MCPcopy Create free account

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

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

Functioncreate_nccl_mem_pool
Create a memory pool using the NCCL allocator.
Megatron-LM/megatron/core/nccl_allocator.py:109
Methodcreate_node
(stream, module, name)
Megatron-LM/megatron/core/models/common/model_chunk_schedule_plan.py:155
Methodcreate_tensor_recv_next
()
Megatron-LM/megatron/core/pipeline_parallel/p2p_communication.py:317
Methodcreate_tensor_recv_prev
()
Megatron-LM/megatron/core/pipeline_parallel/p2p_communication.py:309
Functioncreate_tokens_and_tokentypes
Merge segments A and B, add [CLS] and [SEP] and build tokentypes.
Megatron-LM/megatron/legacy/data/dataset_utils.py:136
Functioncross_entropy_loss_func
(labels, output_tensor)
Megatron-LM/tasks/finetune_utils.py:40
Methodcurrent_trajectory_name
Get the name of the current trajectory file.
swe-rl/mini-swe-agent/src/minisweagent/run/inspector.py:120
Methodcustom
(start, end)
Megatron-LM/megatron/legacy/model/transformer.py:1600
Methodcustom
(start: int, end: int)
Megatron-LM/megatron/core/transformer/transformer_block.py:435
Methodcustom_forward
(*inputs)
Megatron-LM/megatron/legacy/model/transformer.py:629
Methodcustom_forward
(*args, **kwargs)
Megatron-LM/megatron/legacy/model/transformer.py:1601
Methodcustom_forward
(*inputs)
Megatron-LM/megatron/core/transformer/attention.py:273
Methodcustom_forward
( hidden_states, attention_mask, context, context_mask, rotary_pos_emb )
Megatron-LM/megatron/core/transformer/transformer_block.py:436
Methodcustom_forward
(hidden_states)
Megatron-LM/megatron/core/transformer/moe/moe_layer.py:295
Methodcustom_storage
()
Megatron-LM/tests/unit_tests/fusions/test_fused_linear_cross_entropy.py:517
Methodcustom_storage
()
Megatron-LM/tests/unit_tests/fusions/test_fused_linear_cross_entropy.py:949
Methodcustom_storage
()
Megatron-LM/tests/unit_tests/fusions/test_fused_linear_cross_entropy.py:1460
Methodd2h_stream
Get the device-to-host (GPU to CPU) transfer stream.
Megatron-LM/megatron/core/pipeline_parallel/fine_grained_activation_offload.py:54
Functiondataclass_cli
Modified from https://github.com/fastapi/typer/issues/154#issuecomment-1544876144
slime/slime/utils/external_utils/typer_utils.py:8
Functiondebug_msg
Logs a debug message using the current logger stack. This function formats and logs a debug message with the current logger and name stack, p
Megatron-LM/megatron/core/dist_checkpointing/utils.py:318
Methoddecode
Decode embeddings into image tensors. Args: embeddings: Tensor of embeddings to decode. data_batch: Dictionary contai
Megatron-LM/megatron/core/models/mimo/submodules/vision.py:99
Methoddecode
Decode embeddings into a tensor. Args: embeddings (torch.Tensor): Embeddings to decode data_batch (Di
Megatron-LM/megatron/core/models/mimo/submodules/base.py:150
Methoddecode
Decode embeddings into audio data.
Megatron-LM/megatron/core/models/mimo/submodules/audio.py:90
Methoddecode
(self, embeddings: torch.Tensor, data_batch: Dict)
Megatron-LM/tests/unit_tests/models/test_mimo_submodules.py:34
Methoddecode_token_ids
(self, token_ids)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:260
Methoddecoder
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:160
Methoddecoder
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:454
Functiondecorator
(func: Callable)
Megatron-LM/megatron/core/utils.py:2221
Functiondecorator
(func)
Megatron-LM/megatron/core/transformer/moe/moe_utils.py:1221
Functiondecorator
(fn)
slime/slime/utils/timer.py:93
Functiondecrease_brightness
(image1_path: str, image2_path: str)
gui-rl/desktop_env/evaluators/metrics/gimp.py:85
Methoddefault
(self, o)
Megatron-LM/megatron/core/config_logger.py:64
Functiondefault_cache_dir
Provides a default path for the Triton cache directory.
Megatron-LM/megatron/core/ssm/triton_cache_manager.py:26
Functiondefault_embedding_ranks
Return the default ranks that constitute the stages on which the word embeddings live. For most models, these are the first and last pipeline stag
Megatron-LM/megatron/core/parallel_state.py:520
Functiondefault_get_group_data_meta_info
Default implementation for getting meta information about the temporary data collected between get_batch calls.
slime/slime_plugins/rollout_buffer/buffer.py:22
Functiondefault_is_valid_group
(group_data, min_valid_group_size, task_type)
slime/slime_plugins/rollout_buffer/buffer.py:17
Functiondefault_loader
(path: str)
Megatron-LM/megatron/legacy/data/image_folder.py:254
Functiondefault_position_embedding_ranks
Return the default ranks that constitute the stages on which the position embeddings live. For most models, this is only the first pipeline stage.
Megatron-LM/megatron/core/parallel_state.py:529
Functiondelete_checkpoint
(args, iteration_to_delete)
Megatron-LM/megatron/training/checkpointing.py:672
Methoddelete_session
(session_id: str, request: Request)
openclaw-rl/oel/openclaw_oel_api_server.py:580
Methoddelete_vm
(self, vm_path, region=DEFAULT_REGION, lock_needed=True, **kwargs)
gui-rl/desktop_env/providers/aws/manager.py:208
Methoddelete_vm
(self, vm_path, region=None, **kwargs)
gui-rl/desktop_env/providers/docker/manager.py:102
Methoddelete_vm
(self, vm_path, lock_needed=True)
gui-rl/desktop_env/providers/virtualbox/manager.py:356
Methoddelete_vm
(self, vm_path, lock_needed=True, **kwargs)
gui-rl/desktop_env/providers/volcengine/manager.py:305
Methoddelete_vm
(self, vm_path, lock_needed=True)
gui-rl/desktop_env/providers/vmware/manager.py:344
Methoddelete_vm
(self, vm_path, lock_needed=True, **kwargs)
gui-rl/desktop_env/providers/aliyun/manager.py:288
Functiondeprecated
Mark a function as deprecated. This decorator: 1. Adds deprecation metadata to the function 2. Issues a DeprecationWarning when the
Megatron-LM/megatron/core/utils.py:2383
Methoddeprecation
Return a single information on function deprecation notes.
Megatron-LM/tools/retro/config_utils.py:208
Functiondequantize_fp4_tensor
Dequantize a fp4 tensor to a higher precision tensor.
Megatron-LM/megatron/core/fp4_utils.py:73
Methoddeserialize
Deserialize event. Args: obj (dict): Serialized event data. Returns: (DynamicInferenceEvent) Deserialized ev
Megatron-LM/megatron/core/inference/inference_request.py:209
Methoddeserialize
Deserialize request. Args: obj (dict): Serialized request data. Returns: (DynamicInferenceRequest) Deseriali
Megatron-LM/megatron/core/inference/inference_request.py:285
Methoddeserialize
Deserialize record. Args: obj (dict): Serialized record data. Returns: (DynamicInferenceRequestRecord) Deser
Megatron-LM/megatron/core/inference/inference_request.py:508
Methoddestroy
(self)
Megatron-LM/tests/unit_tests/transformer/moe/test_token_dispatcher.py:350
Functiondestroy_global_state
()
Megatron-LM/megatron/training/training.py:153
Functiondestroy_state
()
Megatron-LM/megatron/training/inprocess_restart.py:25
Methoddestroy_weights_update_group
(self, group_name)
slime/slime/backends/sglang_utils/sglang_engine.py:387
Methoddeterministic_empty
(*args, **kwargs)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_local.py:172
Methoddeterministic_empty
(*args, **kwargs)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_nonpersistent.py:44
Methoddetokenize
Text detokenization. Args: ids (list): text to be tokenized. Returns: text: dettokenized text.
Megatron-LM/megatron/core/tokenizers/text/text_tokenizer.py:77
Methoddetokenize
(self, token_ids, **kwargs)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:166
Methoddetokenize
Copy of decode() method for inference pipeline compatibility
Megatron-LM/megatron/training/tokenizer/tokenizer.py:256
Methoddetokenize
(self, token_ids)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:356
Methoddetokenize
(self, ids)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:489
Methoddetokenize
(self, ids)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:554
Methoddetokenize
(self, ids)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:608
Methoddetokenize
(self, tokens: List[int])
Megatron-LM/megatron/training/tokenizer/tokenizer.py:785
Methoddetokenize
(self, ids)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:894
Methoddetokenize
Detokenize tokens.
Megatron-LM/megatron/training/tokenizer/sft_tokenizer.py:155
Methoddfs
(node: str)
Megatron-LM/megatron/core/pipeline_parallel/multimodule_communicator.py:506
Functiondict_map
`map` equivalent for dicts.
Megatron-LM/megatron/core/dist_checkpointing/dict_utils.py:186
Functiondict_map_with_key
`map` equivalent for dicts with a function that accepts tuple (key, value).
Megatron-LM/megatron/core/dist_checkpointing/dict_utils.py:192
Functiondiff
()
swe-rl/server/swe_env_pool_server.py:287
Functiondirect_load_json
(fp)
gui-rl/desktop_env/evaluators/metrics/vscode.py:19
Methoddispatch
Dispatch the hidden_states according to the routing_map.
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:925
Methoddispatch
( self, hidden_states: torch.Tensor, async_finish: bool = True, allocate_on_co
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:1025
Methoddispatch
(self, request: Request, call_next)
slime/slime/router/middleware_hub/radix_tree_middleware.py:68
Methoddispatch_postprocess
After gathering in token_dispatch, this method identifies tokens for local experts and permutes them for expert processing.
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:282
Methoddispatch_postprocess
Post-processes tokens after All-to-All communication. This involves an All-Gather in the tensor parallel dimension and sorting tokens
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:678
Methoddispatch_postprocess
Converts dispatched tokens to a per-expert format for expert processing. This method transforms the output of the fused dispatch into the ten
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:1462
Methoddispatch_preprocess
Reshapes hidden states and caches the routing map.
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:248
Methoddispatch_preprocess
Prepares hidden states and probabilities for dispatch. This method reshapes the hidden states, computes communication metadata, and p
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:595
Methoddispatch_preprocess
Initializes routing metadata and prepares tensors for fused dispatch. This method reshapes input tensors and processes routing information in
Megatron-LM/megatron/core/transformer/moe/token_dispatcher.py:1407
Methoddispose
(self)
slime/slime/utils/metric_utils.py:146
Methoddispose
(self)
slime/slime/ray/rollout.py:123
Functiondistributed_context
()
Megatron-LM/tests/unit_tests/fusions/test_fused_linear_cross_entropy.py:45
Methoddo_post
(self, url, payload, max_retries=60)
slime/slime/utils/http_utils.py:296
Methoddocument_indices
Get the document indices Returns: numpy.ndarray: The document indices
Megatron-LM/megatron/core/datasets/indexed_dataset.py:773
Functiondownload_file
()
gui-rl/desktop_env/server/main.py:1229
Methoddownload_file
(self, Bucket: str, Key: str, Filename: str)
Megatron-LM/tests/unit_tests/data/test_bin_reader.py:73
Functiondownload_seta_env
()
terminal-rl/data_utils/download.py:61
Functiondownload_tbench_adapted
()
terminal-rl/data_utils/download.py:89
Functiondownload_tbench_core
()
terminal-rl/data_utils/download.py:69
Functiondownload_tbench_test
()
terminal-rl/data_utils/download.py:77
Methoddrain_eval_scores
(self)
openclaw-fireworks/fireworks_server.py:865
Methoddrain_eval_scores
(self)
openclaw-tinker/api_server.py:339
Functiondummy_range
(msg)
Megatron-LM/megatron/training/utils.py:769
Functiondummy_step_func
(data_iterator, model)
Megatron-LM/tests/unit_tests/pipeline_parallel/test_schedules.py:475
Methodeffective_response_length
(self)
slime/slime/utils/types.py:146
Methodelapsed
Calculates the elapsed time and restarts timer. Args: reset (bool, optional): Resets timer before restarting. Defaults to True.
Megatron-LM/megatron/core/timers.py:65
Methodelapsed
(self, reset=True, barrier=False)
Megatron-LM/megatron/core/timers.py:93
← previousnext →7,101–7,200 of 10,561, ranked by callers