MCPcopy Create free account

hub / github.com/Gen-Verse/OpenClaw-RL / functions

Functions10,561 in github.com/Gen-Verse/OpenClaw-RL

Method__init__
( self, config: TransformerConfig, submodules: CrossAttentionSubmodules, layer
Megatron-LM/megatron/core/transformer/attention.py:1389
Method__init__
( self, config: TransformerConfig, spec: Union[TransformerBlockSubmodules, ModuleSpec]
Megatron-LM/megatron/core/transformer/transformer_block.py:272
Method__init__
( self, config: TransformerConfig, submodules: MLPSubmodules, is_expert: bool
Megatron-LM/megatron/core/transformer/mlp.py:75
Method__init__
( self, config: TransformerConfig, spec: Union[TransformerBlockSubmodules, ModuleSpec]
Megatron-LM/megatron/core/transformer/multi_token_prediction.py:1081
Method__init__
(self, config: TransformerConfig)
Megatron-LM/megatron/core/transformer/module.py:39
Method__init__
(self, config: TransformerConfig, vp_stage: Optional[int] = None)
Megatron-LM/megatron/core/transformer/module.py:164
Method__init__
( self, config: MLATransformerConfig, submodules: Union[MLASelfAttentionSubmodules],
Megatron-LM/megatron/core/transformer/multi_latent_attention.py:92
Method__init__
( self, config: TransformerConfig, layer_number: int, attn_mask_type: AttnMask
Megatron-LM/megatron/core/transformer/dot_product_attention.py:45
Method__init__
(self, config: TransformerConfig, tp_comm_buffer_name: str, *args, **kwargs)
Megatron-LM/megatron/core/transformer/heterogeneous/linear_replacements.py:58
Method__init__
( self, config: TransformerConfig, submodules: DSAttentionSubmodules, layer_nu
Megatron-LM/megatron/core/transformer/experimental_attention_variant/dsa.py:704
Method__init__
Initialize the MoE Token Dispatcher. Args: config (TransformerConfig): Configuration for the MoE layer. pg_c
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:60
Method__init__
Initialize the AllGather based token dispatcher. Args: num_local_experts (int): Number of local experts. local_expert
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:216
Method__init__
Initialize the AlltoAll token dispatcher. Args: num_local_experts (int): Number of local experts on the current device.
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:367
Method__init__
Initialize the HybridEP dispatcher. Args: group (torch.distributed.ProcessGroup): The process group to use for communica
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:960
Method__init__
Initialize the Flex token dispatcher. Args: num_local_experts (int): Number of local experts on the current device.
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:1329
Method__init__
(self, moe_layer, return_step: str, **kwargs)
Megatron-LM/megatron/core/transformer/moe/moe_utils.py:1065
Method__init__
( self, config: TransformerConfig, submodules: MLPSubmodules, gate: bool,
Megatron-LM/megatron/core/transformer/moe/shared_experts.py:39
Method__init__
( self, num_local_experts, config: TransformerConfig, submodules: MLPSubmodule
Megatron-LM/megatron/core/transformer/moe/experts.py:741
Method__init__
( self, num_local_experts, config: TransformerConfig, submodules: MLPSubmodule
Megatron-LM/megatron/core/transformer/moe/experts.py:1049
Method__init__
( self, config: TransformerConfig, layer_number: Optional[int] = None, pg_coll
Megatron-LM/megatron/core/transformer/moe/moe_layer.py:53
Method__init__
Initialize the zero token dropping router. Args: config (TransformerConfig): The configuration for the transformer model.
Megatron-LM/megatron/core/transformer/moe/router.py:146
Method__init__
Args: config: The config of the model. submodules: Contains the module specs for the input and output linear layers.
Megatron-LM/megatron/core/ssm/gated_delta_net.py:76
Method__init__
( self, config: TransformerConfig, submodules: TransformerLayerSubmodules, lay
Megatron-LM/megatron/core/ssm/mlp_layer.py:16
Method__init__
( self, config: TransformerConfig, submodules: MambaStackSubmodules, residual_
Megatron-LM/megatron/core/ssm/mamba_block.py:73
Method__init__
( self, cp_group: torch.distributed.ProcessGroup, d_inner_local_tp: int, nhead
Megatron-LM/megatron/core/ssm/mamba_context_parallel.py:47
Method__init__
( self, config: TransformerConfig, submodules: MambaMixerSubmodules, d_model,
Megatron-LM/megatron/core/ssm/mamba_mixer.py:138
Method__init__
Initialize Mamba Layer.
Megatron-LM/megatron/core/ssm/mamba_layer.py:57
Method__init__
Input optimizer is the base optimizer (e.g., Adam).
Megatron-LM/megatron/core/optimizer/optimizer.py:109
Method__init__
( self, optimizer: torch.optim.Optimizer, config: OptimizerConfig, grad_scaler
Megatron-LM/megatron/core/optimizer/optimizer.py:451
Method__init__
( self, optimizer: torch.optim.Optimizer, config: OptimizerConfig, grad_scaler
Megatron-LM/megatron/core/optimizer/optimizer.py:640
Method__init__
( self, optimizer: torch.optim.Optimizer, config: OptimizerConfig, init_state_fn: Callable )
Megatron-LM/megatron/core/optimizer/optimizer.py:899
Method__init__
(self, chained_optimizers: List[MegatronOptimizer])
Megatron-LM/megatron/core/optimizer/optimizer.py:1086
Method__init__
( self, params: ParamsT, lr: float = 3e-4, momentum_beta: float = 0.95,
Megatron-LM/megatron/core/optimizer/muon.py:44
Method__init__
Distributed optimizer, for all data types (fp16, bf16, and fp32). The steps in this method create the core mapping between param and
Megatron-LM/megatron/core/optimizer/distrib_optimizer.py:456
Method__init__
Initialize scale value with the input initial scale.
Megatron-LM/megatron/core/optimizer/grad_scaler.py:12
Method__init__
Grad scaler with dynamic scale that gets adjusted during training. Args: initial_scale (float): Initial loss scale value
Megatron-LM/megatron/core/optimizer/grad_scaler.py:61
Method__init__
Initialize LayerWiseDistributedOptimizer. Args: optimizers: List of MegatronOptimizers. config: OptimizerCon
Megatron-LM/megatron/core/optimizer/layer_wise_optimizer.py:42
Method__init__
( self, params, offload_fraction=0.5, cpu_optimizer_cls=None, gpu_opti
Megatron-LM/megatron/core/optimizer/cpu_offloading/hybrid_optimizer.py:45
Method__init__
Constructor for the TRTLLMHelper There are two public API's supported by this helper. a) get_trtllm_pretrained_config_and_model_weig
Megatron-LM/megatron/core/export/trtllm/trtllm_helper.py:42
Method__init__
Constructor for the TRTLLMModelWeightsConverterCPU class This class is responsible to convert the model weights to TRTLLM equivalent weights
Megatron-LM/megatron/core/export/trtllm/trtllm_weights_converter/single_device_trtllm_model_weights_converter.py:43
Method__init__
Constructor for the TRTLLMModelWeightsConverterGPU class This class is responsible to convert the model weights to TRTLLM equivalent weights.
Megatron-LM/megatron/core/export/trtllm/trtllm_weights_converter/distributed_trtllm_model_weights_converter.py:37
Method__init__
(self)
Megatron-LM/megatron/rl/rl_utils.py:93
Method__init__
(self, bin_size: int, pad_token: int, max_sequences_per_bin: int = 16)
Megatron-LM/megatron/rl/rl_utils.py:359
Method__init__
(self, max_k=32, **kwargs)
Megatron-LM/megatron/rl/agent/pass_at_evaluation_agent.py:49
Method__init__
(self, agent_configs: list[AgentConfig])
Megatron-LM/megatron/rl/agent/weighted_multi_task.py:43
Method__init__
(self, sig: signal.Signals = signal.SIGTERM)
Megatron-LM/megatron/training/dist_signal_handler.py:51
Method__init__
(self, src_cfg_class: type, exclude: Optional[list[str]] = None)
Megatron-LM/megatron/training/argument_utils.py:66
Method__init__
(self)
Megatron-LM/megatron/training/log_handler.py:15
Method__init__
(self, dataset_path: str)
Megatron-LM/megatron/training/datasets/sft_dataset.py:30
Method__init__
( self, dataset: LowLevelDataset, dataset_path: Optional[str], indices: np.nda
Megatron-LM/megatron/training/datasets/sft_dataset.py:49
Method__init__
( self, indexed_dataset: IndexedDataset, dataset_path: str, indexed_indices: n
Megatron-LM/megatron/training/datasets/fim_dataset.py:54
Method__init__
( self, total_samples, consumed_samples, micro_batch_size, data_parall
Megatron-LM/megatron/training/datasets/data_samplers.py:104
Method__init__
(self, total_samples, consumed_samples, micro_batch_size, global_batch_size, data_parallel_ra
Megatron-LM/megatron/training/datasets/data_samplers.py:172
Method__init__
( self, dataset, total_samples, consumed_samples, micro_batch_size,
Megatron-LM/megatron/training/datasets/data_samplers.py:258
Method__init__
(self, vocab_file, merges_file, errors='replace', special_tokens=None, max_len=None)
Megatron-LM/megatron/training/tokenizer/gpt2_tokenization.py:157
Method__init__
(self, pretrained_model_name_or_path, trust_remote_code=False, **kwargs)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:127
Method__init__
(self, vocab_file, lower_case=True, vocab_extra_ids=0)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:197
Method__init__
(self, vocab_file, merge_file)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:333
Method__init__
(self, model_file, vocab_extra_ids=0)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:371
Method__init__
(self, model_file)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:541
Method__init__
(self, model_file)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:585
Method__init__
( self, path: str, pattern: str, vocab_size: Optional[int], num_specia
Megatron-LM/megatron/training/tokenizer/tokenizer.py:676
Method__init__
(self, vocab_size, image_token=None, image_token_id=None)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:876
Method__init__
Tokenizer with a support for non-text inputs. Note: Currently, only HuggingFaceTokenizer is supported as the underlying text tokenizer.
Megatron-LM/megatron/training/tokenizer/multimodal_tokenizer.py:74
Method__init__
Note: Currently, only HuggingFaceTokenizer is supported as the underlying text tokenizer. Args: tokenizer_path (str): Un
Megatron-LM/megatron/training/tokenizer/sft_tokenizer.py:17
Method__init__
(self, vocab_file, do_lower_case=True)
Megatron-LM/megatron/training/tokenizer/bert_tokenization.py:164
Method__init__
Constructs a BasicTokenizer. Args: do_lower_case: Whether to lower case the input.
Megatron-LM/megatron/training/tokenizer/bert_tokenization.py:220
Method__init__
(self, vocab, unk_token="[UNK]", max_input_chars_per_word=200)
Megatron-LM/megatron/training/tokenizer/bert_tokenization.py:335
Method__init__
( self, token: str, repo_name: str, slack_token: str = None, webhook_url: str = None )
Megatron-LM/tests/test_utils/python_scripts/auto_reminder_github.py:44
Method__init__
(self, token: str, repo_name: str, pr_number: str)
Megatron-LM/tests/test_utils/python_scripts/swap_pr_labels.py:40
Method__init__
(self, format: str, mp: ModelParallelState, transformer_impl: str = None)
Megatron-LM/tests/functional_tests/test_cases/common/ckpt_converter/__main__.py:102
Method__init__
(self, src: ModelMeta, dst: ModelMeta, num_moe_experts: T.Optional[int] = None)
Megatron-LM/tests/functional_tests/test_cases/common/ckpt_converter/__main__.py:558
Method__init__
( self, src: ModelMeta, dst: ModelMeta, language_model_type: str, vision_model_type: str )
Megatron-LM/tests/functional_tests/test_cases/common/ckpt_converter/__main__.py:600
Method__init__
( self, input_dim: int, output_dim: int, num_layers: int, bias: bool,
Megatron-LM/tests/unit_tests/test_utilities.py:12
Method__init__
(self, config)
Megatron-LM/tests/unit_tests/test_checkpointing.py:39
Method__init__
(self, state_dict)
Megatron-LM/tests/unit_tests/test_checkpointing.py:52
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_fp8_utils.py:39
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_utils.py:28
Method__init__
(self, input_size=80, hidden_size=48, output_size=10)
Megatron-LM/tests/unit_tests/test_layer_wise_optimizer.py:34
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_optimizer.py:38
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_optimizer_cpu_offloading.py:23
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_muon_optimizer.py:27
Method__init__
(self, batch=BATCH, seq=SEQ, vocab=VOCAB)
Megatron-LM/tests/unit_tests/test_rl_utils.py:26
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_rl_utils.py:50
Method__init__
(self)
Megatron-LM/tests/unit_tests/test_tokenizer.py:200
Method__init__
Initialize mock dataset Args: num_samples: Total number of samples sequence_length: Length of each sequence
Megatron-LM/tests/unit_tests/fusions/test_fused_linear_cross_entropy.py:91
Method__init__
(self, hidden_dim)
Megatron-LM/tests/unit_tests/distributed/test_mcore_fully_sharded_data_parallel.py:43
Method__init__
(self, input_dim, output_dim)
Megatron-LM/tests/unit_tests/distributed/test_distributed_data_parallel.py:17
Method__init__
( self, hidden_size: int, num_layers: int, num_moe_experts: int, moe_g
Megatron-LM/tests/unit_tests/distributed/test_grad_sync_with_expert_parallel.py:17
Method__init__
Initialize a dummy model with a few modules.
Megatron-LM/tests/unit_tests/distributed/test_torch_fully_sharded_parallel.py:28
Method__init__
(self, model_dim, num_heads, num_layers, output_dim)
Megatron-LM/tests/unit_tests/distributed/fsdp/test_mfsdp_fully_shard.py:93
Method__init__
(self, model_dim, num_heads, num_layers, output_dim)
Megatron-LM/tests/unit_tests/distributed/fsdp/test_mfsdp_fully_shard.py:112
Method__init__
(self, hidden_size: int = 2048, num_layers: int = 4, dtype=torch.bfloat16)
Megatron-LM/tests/unit_tests/pipeline_parallel/test_fine_grained_activation_offloading.py:21
Method__init__
( self, dataset: LowLevelDataset, dataset_path: Optional[str],
Megatron-LM/tests/unit_tests/data/test_builder.py:63
Method__init__
(self, offset: int, size: int)
Megatron-LM/tests/unit_tests/data/test_bin_reader.py:32
Method__init__
(self, *args: Any)
Megatron-LM/tests/unit_tests/data/test_bin_reader.py:70
Method__init__
(self)
Megatron-LM/tests/unit_tests/inference/test_data_parallel_inference_coordinator.py:41
Method__init__
We cannot call super().__init__() because it requires complex setup.
Megatron-LM/tests/unit_tests/inference/test_data_parallel_inference_coordinator.py:51
Method__init__
(self, value)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_safe_globals.py:15
Method__init__
(self, name: Union[str, Path], sync=True, ignore_cleanup_errors=False)
Megatron-LM/tests/unit_tests/dist_checkpointing/__init__.py:31
← previousnext →5,901–6,000 of 10,561, ranked by callers