Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/Gen-Verse/OpenClaw-RL
/ functions
Functions
10,561 in github.com/Gen-Verse/OpenClaw-RL
⨍
Functions
10,561
◇
Types & classes
1,405
↳
Endpoints
144
Method
log_failure
(message: str)
Megatron-LM/megatron/core/rerun_state_machine.py:535
Function
log_perf_data_raw
( rollout_id: int, args: Namespace, is_primary_rank: bool, compute_total_fwd_flops: Callable )
slime/slime/utils/train_metric_utils.py:13
Method
longest_path_to
(target: str)
Megatron-LM/megatron/core/pipeline_parallel/multimodule_communicator.py:497
Function
lookup_user
()
Megatron-LM/tests/test_utils/python_scripts/auto_reminder.py:217
Method
loss_fn
(output, _)
Megatron-LM/tests/unit_tests/distributed/test_mcore_fully_sharded_data_parallel.py:188
Function
loss_func
(loss_mask, sentence_order, output_tensor)
Megatron-LM/pretrain_bert.py:99
Function
loss_func
Loss function. Args: loss_mask (torch.Tensor): Used to mask out some portions of the loss output_tensor (torch.Tensor): The tenso
Megatron-LM/pretrain_gpt.py:82
Function
loss_func
(output_tensor)
Megatron-LM/pretrain_ict.py:73
Function
loss_func
Loss function. Args: loss_mask (torch.Tensor): Used to mask out some portions of the loss output_tensor (torch.Tensor): The tenso
Megatron-LM/pretrain_mamba.py:64
Function
loss_func
Loss function. Args: loss_mask (torch.Tensor): Used to mask out some portions of the loss kl_term (torch.Tensor): KL term of the
Megatron-LM/train_rl.py:64
Function
loss_func
(model, labels, output_tensor, collect_data=False)
Megatron-LM/pretrain_vision_dino.py:37
Function
loss_func
(images, masks, masked_images, outputs, non_loss_data=False)
Megatron-LM/pretrain_vision_inpaint.py:46
Function
loss_func
(labels, output_tensor)
Megatron-LM/pretrain_vision_classify.py:51
Function
loss_func
(loss_mask, output_tensor)
Megatron-LM/tools/retro/sft/sft_retro.py:153
Function
loss_func
(output_tensor: torch.Tensor)
Megatron-LM/tests/unit_tests/pipeline_parallel/test_pipeline_layout.py:243
Function
loss_func
(output_tensor)
Megatron-LM/tests/unit_tests/pipeline_parallel/test_schedules.py:144
Function
loss_func
(output_tensor)
Megatron-LM/tests/unit_tests/transformer/test_full_cuda_graph.py:37
Function
loss_func
(loss_mask: torch.Tensor, output_tensor: torch.Tensor)
Megatron-LM/tests/unit_tests/export/trtllm/test_distributed_fp8.py:87
Function
loss_func
( loss_mask: torch.Tensor, output_tensor: torch.Tensor )
Megatron-LM/examples/run_simple_mcore_train_loop.py:129
Function
loss_func
Simple loss function for MIMO model training. Args: loss_mask: mask indicating which tokens contribute to the loss output_tensor:
Megatron-LM/examples/mimo/train.py:140
Function
loss_func
(loss_mask, output_tensor)
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/finetune_gpt.py:72
Function
loss_func
(loss_mask, output_tensor)
Megatron-LM/examples/multimodal/train.py:241
Function
loss_function
Dispatch to the configured loss and rescale for Megatron integration. Selects one of "policy_loss", "value_loss", "sft_loss", or a custom loss
slime/slime/backends/megatron_utils/loss.py:1187
Function
lru_cache
()
Megatron-LM/megatron/training/tokenizer/gpt2_tokenization.py:34
Function
main
Minimally send prompts to SGLang using OpenAI endpoints with arguments in the same format as main Slime. Example usage: python -m slime.
slime/slime/utils/debug_utils/send_to_sglang.py:12
Function
main
( # Deliberately make this name consistent with main training arguments load_debug_rollout_data: Annot
slime/slime/utils/debug_utils/display_debug_rollout_data.py:25
Function
main
( rollout_data_path: Annotated[str, typer.Option()], custom_rm_path: Annotated[str, typer.Option()], )
slime/slime/utils/debug_utils/replay_reward_fn.py:22
Function
main
(args: DemoArgs)
slime/slime/utils/external_utils/typer_utils.py:56
Method
main_grad_getter
(p)
Megatron-LM/megatron/core/distributed/fsdp/src/megatron_fsdp/param_and_grad_buffer.py:2314
Method
make_needed_groups
(param_group)
Megatron-LM/megatron/core/optimizer/distrib_optimizer.py:736
Method
mamba_state_shapes_per_request
Returns the Mamba conv and ssm states shapes per request.
Megatron-LM/megatron/core/ssm/mamba_layer.py:87
Function
manager
Create a basic RunBatchProgressManager for testing.
swe-rl/mini-swe-agent/tests/run/test_batch_progress.py:8
Function
manager_with_yaml
Create a RunBatchProgressManager with yaml reporting.
swe-rl/mini-swe-agent/tests/run/test_batch_progress.py:14
Method
many_returns
Return a list of information on function return.
Megatron-LM/tools/retro/config_utils.py:201
Method
mapping_registry
Weight mappings from HF Qwen3.5 to Megatron format.
slime/slime_plugins/megatron_bridge/qwen3_5_text.py:83
Method
mark_bucket_ready_to_use
()
Megatron-LM/megatron/core/distributed/fsdp/src/megatron_fsdp/param_and_grad_buffer.py:3619
Method
mask
Returns id of of mask token.
Megatron-LM/megatron/core/tokenizers/text/text_tokenizer.py:216
Method
mask
Returns mask token.
Megatron-LM/megatron/core/tokenizers/text/libraries/null_tokenizer.py:75
Method
mask
The MASK token id Raises: NotImplementedError: Non-abstract, optional attribute
Megatron-LM/megatron/core/datasets/megatron_tokenizer.py:156
Method
mask
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:287
Method
mask
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:530
Method
mask
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:566
Method
mask
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:620
Method
mask
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:859
Method
mask
(self)
Megatron-LM/megatron/training/tokenizer/tokenizer.py:932
Method
mask_id
Returns id of of mask token. Need for NeMo.
Megatron-LM/megatron/core/tokenizers/text/text_tokenizer.py:221
Method
mask_id
Returns id of mask token.
Megatron-LM/megatron/core/tokenizers/text/libraries/huggingface_tokenizer.py:326
Method
mask_id
Returns id of mask token.
Megatron-LM/megatron/core/tokenizers/text/libraries/sentencepiece_tokenizer.py:363
Method
mask_id
Property alias to match MegatronTokenizer; returns mask_id if available.
Megatron-LM/megatron/core/tokenizers/text/libraries/abstract_tokenizer.py:143
Method
mask_id
Returns id of mask token.
Megatron-LM/megatron/core/tokenizers/text/libraries/tiktoken_tokenizer.py:261
Method
match
Pattern based match.
Megatron-LM/megatron/core/quantization/quant_config.py:119
Method
match
Gives an operator's context, return a QuantizationConfig if necessary, or sentinel (None) denoting no matchers matched.
Megatron-LM/megatron/core/quantization/quant_config.py:202
Function
match_in_list
(result, rules)
gui-rl/desktop_env/evaluators/metrics/general.py:50
Function
maybe_dump_bridge_runtime_layout
(bridge, megatron_model)
slime/slime/utils/megatron_bridge_utils.py:29
Function
maybe_force_nccl_backend_init
(device_id)
Megatron-LM/megatron/training/inprocess_restart.py:148
Function
maybe_reformulate_nd_flattened_tensor
(sh_ten: Any)
Megatron-LM/megatron/core/dist_checkpointing/strategies/resharding.py:125
Function
maybe_setup_simulated_fault
Sets a simulated fault, based on `FT_SIM_FAULT_DESC` env variable. Simulated fault description format: rank_hung|rank_killed;rank_to_fail|"";b
Megatron-LM/megatron/training/ft_integration.py:298
Function
maybe_skip_or_early_return_by_cudagraph
Decorator to skip certain codepaths in the MoE layer forward pass in CUDA graph replay, or early return from the MoE layer forward pass in CU
Megatron-LM/megatron/core/transformer/moe/moe_utils.py:1183
Function
maybe_wrap_for_inprocess_restart
(pretrain)
Megatron-LM/megatron/training/inprocess_restart.py:126
Function
mcore_gpt_load_te_state_dict_pre_hook
Register a pre-hook to fix the state_dict key difference of. This prehook is used when trying to load the megatron/core GPTModel that uses a
Megatron-LM/megatron/core/post_training/modelopt/gpt/state_dict_hooks.py:10
Function
megatron_fsdp_optimizer_step
(optimizer, *args, **kwargs)
Megatron-LM/megatron/core/distributed/fsdp/src/megatron_fsdp/fully_shard.py:269
Function
megatron_fsdp_optimizer_zero_grad
(optimizer, *args, **kwargs)
Megatron-LM/megatron/core/distributed/fsdp/src/megatron_fsdp/fully_shard.py:293
Method
merge
(self, other)
slime/slime/utils/seqlen_balancing.py:64
Method
merge_extra_states
Merge multiple "_extra_state" into one.
Megatron-LM/megatron/core/extensions/kitchen.py:621
Method
merge_extra_states
Merge multiple "_extra_state" into one.
Megatron-LM/megatron/core/extensions/transformer_engine.py:1315
Method
merge_lists
(key)
Megatron-LM/megatron/core/inference/inference_request.py:466
Method
merges_file
Returns merges file path if specified.
Megatron-LM/megatron/core/tokenizers/text/text_tokenizer.py:251
Method
messages
(self)
swe-rl/mini-swe-agent/src/minisweagent/agents/interactive_textual.py:318
Method
metrics
(self)
Megatron-LM/megatron/rl/agent/pass_at_evaluation_agent.py:37
Method
metrics
(self)
Megatron-LM/megatron/rl/agent/reward_only_agent.py:35
Function
metrics_func
(model, epoch, output_predictions=False)
Megatron-LM/tasks/eval_utils.py:33
Method
mkdir_in_googledrive
(paths: List[str])
gui-rl/desktop_env/controllers/setup.py:712
Function
mlp_wrapper
Wrapper for Dense forward.
Megatron-LM/megatron/core/models/gpt/fine_grained_callables.py:487
Function
mock_forward
(*args, **kwargs)
Megatron-LM/tests/unit_tests/test_inference.py:33
Method
mock_get_empty_tensor_for_exchange
(*args, **kwargs)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_fully_parallel.py:345
Function
mock_hadamard_transform
Mock implementation of hadamard_transform for testing without the library installed. This is a simple identity-like transformation that preserves
Megatron-LM/tests/unit_tests/transformer/test_attention_variant_dsa.py:35
Function
mock_input
(prompt)
swe-rl/mini-swe-agent/tests/agents/test_interactive.py:151
Method
mock_local_plan
(self)
Megatron-LM/tests/unit_tests/dist_checkpointing/test_fully_parallel.py:610
Function
mock_optimizer
()
Megatron-LM/tests/unit_tests/test_optimizer_param_scheduler.py:12
Function
mock_query
(*args, **kwargs)
swe-rl/mini-swe-agent/tests/agents/test_interactive.py:141
Function
mock_response
Create a mock successful OpenRouter API response.
swe-rl/mini-swe-agent/tests/models/test_openrouter_model.py:17
Function
mock_response_no_cost
Create a mock OpenRouter API response without cost information.
swe-rl/mini-swe-agent/tests/models/test_openrouter_model.py:38
Method
mock_tokenize_prompt
(prompt, add_BOS=False)
Megatron-LM/tests/unit_tests/inference/engines/test_dynamic_engine.py:726
Function
mock_train_valid_test_datasets_provider
(train_val_test_num_samples)
Megatron-LM/tests/unit_tests/test_training.py:9
Function
model_config_path
Simple fixture returning path to model_config.yaml.
Megatron-LM/tests/functional_tests/python_test_utils/conftest.py:103
Function
model_parallel_is_initialized
Check if model- and data-parallel groups are initialized.
Megatron-LM/megatron/core/parallel_state.py:1347
Function
model_provider
Builds the model. Args: pre_process (bool, optional): Set to true if you need to compute embedings. Defaults to True.
Megatron-LM/pretrain_t5.py:68
Function
model_provider
Builds the model. Note: currently, only LLaVA model is supported. Follow-up changes will make this configurable. Args: pre_process (
Megatron-LM/pretrain_vlm.py:45
Function
model_provider
Build the model. Select between two different model classes: 1. Default model (uses megatron.legacy.models/gpt_model.py). 2. Core mod
Megatron-LM/pretrain_retro.py:66
Function
model_provider
Build the model.
Megatron-LM/pretrain_vision_dino.py:18
Function
model_provider
Build the model.
Megatron-LM/pretrain_vision_inpaint.py:18
Function
model_provider
Build the model.
Megatron-LM/pretrain_vision_classify.py:18
Function
model_provider
Build the model.
Megatron-LM/megatron/legacy/model/biencoder_model.py:22
Function
model_provider
( pre_process=True, post_process=True, layer_spec_fn=get_gpt_layer_with_transformer_engine_spec,
Megatron-LM/tests/unit_tests/transformer/moe/test_upcycling.py:56
Function
model_provider
Model provider for MIMO model training. Args: pre_process: Whether to pre-process the model post_process: Whether to post-process
Megatron-LM/examples/mimo/train.py:195
Function
model_provider
Build the model.
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/finetune_gpt.py:28
Function
model_provider
Builds the model. If you set the use_legacy_models to True, it will return the legacy GPT model and if not the core GPT model. Args:
Megatron-LM/examples/academic_paper_scripts/detxoify_lm/generate_samples_gpt.py:29
Function
model_provider
Builds the model. If you set the use_legacy_models to True, it will return the legacy GPT model and if not the mcore GPT model. Args
slime/slime/backends/megatron_utils/model_provider.py:189
Method
model_provider
( self, pre_process=True, post_process=True, layer_spec_fn=get_gpt_layer_with_
Megatron-LM/tests/unit_tests/transformer/test_multi_token_prediction.py:145
Method
model_provider
( self, pre_process=True, post_process=True, layer_spec_fn=get_gpt_decoder_blo
Megatron-LM/tests/unit_tests/transformer/test_cuda_graphs.py:1048
← previous
next →
8,201–8,300 of 10,561, ranked by callers