MCPcopy Create free account

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

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

↓ 1 callersFunctionget_config_path
Config copy stored within retro project dir.
Megatron-LM/megatron/core/models/retro/utils.py:8
↓ 1 callersFunctionget_context_parallel_rank
Return caller's rank in the context-parallel group.
Megatron-LM/megatron/core/parallel_state.py:1747
↓ 1 callersMethodget_contrastive_rollouts
(self, request: RolloutRequest)
Megatron-LM/megatron/rl/server/agent/fastapi_env_server.py:104
↓ 1 callersFunctionget_conversation
Get a conversation for a given task and evaluation question.
Megatron-LM/examples/multimodal/run_text_generation.py:536
↓ 1 callersFunctionget_conversations
Extract the input for tokenizer.apply_chat_template.
Megatron-LM/examples/post_training/modelopt/generate.py:60
↓ 1 callersMethodget_converter_model_type
Get converter type: 'GPT' or 'Bert'.
Megatron-LM/tests/functional_tests/test_cases/common/ckpt_converter/__main__.py:150
↓ 1 callersFunctionget_core_transformer_block_key
(model_key)
Megatron-LM/tools/checkpoint/schema_core.py:10
↓ 1 callersFunctionget_corpus_scores
(lines)
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/annotations/filter-selfgeneration.py:24
↓ 1 callersFunctionget_cpu_offload_context
Get CPU offload context and sync function.
Megatron-LM/megatron/core/extensions/transformer_engine.py:2063
↓ 1 callersFunctionget_current_memory_info
()
Megatron-LM/examples/post_training/modelopt/validate.py:71
↓ 1 callersFunctionget_current_node_ip
()
slime/slime/utils/misc.py:58
↓ 1 callersFunctionget_current_running_global_batch_size
Get current running global batch size, taking into account number of DP replicas might be incompatible with true global batch size if `decrease_ba
Megatron-LM/megatron/core/num_microbatches_calculator.py:32
↓ 1 callersFunctionget_cursor
()
gui-rl/desktop_env/server/main.py:275
↓ 1 callersFunctionget_data_loader
Build data loader over data subset. Get a subset of the dataset (from start_idx -> end_idx), and wrap it in a sequential sampler and data loa
Megatron-LM/tools/bert_embedding/embed.py:70
↓ 1 callersMethodget_dataset
(self, validation: bool = False)
Megatron-LM/examples/rl/environments/math/aime_agent.py:18
↓ 1 callersFunctionget_days_since
Calculate number of days since the given datetime string.
Megatron-LM/tests/test_utils/python_scripts/auto_reminder.py:140
↓ 1 callersMethodget_db_chunk_and_continuation_text
Get DB chunk along with continuation, as text.
Megatron-LM/tools/retro/cli/cli.py:130
↓ 1 callersMethodget_db_chunk_bert
Get DB chunk as Bert token ids.
Megatron-LM/tools/retro/cli/cli.py:120
↓ 1 callersMethodget_db_indexed_dataset_infos
Dataset infos, including number of training & sampled sets.
Megatron-LM/tools/retro/cli/cli.py:101
↓ 1 callersMethodget_db_num_chunks
Number of DB chunks.
Megatron-LM/tools/retro/cli/cli.py:110
↓ 1 callersMethodget_db_num_indexed_datasets
Number of indexed datasets within blended dataset.
Megatron-LM/tools/retro/cli/cli.py:96
↓ 1 callersFunctionget_default_causal_mask
Return the causal upper triangular mask for softmax input.
Megatron-LM/megatron/core/transformer/utils.py:33
↓ 1 callersFunctionget_default_save_common_strategy
Get default save common strategy.
Megatron-LM/megatron/core/dist_checkpointing/serialization.py:445
↓ 1 callersFunctionget_default_wandb_args
(test_file: str, run_name_prefix: str | None = None, run_id: str | None = None)
slime/slime/utils/external_utils/command_utils.py:200
↓ 1 callersFunctionget_developer_agent_prompt
Generate the prompt for the Lead Software Engineer agent. Args: current_date (str): The current date. system (str): The opera
terminal-rl/agent/camel_agent.py:39
↓ 1 callersFunctionget_device
(local_rank=None)
Megatron-LM/megatron/training/dist_signal_handler.py:14
↓ 1 callersFunctionget_docker_image_name
Get the image name for a SWEBench/SWE-Gym instance.
swe-rl/swe_utils.py:11
↓ 1 callersFunctionget_dsa_module_spec_for_backend
Helper function to get module spec for Sparse Attention.
Megatron-LM/megatron/core/models/gpt/experimental_attention_variant_module_specs.py:45
↓ 1 callersFunctionget_dynamic_inference_engine
Get the relevant backend for running inference. This function will automatically choose the TRTLLMBackend when possible, and default to Mcore
Megatron-LM/megatron/rl/inference/megatron.py:89
↓ 1 callersMethodget_early_return_outputs
Get the CUDA graph early return outputs for the MoE layer, including the intermediate tensors and the intermediate attributes of the
Megatron-LM/megatron/core/transformer/moe/moe_utils.py:1070
↓ 1 callersMethodget_emb
Forward pass of Yarn Rotary Embedding. Args: max_seq_len (int): Maximum size of sequence offset (int, optional): RoPE
Megatron-LM/megatron/core/models/common/embeddings/yarn_rotary_pos_embedding.py:107
↓ 1 callersMethodget_emb
Forward pass of RoPE embedding before CP sharding. Args: max_seq_len (int): Maximum size of sequence offset (int, opt
Megatron-LM/megatron/core/models/common/embeddings/rotary_pos_embedding.py:151
↓ 1 callersFunctionget_embedding_ranks
Get the embedding ranks.
Megatron-LM/pretrain_gpt.py:284
↓ 1 callersFunctionget_environment
(config: dict, *, default_type: str = "")
swe-rl/mini-swe-agent/src/minisweagent/environments/__init__.py:28
↓ 1 callersFunctionget_environment_rollouts
Sample environment rollouts from an LLM. Args: model: Model to sample from. n_prompts: Number of prompts to sample for across *al
Megatron-LM/megatron/rl/rl_utils.py:589
↓ 1 callersFunctionget_evaluation_dataloader
Build evaluation dataset.
Megatron-LM/examples/multimodal/run_text_generation.py:107
↓ 1 callersFunctionget_evaluation_dataset
Get an evaluation dataset.
Megatron-LM/examples/multimodal/evaluation/evaluation_datasets.py:1408
↓ 1 callersMethodget_expected_keys
Get expected keys from the dataclass fields.
Megatron-LM/megatron/core/extensions/kitchen.py:94
↓ 1 callersMethodget_experience_for_turn
Get experience text for the current turn. In session-experience mode, returns per-session experience (no cross-session). In multi-exp
openclaw-rl/oel/openclaw_oel_api_server.py:516
↓ 1 callersMethodget_experience_stats
Return experience statistics for logging.
openclaw-rl/oel/openclaw_oel_api_server.py:1547
↓ 1 callersFunctionget_experimental_attention_variant_module_spec_for_backend
Helper function to get module spec for Attention
Megatron-LM/megatron/core/models/gpt/experimental_attention_variant_module_specs.py:109
↓ 1 callersFunctionget_expert_data_parallel_rank
Return caller's rank in the expert data parallel group.
Megatron-LM/megatron/core/parallel_state.py:1933
↓ 1 callersFunctionget_expert_model_parallel_world_size
Return world size for the expert-model-parallel group.
Megatron-LM/megatron/core/parallel_state.py:1781
↓ 1 callersFunctionget_expert_tensor_parallel_rank
Return my rank for the expert tensor parallel group.
Megatron-LM/megatron/core/parallel_state.py:1840
↓ 1 callersFunctionget_fa_version_str
()
Megatron-LM/megatron/core/utils.py:379
↓ 1 callersFunctionget_failed_stage
(pipeline_id: int)
Megatron-LM/tests/test_utils/python_scripts/dashboard.py:87
↓ 1 callersFunctionget_feature_bank
()
Megatron-LM/megatron/legacy/model/vision/knn_monitor.py:96
↓ 1 callersFunctionget_file_name
(args, file_id)
Megatron-LM/tools/preprocess_data.py:247
↓ 1 callersFunctionget_filtered_paths
Recursively collect all object paths with exempt decorators from a package. This function traverses the entire package tree and identifies ob
Megatron-LM/scripts/check_api_backwards_compatibility.py:87
↓ 1 callersFunctionget_fp4_align_size
Get the alignment size required for FP4 GEMM. FP4 GEMM requires Blackwell and later architectures. The value 32 is a hardware requiremen
Megatron-LM/megatron/core/fp4_utils.py:51
↓ 1 callersMethodget_fp4_context
Return a new fp4 context in cudagraph mode.
Megatron-LM/megatron/core/transformer/cuda_graphs.py:631
↓ 1 callersMethodget_fp8_context
Return a new fp8 context in cudagraph mode.
Megatron-LM/megatron/core/transformer/cuda_graphs.py:625
↓ 1 callersFunctionget_free_port
(start_port=10000, consecutive=1)
slime/slime/utils/misc.py:65
↓ 1 callersFunctionget_full_tensor_if_necessary
For DTensor gets full tensor if some ranks will not have a local copy
Megatron-LM/megatron/core/utils.py:1011
↓ 1 callersFunctionget_gated_delta_net_module_spec_for_backend
Helper function to get module spec for Linear Attention
Megatron-LM/megatron/core/models/gpt/experimental_attention_variant_module_specs.py:28
↓ 1 callersFunctionget_gitlab_handle
()
Megatron-LM/tests/test_utils/python_scripts/check_status_of_main.py:21
↓ 1 callersMethodget_global_id_seqlens
Calculates the global ID for each subsample. We assign a unique global ID to each subsample. Returns: global_id_seq
Megatron-LM/megatron/core/datasets/data_schedule.py:105
↓ 1 callersFunctionget_global_proxy_pool
Get global proxy pool instance
gui-rl/desktop_env/providers/aws/proxy_pool.py:182
↓ 1 callersMethodget_global_seqlens
Gathers the sequence lengths of all subsamples from all DP ranks. Each DP rank loads the same number of microbatches but each microba
Megatron-LM/megatron/core/datasets/data_schedule.py:55
↓ 1 callersFunctionget_global_worker
(args, data_buffer)
openclaw-rl/openclaw_rollout.py:17
↓ 1 callersFunctionget_global_worker
(args, data_buffer)
openclaw-rl/oel/openclaw_oel_rollout.py:26
↓ 1 callersFunctionget_global_worker
(args, data_buffer)
openclaw-opd/openclaw_opd_rollout.py:17
↓ 1 callersFunctionget_global_worker
(args, data_buffer)
openclaw-combine/openclaw_combine_rollout.py:18
↓ 1 callersFunctionget_global_worker
(args, data_buffer)
openclaw-combine/openclaw_combine_select_rollout.py:32
↓ 1 callersFunctionget_gpt_chunk_datasets
(config)
Megatron-LM/tools/retro/preprocess_data.py:95
↓ 1 callersFunctionget_gpt_mtp_block_spec_for_backend
GPT Multi-Token Prediction (MTP) block spec.
Megatron-LM/megatron/core/models/gpt/gpt_layer_specs.py:809
↓ 1 callersFunctionget_gpt_pipelines
Get GPT (non-MoE) pipelines.
Megatron-LM/tests/functional_tests/test_cases/common/ckpt_converter/__main__.py:833
↓ 1 callersFunctionget_gpt_tokenizer
GPT (BPE) tokenizer.
Megatron-LM/tools/retro/preprocess_data.py:152
↓ 1 callersFunctionget_group_members
(group)
Megatron-LM/tests/test_utils/python_scripts/auto_reminder.py:113
↓ 1 callersFunctionget_group_timestamp
(group_items)
slime/slime_plugins/rollout_buffer/rollout_buffer_example.py:52
↓ 1 callersMethodget_grouped_rollouts
( self, request: GroupedRolloutRequest )
Megatron-LM/megatron/rl/server/agent/fastapi_env_server.py:124
↓ 1 callersMethodget_grouped_rollouts
(self, request: GroupedRolloutRequest)
Megatron-LM/megatron/rl/agent/api.py:180
↓ 1 callersMethodget_groups_and_subsamples
This function recursively forms groups of sub-samples such that all DPxCP ranks have a roughly balanced workload in the group.
Megatron-LM/megatron/core/pipeline_parallel/hybrid_cp_schedule.py:456
↓ 1 callersFunctionget_grpo_returns
( rewards: torch.Tensor, kl: list[torch.Tensor], )
slime/slime/utils/ppo_utils.py:201
↓ 1 callersMethodget_hf_weight_chunks
Mental model of the API: megatron_model.to_hf_magically().named_parameters()
slime/slime/backends/megatron_utils/update_weight/hf_weight_iterator_base.py:24
↓ 1 callersMethodget_history
(self, current_turn_idx: int)
terminal-rl/agent/prm_agent.py:138
↓ 1 callersFunctionget_ict_batch
(data_iterator)
Megatron-LM/megatron/legacy/data/biencoder_dataset_utils.py:54
↓ 1 callersFunctionget_index
Read index from disk. Args: config (RetroPreprocessingConfig): Retro preprocessing config. ondisk (bool): If `ondisk = True`, mem
Megatron-LM/megatron/core/datasets/retro/query/query.py:65
↓ 1 callersMethodget_index_class
Get an index class, given a type string. Args: index_type (str): One of 'faiss-base' (naive Faiss index wrapper) or 'faiss-par-ad
Megatron-LM/megatron/core/datasets/retro/index/factory.py:17
↓ 1 callersFunctionget_individual_chunk_db
Load individual dataset's chunk DB. Args: project_dir (str): Path to Retro project dir. ds_id (int): Index of dataset within blen
Megatron-LM/megatron/core/datasets/retro/db/utils.py:174
↓ 1 callersFunctionget_individual_doc_offsets
Load individual dataset's document offsets. Args: project_dir (str): Path to Retro project dir. ds_id (int): Index of dataset wit
Megatron-LM/megatron/core/datasets/retro/db/utils.py:209
↓ 1 callersFunctionget_inference_engine
Get the relevant backend for running inference This function will automatically choose the TRTLLMBackend when possible, and default to Mcore
Megatron-LM/tools/run_text_generation_server.py:49
↓ 1 callersFunctionget_inference_engine
Utility to get the relevant backend for running inference Args: args (Namespace): The user arguments parsed from command line mod
Megatron-LM/tools/run_inference_performance_test.py:67
↓ 1 callersFunctionget_inference_engine
Utility to get the relevant backend for running inference This function will automatically choose the TRTLLMBackend when possible, and if not rev
Megatron-LM/examples/inference/gpt/gpt_static_inference.py:70
↓ 1 callersFunctionget_inference_engine
Utility to get the relevant backend for running inference This function will automatically chose the TRTLLMBackend when possible, and if not reve
Megatron-LM/examples/inference/t5/simple_t5_batch_inference.py:68
↓ 1 callersFunctionget_inference_interface
(args, loop, model)
Megatron-LM/megatron/rl/rl_utils.py:530
↓ 1 callersFunctionget_input_data
Prepare inputs for the MIMO model forward pass.
Megatron-LM/examples/mimo/avlm_inference.py:32
↓ 1 callersMethodget_input_embeddings
(self)
slime/slime/backends/sglang_utils/qwen3_5.py:114
↓ 1 callersMethodget_input_grads_with_dummy_flags
Get the inputs grads that are returned by the bwd cudagraph call. If using grad accum fusion, wgrads have already been accumulated, so return
Megatron-LM/megatron/core/transformer/cuda_graphs.py:821
↓ 1 callersMethodget_input_ids
Randomly initialize input token IDs.
Megatron-LM/tests/functional_tests/test_cases/common/ckpt_converter/__main__.py:239
↓ 1 callersFunctionget_internvit300M_layer_spec
(use_te)
Megatron-LM/examples/multimodal/nvlm/internvit.py:259
↓ 1 callersFunctionget_internvit_layer_spec
(use_te)
Megatron-LM/examples/multimodal/nvlm/internvit.py:234
↓ 1 callersMethodget_ip_address
Method to get the private IP address of the VM. Private IP means inside the VPC.
gui-rl/desktop_env/providers/base.py:19
↓ 1 callersMethodget_ip_address
(self, path_to_vm: str)
gui-rl/desktop_env/providers/aliyun/provider.py:83
↓ 1 callersFunctionget_key_subresponses
(response)
Megatron-LM/examples/multimodal/evaluation/mmmu_utils.py:273
↓ 1 callersFunctionget_language_model_schema
( prefix: T.Optional[str] = "", layer_prefix: T.Optional[str] = "", use_swiglu=False, )
Megatron-LM/tools/checkpoint/schema_hf.py:173
↓ 1 callersMethodget_layer_id_list
Get the list of layer_id for each layer in the pipeline stage.
Megatron-LM/megatron/core/transformer/pipeline_parallel_layer_layout.py:194
↓ 1 callersFunctionget_layer_maps_from_layer_type_list
Returns maps from global layer index to the corresponding layer index for each layer type in [Attention, Mamba, MLP, MoE] given a layer type
Megatron-LM/megatron/core/ssm/mamba_hybrid_layer_allocation.py:171
↓ 1 callersFunctionget_layer_name_without_prefix
Get TRTLayer name without prefix Given a layer e.g TRTLLMLayers.attention_qkv_weight it returns 'attention.qkv.weight' Args: layer (
Megatron-LM/megatron/core/export/trtllm/trtllm_layers.py:157
↓ 1 callersFunctionget_layer_param
(args, name, param)
slime/tools/convert_torch_dist_to_hf.py:83
← previousnext →4,101–4,200 of 10,561, ranked by callers