MCPcopy Create free account

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

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

↓ 1 callersMethod__init__
(self, *args, **kwargs)
Megatron-LM/megatron/core/transformer/identity_op.py:10
↓ 1 callersMethod__init__
( self, config: TransformerConfig, submodules: MultiTokenPredictionLayerSubmodules,
Megatron-LM/megatron/core/transformer/multi_token_prediction.py:615
↓ 1 callersMethod__init__
( self, config: MLATransformerConfig, submodules: MLASelfAttentionSubmodules,
Megatron-LM/megatron/core/transformer/multi_latent_attention.py:396
↓ 1 callersMethod__init__
(self, config: TransformerConfig, *args, **kwargs)
Megatron-LM/megatron/core/transformer/heterogeneous/linear_replacements.py:87
↓ 1 callersMethod__init__
Initialize the indexer. Args: config (TransformerConfig): The configuration for the transformer model. submodules (DS
Megatron-LM/megatron/core/transformer/experimental_attention_variant/dsa.py:352
↓ 1 callersMethod__init__
( self, config: TransformerConfig, submodules: Optional[MoESubmodules] = None,
Megatron-LM/megatron/core/transformer/moe/moe_layer.py:105
↓ 1 callersMethod__init__
Initialize the Router module. Args: config (TransformerConfig): Configuration object for the Transformer model.
Megatron-LM/megatron/core/transformer/moe/router.py:30
↓ 1 callersMethod__init__
(self, start: int, end: int)
Megatron-LM/megatron/core/optimizer/distrib_optimizer.py:69
↓ 1 callersMethod__init__
(self, **data)
Megatron-LM/megatron/rl/agent/weighted_multi_task.py:32
↓ 1 callersMethod__init__
(self, dataset, seed)
Megatron-LM/megatron/training/datasets/data_samplers.py:225
↓ 1 callersMethod__init__
(self, in_features, out_features)
Megatron-LM/tests/unit_tests/test_fp8_utils.py:16
↓ 1 callersMethod__init__
(self)
Megatron-LM/tests/unit_tests/test_layer_wise_optimizer.py:54
↓ 1 callersMethod__init__
(self, input_dim, output_dim)
Megatron-LM/tests/unit_tests/distributed/test_mcore_fully_sharded_data_parallel.py:26
↓ 1 callersMethod__init__
(self)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_fully_parallel.py:55
↓ 1 callersMethod__init__
Custom RMSNorm for InternViT. Args: config (TransformerConfig): Config. hidden_size (int): Input hidden size.
Megatron-LM/examples/multimodal/nvlm/internvit.py:61
↓ 1 callersMethod__init__
( self, args, config, layer_number: int, cp_comm_type: str = "p2p",
slime/slime_plugins/models/qwen3_next.py:166
↓ 1 callersMethod__init__
( self, args, config, layer_number: int, cp_comm_type: str = "p2p",
slime/slime_plugins/models/qwen3_5.py:146
↓ 1 callersMethod__init__
(self, args)
slime/slime/utils/profile_utils.py:14
↓ 1 callersMethod__init__
(self, fn, xs)
slime/slime/backends/megatron_utils/update_weight/hf_weight_iterator_bridge.py:62
↓ 1 callersMethod__init__
(self, num_experts: int, num_local_experts: int)
slime/slime/backends/fsdp_utils/models/qwen3_moe.py:49
↓ 1 callersMethod__init__
(self, args)
slime/slime/rollout/data_source.py:45
↓ 1 callersMethod__new__
Constructor Creates an instance of the class if not created Args: cls (Type['StragglerDetector']): The class type
Megatron-LM/megatron/core/utils.py:1327
↓ 1 callersMethod__next__
(self)
Megatron-LM/megatron/core/inference/utils.py:22
↓ 1 callersMethod__post_init__
(self)
Megatron-LM/tools/retro/sft/dataset_conv.py:44
↓ 1 callersMethod__post_init__
Python dataclass method that is used to modify attributes after initialization. See https://docs.python.org/3/library/dataclasses.html#post-in
Megatron-LM/megatron/core/transformer/transformer_config.py:856
↓ 1 callersMethod__repr__
(self)
Megatron-LM/megatron/core/dist_checkpointing/strategies/checkpointable.py:195
↓ 1 callersMethod_add
Add to index (rank 0's method). Args: config (RetroPreprocessingConfig): Retro preprocessing config. text_dataset (GP
Megatron-LM/megatron/core/datasets/retro/index/indexes/faiss_base.py:108
↓ 1 callersFunction_add_autoresume_args
(parser)
Megatron-LM/megatron/training/arguments.py:3042
↓ 1 callersFunction_add_biencoder_args
(parser)
Megatron-LM/megatron/training/arguments.py:3054
↓ 1 callersFunction_add_checkpointing_args
(parser)
Megatron-LM/megatron/training/arguments.py:2495
↓ 1 callersFunction_add_config_logger_args
(parser)
Megatron-LM/megatron/training/arguments.py:1922
↓ 1 callersFunction_add_data_args
(parser)
Megatron-LM/megatron/training/arguments.py:2925
↓ 1 callersFunction_add_distributed_args
(parser)
Megatron-LM/megatron/training/arguments.py:2672
↓ 1 callersFunction_add_experimental_args
(parser)
Megatron-LM/megatron/training/arguments.py:3441
↓ 1 callersFunction_add_experimental_attention_variant_args
(parser)
Megatron-LM/megatron/training/arguments.py:3349
↓ 1 callersFunction_add_ft_package_args
(parser)
Megatron-LM/megatron/training/arguments.py:1911
↓ 1 callersFunction_add_heterogeneous_args
Heterogeneous models refer to transformer architectures where individual layers can differ in configuration. Specifically: - Attenti
Megatron-LM/megatron/training/arguments.py:3391
↓ 1 callersFunction_add_inference_args
(parser)
Megatron-LM/megatron/training/arguments.py:1492
↓ 1 callersFunction_add_initialization_args
(parser)
Megatron-LM/megatron/training/arguments.py:2404
↓ 1 callersFunction_add_inprocess_restart_args
(parser)
Megatron-LM/megatron/training/arguments.py:1833
↓ 1 callersFunction_add_kitchen_quantization_arguments
Add quant-specific arguments to the main parser If kitchen isn't available, nothing to do here, return unchanged parser
Megatron-LM/megatron/training/arguments.py:3512
↓ 1 callersFunction_add_learning_rate_args
(parser)
Megatron-LM/megatron/training/arguments.py:2432
↓ 1 callersFunction_add_logging_args
(parser)
Megatron-LM/megatron/training/arguments.py:1930
↓ 1 callersFunction_add_mixed_precision_args
(parser)
Megatron-LM/megatron/training/arguments.py:2629
↓ 1 callersFunction_add_mla_args
(parser)
Megatron-LM/megatron/training/arguments.py:3326
↓ 1 callersFunction_add_moe_args
(parser)
Megatron-LM/megatron/training/arguments.py:3184
↓ 1 callersFunction_add_msc_args
(parser)
Megatron-LM/megatron/training/arguments.py:3506
↓ 1 callersFunction_add_network_size_args
(parser)
Megatron-LM/megatron/training/arguments.py:1674
↓ 1 callersFunction_add_one_logger_args
(parser)
Megatron-LM/megatron/training/arguments.py:1882
↓ 1 callersFunction_add_regularization_args
(parser)
Megatron-LM/megatron/training/arguments.py:2011
↓ 1 callersFunction_add_rerun_machine_args
(parser)
Megatron-LM/megatron/training/arguments.py:2386
↓ 1 callersFunction_add_retro_args
(parser)
Megatron-LM/megatron/training/arguments.py:1626
↓ 1 callersFunction_add_rl_args
(parser)
Megatron-LM/megatron/training/arguments.py:2070
↓ 1 callersFunction_add_sft_args
(parser)
Megatron-LM/megatron/training/arguments.py:3542
↓ 1 callersFunction_add_straggler_detector_args
(parser)
Megatron-LM/megatron/training/arguments.py:1815
↓ 1 callersFunction_add_to_index
Add DB chunks to index. Args: config (RetroPreprocessingConfig): Retro preprocessing config. Returns: Path to the populated
Megatron-LM/megatron/core/datasets/retro/index/build.py:277
↓ 1 callersFunction_add_training_args
(parser)
Megatron-LM/megatron/training/arguments.py:2136
↓ 1 callersFunction_add_transformer_engine_args
(parser)
Megatron-LM/megatron/training/arguments.py:1419
↓ 1 callersFunction_add_validation_args
(parser)
Megatron-LM/megatron/training/arguments.py:2868
↓ 1 callersFunction_add_vision_args
(parser)
Megatron-LM/megatron/training/arguments.py:3111
↓ 1 callersFunction_add_workload_inspector_server_args
(parser)
Megatron-LM/megatron/training/arguments.py:1827
↓ 1 callersFunction_adv_diff_clip
()
openclaw-combine/openclaw_topk_select_loss.py:120
↓ 1 callersFunction_all_to_all_hp2cp
Perform AlltoAll communication on a context parallel group, transform the input tensor from shape [global-sequence, batch, local-hidden/c
Megatron-LM/megatron/core/ssm/mamba_context_parallel.py:324
↓ 1 callersFunction_alloc_storage
Allocate storage for ``tensor`` with the given size. Returns: bool: ``True`` if this method allocated storage and ``False`` if the
Megatron-LM/megatron/core/distributed/fsdp/src/megatron_fsdp/param_and_grad_buffer.py:115
↓ 1 callersFunction_allocate_auto
( total_layers_count: int, target_attention_ratio: float, target_mlp_ratio: float )
Megatron-LM/megatron/core/ssm/mamba_hybrid_layer_allocation.py:35
↓ 1 callersMethod_allocate_gpus_for_actor
(self, pg, num_gpus_per_actor)
slime/slime/ray/actor_group.py:46
↓ 1 callersFunction_allocate_override
(total_layers_count: int, override_pattern: str)
Megatron-LM/megatron/core/ssm/mamba_hybrid_layer_allocation.py:72
↓ 1 callersFunction_allocate_prm_engine_addr_and_ports
(*, args, num_engines, prm_engines)
slime/slime/ray/rollout.py:996
↓ 1 callersFunction_allocate_rollout_engine_addr_and_ports_external
(args, rollout_engines)
slime/slime/ray/rollout.py:980
↓ 1 callersFunction_allocate_rollout_engine_addr_and_ports_normal
(*, args, num_engines, rollout_engines)
slime/slime/ray/rollout.py:1052
↓ 1 callersFunction_allocate_vm
(region)
gui-rl/desktop_env/providers/azure/manager.py:16
↓ 1 callersFunction_allocate_vm
(region=DEFAULT_REGION, screen_size=(1920, 1080))
gui-rl/desktop_env/providers/aws/manager.py:54
↓ 1 callersFunction_allreduce_position_embedding_grads
All-reduce position_embeddings grad across encoder and decoder stages to ensure that position embeddings parameters stay in sync.
Megatron-LM/megatron/core/distributed/finalize_model_grads.py:262
↓ 1 callersFunction_analyze_patch_policy
(patch_text: str, instance: dict | None = None)
swe-rl/generate_with_swe_trajectory.py:279
↓ 1 callersFunction_append_hint_to_messages
(messages: list[dict], hint: str)
openclaw-fireworks/fireworks_server.py:227
↓ 1 callersFunction_append_visible_static_text
(container)
swe-rl/mini-swe-agent/tests/run/test_inspector.py:16
↓ 1 callersMethod_apply_aux_loss
Apply the auxiliary loss for the given scores and routing map.
Megatron-LM/megatron/core/transformer/moe/router.py:273
↓ 1 callersMethod_apply_bias
(intermediate_parallel, bias_parallel, tokens_per_expert, permuted_probs)
Megatron-LM/megatron/core/transformer/moe/experts.py:827
↓ 1 callersMethod_apply_chat_template
( self, messages: List[dict[str, Any]], tools: List[dict[str, Any]] | None, )
terminal-rl/inference_client.py:297
↓ 1 callersFunction_apply_dataset_field_overrides
( args: Any, dataset_cfg: dict[str, Any], defaults: dict[str, Any], spec_names: dict[str, Any] )
slime/slime/utils/eval_config.py:182
↓ 1 callersMethod_apply_expert_bias
Update expert bias and tokens_per_expert Prevent extra local tokens accumulation on evaluation or activation recomputation
Megatron-LM/megatron/core/transformer/moe/router.py:476
↓ 1 callersMethod_apply_gated_norm
(self, x, gate)
Megatron-LM/megatron/core/ssm/gated_delta_net.py:417
↓ 1 callersMethod_apply_global_aux_loss
Apply the global auxiliary loss for the given scores and routing map.
Megatron-LM/megatron/core/transformer/moe/router.py:345
↓ 1 callersMethod_apply_output_gate
(self, x, gate)
Megatron-LM/megatron/core/transformer/attention.py:1030
↓ 1 callersFunction_apply_rotary_pos_emb_thd
A baseline implementation of applying RoPE for `thd` format. Args: t (Tensor): Input tensor T is of shape [t, h, d] cu_seqlens(Te
Megatron-LM/megatron/core/models/common/embeddings/rope_utils.py:178
↓ 1 callersMethod_apply_scaling
( self, freqs, factor=8, low_freq_factor=1, high_freq_factor=4,
Megatron-LM/megatron/core/models/common/embeddings/rotary_pos_embedding.py:92
↓ 1 callersMethod_apply_seq_aux_loss
Apply the sequence-level auxiliary loss for the given scores and routing map. To calculate the sequence-level aux loss, we reshape the batch_
Megatron-LM/megatron/core/transformer/moe/router.py:301
↓ 1 callersMethod_apply_tile_tagging
Apply tile tagging. The image embeddings of multiple tiles are prepended with tile tags such as <tile_1>. This implements the method
Megatron-LM/megatron/core/models/multimodal/llava_model.py:754
↓ 1 callersFunction_assemble_full_tensor_from_uneven_chunks
Assemble the full tensor from unevenly sized chunks gathered from all ranks. Args: dtensor (DTensor): The original distributed tenso
Megatron-LM/megatron/core/distributed/fsdp/src/megatron_fsdp/uneven_dtensor.py:314
↓ 1 callersMethod_assert_has_kv_scales
(self, state_dict, kv_quantized)
Megatron-LM/tests/unit_tests/export/trtllm/test_single_device_fp8.py:152
↓ 1 callersMethod_assert_has_kv_scales
(self, state_dict, kv_quantized)
Megatron-LM/tests/unit_tests/export/trtllm/test_distributed_fp8.py:155
↓ 1 callersMethod_assert_has_scales
(self, state_dict, quantized)
Megatron-LM/tests/unit_tests/export/trtllm/test_single_device_fp8.py:139
↓ 1 callersMethod_assert_has_scales
(self, state_dict, quantized)
Megatron-LM/tests/unit_tests/export/trtllm/test_distributed_fp8.py:142
↓ 1 callersMethod_assert_non_quantizable_layers
(self, state_dict)
Megatron-LM/tests/unit_tests/export/trtllm/test_single_device_fp8.py:177
↓ 1 callersMethod_assert_non_quantizable_layers
(self, state_dict)
Megatron-LM/tests/unit_tests/export/trtllm/test_distributed_fp8.py:180
↓ 1 callersFunction_assert_optimizer_tensors_equal
( left: Dict[str, torch.Tensor], right: Dict[str, torch.Tensor], left_empty: Dict[str, torch.Tenso
Megatron-LM/tests/functional_tests/python_test_utils/test_optimizer_grads_match.py:154
↓ 1 callersMethod_assert_quantizable_layers
(self, state_dict, quantized)
Megatron-LM/tests/unit_tests/export/trtllm/test_single_device_fp8.py:165
↓ 1 callersMethod_assert_quantizable_layers
(self, state_dict, quantized)
Megatron-LM/tests/unit_tests/export/trtllm/test_distributed_fp8.py:168
↓ 1 callersMethod_backward
(self, *output_grad)
Megatron-LM/megatron/core/pipeline_parallel/utils.py:248
↓ 1 callersMethod_backward_kv_proj
Computes weight gradients of KV projection layers
Megatron-LM/megatron/core/transformer/multi_latent_attention.py:987
← previousnext →2,501–2,600 of 10,561, ranked by callers