: Union[nn.Module, Dict], # [verl] model itself or its parameter dict mod
training/verl/third_party/vllm/vllm_v_0_5_4/model_runner.py:54
Method__init__
( self, model, # pytorch model itself or its parameter dict model_config: ModelConfig,
training/verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py:37
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict model_conf
training/verl/third_party/vllm/vllm_v_0_5_4/worker.py:52
Method__init__
( self, hf_config: PretrainedConfig, dtype: str, seed: int, revision:
training/verl/third_party/vllm/vllm_v_0_4_2/config.py:98
Method__init__
( self, # NOTE(sgm): first two arguments are added for verl model: Union[nn.Module, Di
training/verl/third_party/vllm/vllm_v_0_4_2/llm_engine_sp.py:74
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_4_2/model_loader.py:97
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_4_2/model_loader.py:118
Method__init__
(self, load_config: LoadConfig)
training/verl/third_party/vllm/vllm_v_0_4_2/model_loader.py:203
Method__init__
(self, tokenizer: PreTrainedTokenizer, enable_lora: bool, max_num_seqs: int, max_input_length
training/verl/third_party/vllm/vllm_v_0_4_2/tokenizer.py:28
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict model_conf
training/verl/third_party/vllm/vllm_v_0_4_2/model_runner.py:50
Method__init__
( self, model, # pytorch model itself or its parameter dict model_config: ModelConfig,
training/verl/third_party/vllm/vllm_v_0_4_2/spmd_gpu_executor.py:36
Method__init__
( self, model: Union[nn.Module, Dict], # model itself or its parameter dict model_conf
training/verl/third_party/vllm/vllm_v_0_4_2/worker.py:50
Method__init__
(self, max_per_second)
data_preprocessing/stage4_judge.py:21
Method__init__
(self, model_path: str, tp_size: int, try_num: int, use_fp8)
data_preprocessing/stage4_judge.py:35
Method__init__
(self, input_file: str, output_path: str)
data_preprocessing/stage4_judge.py:264
Method__iter__
(self)
eval/Math/Qwen25-Math/evaluation/python_executor.py:68
Method__len__
(self)
training/verl/protocol.py:114
Method__len__
(self)
training/verl/utils/dataset/rm_dataset.py:96
Method__len__
(self)
training/verl/utils/dataset/sft_dataset.py:107
Method__len__
(self)
training/verl/utils/dataset/rl_dataset.py:121
Method__len__
(self)
training/verl/utils/dataset/rl_dataset.py:171
Method__new__
(cls, *args, **kwargs)
training/verl/single_controller/base/worker.py:85
Method__post_init__
(self)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/benchmarks/code_generation.py:35
Method__post_init__
(self)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/benchmarks/code_generation.py:56
Method__post_init__
(self)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/benchmarks/test_output_prediction.py:29
Method__post_init__
(self)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/benchmarks/code_execution.py:23
Method__post_init__
(self)
training/verl/protocol.py:110
Method__post_init__
(self)
training/verl/third_party/vllm/vllm_v_0_3_1/config.py:431
Method__post_init__
(self)
training/verl/third_party/vllm/vllm_v_0_6_3/config.py:80
Method__post_init__
(self)
training/verl/third_party/vllm/vllm_v_0_6_3/arg_utils.py:30
Method__post_init__
(self)
training/verl/third_party/vllm/vllm_v_0_5_4/config.py:221
Method__post_init__
(self)
training/verl/third_party/vllm/vllm_v_0_4_2/config.py:180
Method__run_single
(counter)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/runner/cohere_runner.py:27
Method__run_single
(counter)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/runner/claude_runner.py:26
Method__run_single
(counter)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/runner/mistral_runner.py:28
Method__run_single
(counter)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/runner/deepseek_runner.py:35
Method__run_single
(counter)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/runner/gemini_runner.py:46
Method__run_single
(counter)
eval/Coding/livecodebench/LiveCodeBench-main/lcb_runner/runner/claude3_runner.py:26
Method__setstate__
(self, data)
training/verl/protocol.py:144
Method_decode_sequence
Decodes the new token for a sequence.
training/verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py:662
Method_download_files
()
training/verl/utils/dataset/rm_dataset.py:72
Method_finalize_sequence
(self, seq: Sequence, sampling_params: SamplingParams, stop_string: str)
training/verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py:710
Method_forward_head
(self, hidden_states)
training/verl/models/llama/megatron/modeling_llama_megatron.py:378
Method_forward_head
(self, hidden_states)
training/verl/models/llama/megatron/modeling_llama_megatron.py:644
Function_get_logits
(self, hidden_states: torch.Tensor, embedding: torch.Tensor, embedding_bias: Optional[torch.Te
training/verl/third_party/vllm/vllm_v_0_3_1/model_loader.py:193
Function_get_logits
(self, hidden_states: torch.Tensor, embedding: torch.Tensor, embedding_bias: Optional[torch.Te
training/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py:291
Function_get_logits
(self, hidden_states: torch.Tensor, embedding: torch.Tensor, embedding_bias: Optional[torch.Te
training/verl/third_party/vllm/vllm_v_0_5_4/model_loader.py:263
Function_get_logits
(self, hidden_states: torch.Tensor, embedding: torch.Tensor, embedding_bias: Optional[torch.Te
training/verl/third_party/vllm/vllm_v_0_4_2/model_loader.py:250
Method_get_pid
(self)
training/verl/single_controller/base/worker.py:66
Method_get_weights_iterator
(actor_model: Union[PreTrainedModel, Dict])
training/verl/third_party/vllm/vllm_v_0_6_3/model_loader.py:147
← previousnext →1,501–1,600 of 2,604, ranked by callers