Code
Hub
Workspaces
Following
Trending
Connect
MCP
copy
Create free account
hub
/
github.com/Jiayi-Pan/TinyZero
/ functions
Functions
1,293 in github.com/Jiayi-Pan/TinyZero
⨍
Functions
1,293
◇
Types & classes
173
↳
Endpoints
30
Method
execute_model
( self, seq_group_metadata_list: List[SequenceGroupMetadata], kv_caches: List[torch.Te
verl/third_party/vllm/vllm_v_0_4_2/model_runner.py:238
Method
execute_model
(self, execute_model_req)
verl/third_party/vllm/vllm_v_0_4_2/spmd_gpu_executor.py:146
Method
execute_model
(self, execute_model_req: Optional[ExecuteModelRequest] = None)
verl/third_party/vllm/vllm_v_0_4_2/worker.py:209
Method
execute_model_async
Executes one model step on the given sequences.
verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py:249
Method
execute_model_async
Executes one model step on the given sequences.
verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py:246
Method
execute_model_async
Executes one model step on the given sequences.
verl/third_party/vllm/vllm_v_0_4_2/spmd_gpu_executor.py:211
Method
execute_rank_zero
(self, method_name: str, *args, **kwargs)
verl/single_controller/ray/base.py:326
Method
execute_rank_zero_sync
(self, method_name: str, *args, **kwargs)
verl/single_controller/ray/base.py:319
Function
exists
r"""Works like os.path.exists() but supports hdfs. Test whether a path exists. Returns False for broken symbolic links. Args: path (
verl/utils/hdfs_io.py:27
Function
extract_pg_from_exist
(resource_pools: Dict[str, RayResourcePool], src_role_names: List[str], resource_poo
verl/single_controller/ray/base.py:91
Function
extract_solution
(solution_str, *args)
examples/data_preprocess/multiply.py:62
Function
extract_step
(path)
verl/trainer/fsdp_sft_trainer.py:51
Function
fit
The training loop of PPO. The driver process only need to call the compute functions of the worker group through RPC to construct the PPO dat
examples/split_placement/split_monkey_patch.py:25
Method
flush
(self)
verl/utils/logger/aggregate_logger.py:37
Method
foo
(self, y)
tests/ray/test_worker_group_basics.py:48
Function
forward
( self, embedding: torch.Tensor, hidden_states: torch.Tensor, sampling_metadata: SamplingMetad
verl/third_party/vllm/vllm_v_0_3_1/model_loader.py:206
Method
forward
( ctx: Any, group: dist.ProcessGroup, local_input: Tensor, scatter_dim: int,
verl/utils/ulysses.py:167
Method
forward
(ctx: Any, group: dist.ProcessGroup, local_tensor: Tensor, gat
verl/utils/ulysses.py:200
Method
forward
(self, *args, **kwargs)
verl/utils/model.py:34
Method
forward
(ctx, vocab_parallel_logits: torch.Tensor)
verl/utils/megatron/tensor_parallel.py:103
Method
forward
r""" Args: labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): Labels for computing
verl/models/llama/megatron/modeling_llama_megatron.py:174
Method
forward
Args: input_ids: input ids. shape (1, totol_nnz) position_ids: position ids. shape (batch_size, seq_length)
verl/models/llama/megatron/modeling_llama_megatron.py:240
Method
forward
r""" Args: labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): Labels for computing
verl/models/llama/megatron/modeling_llama_megatron.py:308
Method
forward
( self, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.Tensor] = N
verl/models/llama/megatron/modeling_llama_megatron.py:384
Method
forward
Args: input_ids: input ids. shape (1, totol_nnz) position_ids: position ids. shape (batch_size, seq_length)
verl/models/llama/megatron/modeling_llama_megatron.py:467
Method
forward
r""" Args: labels (`torch.LongTensor` of shape `(batch_size, sequence_length)`, *optional*): Labels for computing
verl/models/llama/megatron/modeling_llama_megatron.py:562
Method
forward
( self, *, input_ids: torch.LongTensor = None, attention_mask: Optional[torch.
verl/models/llama/megatron/modeling_llama_megatron.py:644
Method
forward
Args: hidden_states (`torch.FloatTensor`): input to the layer of shape `(batch, seq_len, embed_dim)` attention_mask (
verl/models/llama/megatron/layers/parallel_decoder.py:44
Method
forward
( self, hidden_states: torch.Tensor, position_ids: Optional[torch.LongTensor] = None,
verl/models/llama/megatron/layers/parallel_decoder.py:112
Method
forward
(self, x)
verl/models/llama/megatron/layers/parallel_mlp.py:71
Method
forward
(self, x, seq_len=None)
verl/models/llama/megatron/layers/parallel_attention.py:61
Method
forward
( self, hidden_states: torch.Tensor, attention_mask: Optional[torch.Tensor] = None,
verl/models/llama/megatron/layers/parallel_attention.py:234
Method
forward
(self, hidden_states: torch.Tensor, position_ids: Optional[torch.LongTensor] =
verl/models/llama/megatron/layers/parallel_attention.py:340
Method
forward
(self, hidden_states)
verl/models/llama/megatron/layers/parallel_rmsnorm.py:41
Method
forward_step
(batch_iter, model)
verl/workers/critic/megatron_critic.py:165
Method
forward_step
(batch_iter, model)
verl/workers/actor/megatron_actor.py:286
Method
forward_step
(batch_iter, model)
verl/workers/reward_model/megatron/reward_model.py:222
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py:218
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_3_1/worker.py:176
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_6_3/llm.py:148
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_6_3/llm_engine_sp.py:349
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_6_3/spmd_gpu_executor.py:159
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_6_3/worker.py:239
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_5_4/llm.py:154
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_5_4/llm_engine_sp.py:275
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_5_4/spmd_gpu_executor.py:153
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_5_4/worker.py:233
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_4_2/llm.py:140
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_4_2/llm_engine_sp.py:241
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_4_2/spmd_gpu_executor.py:143
Method
free_cache_engine
(self)
verl/third_party/vllm/vllm_v_0_4_2/worker.py:203
Method
from_cli_args
(cls, args: argparse.Namespace)
verl/third_party/vllm/vllm_v_0_3_1/arg_utils.py:178
Method
from_cli_args
(cls, args: argparse.Namespace)
verl/third_party/vllm/vllm_v_0_5_4/arg_utils.py:257
Method
from_cli_args
(cls, args: argparse.Namespace)
verl/third_party/vllm/vllm_v_0_4_2/arg_utils.py:223
Function
gemma_dtensor_weight_loader
(actor_weights: Dict, vllm_model: nn.Module)
verl/third_party/vllm/vllm_v_0_6_3/dtensor_weight_loaders.py:24
Function
gemma_dtensor_weight_loader
(actor_weights: Dict, vllm_model: nn.Module)
verl/third_party/vllm/vllm_v_0_5_4/dtensor_weight_loaders.py:27
Function
gemma_dtensor_weight_loader
(actor_weights: Dict, vllm_model: nn.Module)
verl/third_party/vllm/vllm_v_0_4_2/dtensor_weight_loaders.py:26
Function
gemma_load_weights
(self, weights: Iterable[Tuple[str, torch.Tensor]])
verl/third_party/vllm/vllm_v_0_4_2/hf_weight_loader.py:30
Function
gen_dataset
Generate dataset for countdown task. Args: num_samples: Number of samples to generate num_operands: Number of numbers provide
examples/data_preprocess/countdown.py:15
Method
generate
Generates the completions for the input prompts. NOTE: This class automatically batches the given prompts, considering the memory con
verl/third_party/vllm/vllm_v_0_4_2/llm.py:152
Method
generate_sequences
(self, prompts: DataProto)
verl/workers/megatron_workers.py:345
Method
generate_sequences
(self, prompts: DataProto)
verl/workers/fsdp_workers.py:401
Method
generate_sequences
(self, prompts: DataProto)
verl/workers/rollout/hf_rollout.py:42
Method
generate_sequences
Generate sequences
verl/workers/rollout/naive/naive_rollout.py:52
Method
generate_sequences
(self, prompts: DataProto, **kwargs)
verl/workers/rollout/vllm_rollout/vllm_rollout.py:142
Method
get_added_vocab
Returns the added tokens in the vocabulary as a dictionary of token to index. Results might be different from the fast call because f
verl/workers/rollout/tokenizer.py:135
Method
get_cuda_visible_devices
(self)
verl/single_controller/base/worker.py:165
Method
get_data
()
verl/protocol.py:474
Function
get_default_kwargs_for_column_parallel_linear
()
verl/utils/megatron/tensor_parallel.py:58
Function
get_default_kwargs_for_parallel_embedding
()
verl/utils/megatron/tensor_parallel.py:77
Function
get_default_kwargs_for_row_parallel_linear
()
verl/utils/megatron/tensor_parallel.py:72
Function
get_device_mesh
()
verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py:281
Function
get_device_mesh
()
verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py:272
Function
get_device_mesh
()
verl/third_party/vllm/vllm_v_0_4_2/parallel_state.py:263
Method
get_head_size
(self)
verl/third_party/vllm/vllm_v_0_3_1/config.py:190
Method
get_hidden_size
(self)
verl/third_party/vllm/vllm_v_0_3_1/config.py:187
Function
get_kl_controller
(config)
verl/trainer/ppo/core_algos.py:56
Method
get_master_addr_port
(self)
verl/single_controller/base/worker.py:162
Method
get_max_input_len
Get the maximum input length for the LoRA request.
verl/third_party/vllm/vllm_v_0_5_4/tokenizer.py:39
Method
get_max_input_len
Get the maximum input length for the LoRA request.
verl/third_party/vllm/vllm_v_0_4_2/tokenizer.py:39
Method
get_megatron_global_info
(self)
verl/single_controller/base/megatron/worker_group.py:50
Method
get_megatron_rank_info
(self, rank: int)
verl/single_controller/base/megatron/worker_group.py:31
Method
get_memory_buffers
(self)
verl/utils/memory_buffer.py:153
Function
get_micro_data_parallel_rank
()
verl/third_party/vllm/vllm_v_0_3_1/parallel_state.py:146
Function
get_micro_data_parallel_world_size
()
verl/third_party/vllm/vllm_v_0_3_1/parallel_state.py:142
Method
get_model_config
Gets the model configuration.
verl/third_party/vllm/vllm_v_0_3_1/llm_engine_sp.py:336
Method
get_node_id
(self)
tests/ray/test_high_level_scheduling_api.py:29
Method
get_node_ip_by_sdk
()
verl/single_controller/base/worker.py:41
Method
get_num_kv_heads
Returns the number of KV heads per GPU.
verl/third_party/vllm/vllm_v_0_3_1/config.py:226
Method
get_rank_zero_info
(self)
verl/single_controller/base/register_center/ray.py:24
Method
get_special_tokens_mask
( self, token_ids_0: List[int], token_ids_1: Optional[List[int]] = None, alrea
tests/e2e/envs/digit_completion/tokenizer.py:114
Method
get_state
(self)
tests/e2e/envs/digit_completion/task.py:61
Function
get_tensor_model_parallel_rank
Return my rank for the tensor model parallel group.
verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py:302
Function
get_tensor_model_parallel_rank
Return my rank for the tensor model parallel group.
verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py:293
Function
get_tensor_model_parallel_src_rank
Calculate the global rank corresponding to the first local rank in the tensor model parallel group.
verl/third_party/vllm/vllm_v_0_3_1/parallel_state.py:124
Function
get_tensor_model_parallel_src_rank
Calculate the global rank corresponding to the first local rank in the tensor model parallel group.
verl/third_party/vllm/vllm_v_0_6_3/parallel_state.py:307
Function
get_tensor_model_parallel_src_rank
Calculate the global rank corresponding to the first local rank in the tensor model parallel group.
verl/third_party/vllm/vllm_v_0_5_4/parallel_state.py:298
Function
get_tensor_model_parallel_src_rank
Calculate the global rank corresponding to the first local rank in the tensor model parallel group.
verl/third_party/vllm/vllm_v_0_4_2/parallel_state.py:289
Function
get_tensor_parallel_partition_dim
(param)
verl/utils/megatron/tensor_parallel.py:90
Function
get_tensor_parallel_partition_stride
(param)
verl/utils/megatron/tensor_parallel.py:95
← previous
next →
901–1,000 of 1,293, ranked by callers