MCPcopy Create free account

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

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

Methodpad
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:283
Methodpad
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:514
Methodpad
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:757
Methodpad
Pad token ID.
Megatron-LM/megatron/training/tokenizer/multimodal_tokenizer.py:318
Functionpad_and_convert_to_numpy
Pad sequences and convert them to numpy.
Megatron-LM/megatron/legacy/data/dataset_utils.py:378
Methodpad_func
(experts, pad)
slime/slime/backends/megatron_utils/actor.py:533
Methodpad_id
Get the padding ID.
Megatron-LM/megatron/core/tokenizers/text/libraries/bytelevel_tokenizer.py:132
Methodpad_id
Returns id of padding token.
Megatron-LM/megatron/core/tokenizers/text/libraries/huggingface_tokenizer.py:279
Methodpad_id
Returns id of padding token.
Megatron-LM/megatron/core/tokenizers/text/libraries/sentencepiece_tokenizer.py:314
Methodpad_id
Property alias to match MegatronTokenizer; returns pad_id if available.
Megatron-LM/megatron/core/tokenizers/text/libraries/abstract_tokenizer.py:110
Methodpad_id
Returns id of padding token.
Megatron-LM/megatron/core/tokenizers/text/libraries/tiktoken_tokenizer.py:266
Methodpad_weight
(orig_word_embed, true_vocab_size)
Megatron-LM/tools/checkpoint/saver_base.py:374
Methodparam_copy_back_gpu_hook
(optimizer, args, kwargs)
Megatron-LM/megatron/core/optimizer/cpu_offloading/hybrid_optimizer.py:119
Methodparam_copy_back_gpu_hook_closure
()
Megatron-LM/megatron/core/optimizer/cpu_offloading/hybrid_optimizer.py:118
Methodparam_groups
Get param_groups aggregated over underlying optimizers.
Megatron-LM/megatron/core/optimizer/optimizer.py:1118
Methodparams
Return a list of information on function params.
Megatron-LM/tools/retro/config_utils.py:176
Methodparse_coordinate
(raw_coord)
gui-rl/agents/qwen35_agent.py:523
Methodparse_error_count
(self)
terminal-rl/rollout_agent.py:13
Methodparse_keys
(raw_keys)
gui-rl/agents/qwen35_agent.py:513
Methodparse_response
( self, response: str, original_width: int, original_height: int, proc
gui-rl/agents/qwen3vl_agent.py:378
Methodparse_response
(self, responses: list[InferenceResponse])
Megatron-LM/megatron/rl/inference/chat_templates.py:48
Methodparse_xml_tool_call
(xml_content: str)
gui-rl/agents/qwen35_agent.py:494
Functionpart_filter
(part: str)
Megatron-LM/examples/mimo/data/avlm_sample_loader.py:85
Methodpartial_weight_for_target
(target: str)
Megatron-LM/megatron/core/pipeline_parallel/multimodule_communicator.py:485
Functionpatch_hadamard_if_needed
Automatically patch hadamard_transform in dsa module if not installed.
Megatron-LM/tests/unit_tests/transformer/test_attention_variant_dsa.py:44
Methodpause_generation
(self)
slime/slime/backends/sglang_utils/sglang_engine.py:416
Functionpipeline
( compare_approximate_results: bool, golden_values: Dict[str, GoldenValueMetric], actual_values: D
Megatron-LM/tests/functional_tests/python_test_utils/common.py:178
Functionpoint_transform
(x)
gui-rl/desktop_env/evaluators/metrics/gimp.py:167
Functionpolicy_loss_function
Compute policy loss (PPO/GSPO) and metrics. Computes current log-probabilities and entropy from model logits, then calculates PPO-style clipp
slime/slime/backends/megatron_utils/loss.py:845
Functionpop_first
(args, rollout_id, buffer: list[list[Sample]], num_samples: int)
slime/slime/rollout/data_source.py:214
Functionport
(consecutive=1)
slime/slime/ray/rollout.py:1011
Methodpost_load_state_dict_hook
(self)
Megatron-LM/megatron/core/optimizer/cpu_offloading/hybrid_optimizer.py:417
Functionpost_process_rewards
(args, samples: list[Sample], **kwargs)
slime/examples/on_policy_distillation/on_policy_distillation.py:35
Methodpost_process_weights
Update model weights from tensor data. The HTTP server will only post meta data, and the real weights will be copied directly from GPUs.
slime/slime/backends/sglang_utils/sglang_engine.py:426
Methodpostprocess
(self, model_outputs)
Megatron-LM/tools/bert_embedding/huggingface.py:53
Functionpostprocess_openclaw_prm_teacher
(args: Namespace, rollout_data: RolloutBatch)
openclaw-combine/prm_teacher_postprocess.py:116
Functionpp_post_backward
(input_tensor_grad, vp_stage=None)
Megatron-LM/megatron/core/pipeline_parallel/schedules.py:1694
Functionpp_post_forward
(output_tensor, vp_stage=None)
Megatron-LM/megatron/core/pipeline_parallel/schedules.py:1623
Functionpp_pre_backward
(vp_stage=None)
Megatron-LM/megatron/core/pipeline_parallel/schedules.py:1676
Functionpp_pre_forward
(vp_stage=None)
Megatron-LM/megatron/core/pipeline_parallel/schedules.py:1604
Methodpp_size
(name: str)
Megatron-LM/megatron/core/pipeline_parallel/multimodule_communicator.py:480
Functionpre_forward_hook
(*args, **kwargs)
slime/slime/utils/routing_replay.py:89
Methodpre_load_state_dict_hook
Pre-load state dictionary hook to prevent loss of precision in mixed-precision training. When loading a state di
Megatron-LM/megatron/core/optimizer/cpu_offloading/hybrid_optimizer.py:384
Methodprefix_cache_hit_rate
(self)
slime/slime/utils/types.py:95
Methodpreload_tensors
Preloads tensors in `state_dict` to host memory via CPU memory. Args: write_buckets (List): List of `WriteBucket` object
Megatron-LM/megatron/core/dist_checkpointing/strategies/filesystem_async.py:212
Methodpreload_tensors
(*args, **kwargs)
slime/slime/utils/rocm_checkpoint_writer.py:14
Methodprep_inference_input
Prepares the inference input data. Args: prompts_tokens (torch.Tensor): A tensor of shape [batch_size, max_seq_len] Retu
Megatron-LM/megatron/core/inference/model_inference_wrappers/abstract_model_inference_wrapper.py:129
Methodprep_inference_input
Prepares the inference input data. Args: prompts_tokens (torch.Tensor): A tensor of shape [batch_size, max_seq_len] Retu
Megatron-LM/megatron/core/inference/model_inference_wrappers/gpt/gpt_inference_wrapper.py:41
Methodprep_inference_input
Prepares the inference input data. Args: prompts_tokens (torch.Tensor): A tensor of shape [batch_size, max_seq_len] n
Megatron-LM/megatron/core/inference/model_inference_wrappers/multimodal/vlm_inference_wrapper.py:55
Methodprep_inference_input
Preparing input data for inference, using respective wrapper's prep_inference_input method # pylint: disable=line-too-long Currently only sup
Megatron-LM/megatron/core/inference/text_generation_controllers/vlm_text_generation_controller.py:16
Methodprep_inference_input
Preparing input data for inference, using respective wrapper's prep_inference_input method # pylint: disable=line-too-long Args:
Megatron-LM/megatron/core/inference/text_generation_controllers/encoder_decoder_text_generation_controller.py:21
Methodprep_model_for_inference
A utility function for preparing model for inference The function gets called once before the auto regressive inference loop. It puts
Megatron-LM/megatron/core/inference/model_inference_wrappers/abstract_model_inference_wrapper.py:104
Methodprepare_grads
Pre-processing gradients before the optimizer step, returns whether inf/nan is found.
Megatron-LM/megatron/core/optimizer/optimizer.py:185
Methodprepare_request
( self, prompts: list[str | list[LLMChatMessage]], generation_args: GenericGenerationArgs )
Megatron-LM/megatron/rl/inference/inference_interface.py:103
Functionprepare_task_docker_image
( task: dict[str, Any], timeout: float = 1200.0, )
terminal-rl/remote/docker_compose_utils.py:195
Functionpreprocess_common_state_dict
(common_state_dict)
Megatron-LM/megatron/training/training.py:582
Methodpreprocess_fn
(x)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_serialization.py:117
Methodpreprocess_fn
(optim_common_dict)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_optimizer.py:890
Methodpress_modifier_keys
()
gui-rl/agents/qwen35_agent.py:542
Functionpretrain_ict_model_provider
(pre_process=True, post_process=True)
Megatron-LM/pretrain_ict.py:24
Functionprint_error_handler
(exc: Exception, key: Optional[str])
Megatron-LM/examples/multimodal/dataset_helpers.py:945
Methodprint_neighbor_texts
(cls, sample_id: int, chunk_id: int, data_key: str="train")
Megatron-LM/tools/retro/cli/cli.py:193
Functionprint_params_min_max_norm
Print min, max, and norm of all parameters.
Megatron-LM/megatron/training/utils.py:293
Functionprint_tensor
Print tensor metadata from a Megatron Core Distributed Checkpoint
Megatron-LM/tools/checkpoint/checkpoint_inspector.py:190
Functionprint_torch_dcp_in_json
(torch_dcp_dir, model_weight_prefix="model.module")
Megatron-LM/tools/checkpoint/checkpoint_inspector.py:946
Functionprintable_text
Returns text encoded in a way suitable for print or `tf.logging`.
Megatron-LM/megatron/training/tokenizer/bert_tokenization.py:98
Methodprocess_arg
(arg)
Megatron-LM/megatron/core/transformer/cuda_graphs.py:991
Functionprocess_file
(input_path, output_path, filename, strategy, block_size, result_collector)
slime/tools/convert_hf_to_fp8.py:116
Functionprocess_file
(input_path, output_path, filename, group_size, is_symmetric, ignore_rules, result_collector)
slime/tools/convert_hf_to_int4_direct.py:167
Functionprocess_instance
Process a single SWEBench instance.
swe-rl/mini-swe-agent/src/minisweagent/run/extra/swebench.py:122
Methodprocess_json_file
(self, file_name)
Megatron-LM/tools/preprocess_data.py:152
Functionprocess_non_loss_data
(data, iteration, writer)
Megatron-LM/pretrain_vision_inpaint.py:99
Methodprocess_tool_call
(json_str: str)
gui-rl/agents/qwen3vl_agent.py:407
Methodprocess_tool_call_params
(params: Dict)
gui-rl/agents/qwen35_agent.py:533
Functionprocess_typewrite_match
(match)
gui-rl/desktop_env/desktop_env.py:55
Methodproject_embeddings
Project embeddings into a tensor. Args: embeddings (List[torch.Tensor]): List of embeddings to project
Megatron-LM/megatron/core/models/mimo/submodules/base.py:165
Methodproject_embeddings
( self, embeddings: List[torch.Tensor], is_input: bool = True )
Megatron-LM/tests/unit_tests/models/test_mimo_submodules.py:37
Methodprovide
(self, pre_process=None, post_process=None, vp_stage=None)
slime/slime_plugins/megatron_bridge/qwen35_vl.py:901
Methodprovide_language_model
(self, pre_process=None, post_process=None, vp_stage=None)
slime/slime_plugins/megatron_bridge/qwen35_vl.py:921
Methodprovider_bridge
Create a GPT ModelProvider from Qwen3.5 HF config.
slime/slime_plugins/megatron_bridge/qwen3_5_text.py:42
Methodproxy
Proxy all other requests to the SGLang router
slime/slime/router/router.py:131
Methodpurge_record_files
(self)
openclaw-fireworks/fireworks_server.py:875
Methodput
A patched version of put, implementing PR 3544 and PR 4295.
Megatron-LM/megatron/core/ssm/triton_cache_manager.py:51
Functionpytest_addoption
Additional command-line arguments passed to pytest.
Megatron-LM/tests/functional_tests/python_test_utils/conftest.py:6
Functionpytest_addoption
Additional command-line arguments passed to pytest. For now: --experimental: Enable the Mcore experimental flag (DEFAULT: False)
Megatron-LM/tests/unit_tests/conftest.py:17
Functionpytest_sessionfinish
(session, exitstatus)
Megatron-LM/tests/unit_tests/conftest.py:36
Functionpytest_sessionfinish
(session, exitstatus)
Megatron-LM/tests/unit_tests/dist_checkpointing/conftest.py:9
Functionpytest_sessionfinish
(session, exitstatus)
Megatron-LM/tests/unit_tests/transformer/moe/conftest.py:13
Methodqkv_up_proj_and_rope_apply
Apply the up projection and RoPE to the query and key. When sequence packing enabled, the input tensors adopt a packed shape
Megatron-LM/megatron/core/transformer/multi_latent_attention.py:746
Methodqkv_up_proj_and_rope_apply_for_cached_latent_kv
( q_compressed, kv_compressed, k_pos_emb, rotary_pos_emb )
Megatron-LM/megatron/core/transformer/multi_latent_attention.py:678
Methodquery
(self, messages: list[dict[str, str]], **kwargs)
swe-rl/mini-swe-agent/src/minisweagent/models/litellm_model.py:62
Methodquery
(self, *args, **kwargs)
swe-rl/mini-swe-agent/src/minisweagent/models/extra/roulette.py:36
Methodquery
(self)
swe-rl/mini-swe-agent/src/minisweagent/agents/interactive.py:55
Functionquery_single_turn
(client, messages, sampling_params, tools=None)
slime/slime_plugins/rollout_buffer/generator/base_generator.py:37
Functionquick_gelu
Quick GELU activation
Megatron-LM/megatron/core/activations.py:15
Functionraise_not_implemented
Raise NotImplementedError for Dense layer.
Megatron-LM/megatron/core/models/gpt/fine_grained_callables.py:491
Methodrank
(self)
slime/slime/utils/reloadable_process_group.py:160
Functionray_noset_visible_devices
(env_vars=os.environ)
slime/slime/ray/utils.py:28
Methodread
Read bytes into a numpy array. Args: dtype (Type[numpy.number]): Data-type of the returned array. count (int): Numbe
Megatron-LM/megatron/core/datasets/indexed_dataset.py:389
Methodread
Read bytes into a numpy array. Args: dtype (Type[numpy.number]): Data-type of the returned array. count (int): Numbe
Megatron-LM/megatron/core/datasets/indexed_dataset.py:425
← previousnext →8,401–8,500 of 10,561, ranked by callers