import copy import functools from typing import ( Any, Dict, Iterable, Iterator, List, NamedTuple, Optional, Set, Tuple, Union, ) import torch import torch.distributed as dist # Import the entire FSDP file to avoid circular imports import torch.distributed.fsdp.fully_sharded_data_parallel as FSDP from torch.distributed.fsdp.flatten_params_wrapper import FlatParameter class _ConsolidatedOptimState: """ This holds the consolidated optimizer state on the target rank. Positive- dimension tensor state is communicated across ranks, while zero-dimension tensor state and non-tensor state is taken directly from the target rank. PyTorch version 1.12 moved to using zero-dimension tensors for scalar values, but user implemented optimizers may still use float (i.e. a non-tensor). Thus, we support both and handle them identically. Attributes: tensor_state (Dict[str, torch.Tensor]): Mapping from positive-dimension tensor state name to the unsharded flattened tensor representing the state. zero_dim_tensor_state (Dict[str, torch.Tensor]): Mapping from zero- dimension tensor state name to its value. non_tensor_state (Dict[str, Any]): Mapping from non-tensor state name to its value. """ tensor_state: Dict[str, torch.Tensor] = {} zero_dim_tensor_state: Dict[str, torch.Tensor] = {} non_tensor_state: Dict[str, Any] = {} class _PosDimTensorInfo(NamedTuple): """ Meatadata for positive-dimension tensors used internally for :meth:`scatter_full_optim_state_dict`. Attributes: shape (torch.Size): Sharded tensor shape (which is equal to the unsharded tensor shape if the tensor is optimizer state for a non-FSDP parameter and is hence not sharded). dtype (torch.dtype): Data type of the tensor. """ shape: torch.Size dtype: torch.dtype def _unflatten_optim_state( fsdp_module, flat_param: FlatParameter, flat_param_state: Dict[str, Any], to_save: bool, ) -> List[Dict[str, Any]]: """ Unflattens the optimizer state, consisting of the "state" part and the "param_groups" part. Unflattening the "state" part involves consolidating the state on the target rank and remapping from flattened to unflattened parameter IDs, and the "param_groups" part only involves remapping from flattened to unflattened parameter IDs. Args: fsdp_module (FullyShardedDataParallel): FSDP module that owns ``flat_param``, i.e. holds it in ``self.params``. flat_param (FlatParameter): The flattened parameter. flat_param_state (Dict[str, Any]): Entry for the flattened parameter in the "state" part of the optimizer state dict. to_save (bool): Whether to save the state on this rank. Returns: List[Dict[str, Any]]: A :class:`list` holding the entries in the "state" part of the optimizer state dict corresponding to the unflattened parameters comprising the flattened parameter ``flat_param`` if on the target rank or an empty :class:`list` otherwise. The final optimizer state dict will need to map these entries using the proper unflattened parameter IDs. """ assert sum(p is flat_param for p in fsdp_module.params) == 1, \ "`fsdp_module` must own `flat_param`" consolidated_state = _communicate_optim_state( fsdp_module, flat_param, flat_param_state, to_save, ) unflat_param_state = _unflatten_communicated_optim_state( fsdp_module, flat_param, consolidated_state, ) if to_save else [] return unflat_param_state def _communicate_optim_state( fsdp_module, flat_param: FlatParameter, flat_param_state: Dict[str, Any], to_save: bool, ) -> _ConsolidatedOptimState: """ Communicates the optimizer state for a flattened parameter ``flat_param`` across ranks so that the target rank holds the entire non-sharded optimizer state. If ``N`` is the number of tensor optimizer states in the optimizer state dict, then the communication complexity is 0 if ``N = 0`` and ``N + 1`` otherwise (where the plus 1 comes from all-gathering the padding per rank). Args: flat_param (FlatParameter): The flattened parameter. flat_param_state (Dict[str, Any]): The entry in the "state" part of the optimizer state dict corresponding to the flattened parameter. to_save (bool): Whether to save the state on this rank. Returns: ConsolidatedOptimState: Consolidated optimizer state for ``flat_param``; the state is not populated for non-target ranks. """ param_index = -1 for i, param in enumerate(fsdp_module.params): if param is flat_param: param_index = i break assert param_index >= 0, "`fsdp_module` must own `flat_param`" state = _ConsolidatedOptimState() tensor_state, zero_dim_tensor_state, non_tensor_state = \ state.tensor_state, state.zero_dim_tensor_state, state.non_tensor_state process_group = fsdp_module.process_group tensor_buffer = None # initialize lazily in case it is not needed for state_name, value in flat_param_state.items(): # Positive-dimension tensor state: communicate across ranks if torch.is_tensor(value) and value.dim() > 0: # If the parameter is not sharded (e.g. world size of 1), then # neither is the positive-dimension tensor state, so no need to # communicate it -- we take the target rank's value if not flat_param._is_sharded: tensor_state[state_name] = value.cpu() continue if tensor_buffer is None: # Assume that positive-dimension tensor optimizer state # has the same shape as the sharded flattened parameter buffer_size = flat_param._full_param_padded.size() # type: ignore[attr-defined] tensor_buffer = value.new_zeros(*buffer_size) dist._all_gather_base(tensor_buffer, value, group=process_group) if to_save: assert hasattr(flat_param, "_orig_size"), \ "Sharded flattened parameter should have `_orig_size` set" unpadded_numel = flat_param._orig_size.numel() # type: ignore[attr-defined] tensor_state[state_name] = tensor_buffer[:unpadded_numel].cpu() # Zero-dimension tensor state and non-tensor state: take this rank's # value directly elif to_save: if _is_zero_dim_tensor(value): zero_dim_tensor_state[state_name] = value.cpu() else: non_tensor_state[state_name] = value return state def _unflatten_communicated_optim_state( fsdp_module, flat_param: FlatParameter, state: _ConsolidatedOptimState, ) -> List[Dict[str, Any]]: """ Unflattens the communicated optimizer state (given by ``tensor_state``, ``non_tensor_state``, and ``zero_dim_tensor_state``) for a single flattened parameter ``flat_param``. This should only be called on the target rank. Args: fsdp_module (FullyShardedDataParallel): FSDP module that owns ``flat_param``, i.e. holds it in ``self.params``. flat_param (FlatParameter): The flattened parameter. state (_ConsolidatedOptimState): Consolidated optimizer state. Returns: List[Dict[str, Any]]: A :class:`list` holding the entries in the "state" part of the optimizer state dict corresponding to the unflattened parameters comprising the flattened parameter ``flat_param``. The final optimizer state dict will need to map these entries using the proper unflattened parameter IDs. """ assert sum(p is flat_param for p in fsdp_module.params) == 1, \ "`fsdp_module` must own `flat_param`" unflat_param_state: List[Dict[str, Any]] = [] flat_param_views: Dict[str, Iterator] = {} num_unflat_params = flat_param._num_unflattened_params tensor_state, zero_dim_tensor_state, non_tensor_state = \ state.tensor_state, state.zero_dim_tensor_state, state.non_tensor_state for _ in range(num_unflat_params): unflat_state_param = {} # Add positive-dimension tensor state: unflatten with views for state_name, flat_tensor in tensor_state.items(): views_generated = state_name in flat_param_views if not views_generated: param_views = flat_param.get_param_views(flat_tensor) flat_param_views[state_name] = param_views else: param_views = flat_param_views[state_name] unflat_state_param[state_name] = next(param_views) # Add zero-dimension tensor state: take the target rank's value for state_name, zero_dim_tensor in zero_dim_tensor_state.items(): unflat_state_param[state_name] = zero_dim_tensor # Add non-tensor state: take the target rank's value for state_name, non_tensor in non_tensor_state.items(): unflat_state_param[state_name] = non_tensor unflat_param_state.append(unflat_state_param) return unflat_param_state def _flatten_full_optim_state_dict( full_optim_state_dict: Dict[str, Any], model: torch.nn.Module, shard_state: bool, optim_input: Optional[Union[ List[Dict[str, Any]], Iterable[torch.nn.Parameter], ]] = None, ) -> Tuple[Dict[str, Any], Set[int]]: """ Args: shard_state (bool): Whether to shard flattened positive-dimension tensor state; if ``False``, then the full flattened tensor is kept in the returned :class:`dict. Returns: Tuple[Dict[str, Any], Set[int]]: The flattened optimizer state dict and a set of the parameter IDs corresponding to FSDP parameters. """ full_osd = full_optim_state_dict # alias if "state" not in full_osd or "param_groups" not in full_osd: raise ValueError( "`full_optim_state_dict` must have the keys \"state\" and " "\"param_groups\" to be a valid optimizer state dict" ) flat_param_id_to_param = _get_param_id_to_param(model, optim_input) flat_param_to_fsdp_module = _get_flat_param_to_fsdp_module(model) param_to_unflat_param_names = FSDP._get_param_to_unflat_param_names(model) # Handle the "state" part of the optimizer state dict flat_osd_state: Dict[int, Any] = {} full_osd_state = full_osd["state"] unflat_param_names_to_flat_param_id: Dict[str, int] = {} fsdp_flat_param_ids = set() # save which IDs are for FSDP parameters for flat_param_id, param in enumerate(flat_param_id_to_param): # type: ignore[assignment] assert param in param_to_unflat_param_names, \ "Check the `param_to_unflat_params` construction\n" \ f"param: {param}" unflat_param_names = param_to_unflat_param_names[param] # For FSDP parameters, we need to flatten if isinstance(param, FlatParameter): assert param in flat_param_to_fsdp_module, \ "Check the `flat_param_to_fsdp_module` mapping " \ f"construction\nparam={param}" unflat_param_names = param_to_unflat_param_names[param] fsdp_module = flat_param_to_fsdp_module[param] flat_state = _flatten_optim_state( full_osd_state, unflat_param_names, fsdp_module, param, shard_state, ) flat_osd_state[flat_param_id] = flat_state for unflat_param_name in unflat_param_names: unflat_param_names_to_flat_param_id[unflat_param_name] = flat_param_id fsdp_flat_param_ids.add(flat_param_id) # For parameters from non-FSDP modules, we do not need to flatten else: assert len(unflat_param_names) == 1 unflat_param_name = unflat_param_names[0] if unflat_param_name not in full_osd_state: # A non-FSDP module's parameter may be ignored and hence not # have an entry in the optimizer state continue # Remap from unflattened to flattened parameter ID -- do not # deepcopy to avoid unnecessarily duplicating tensor storage flat_osd_state[flat_param_id] = \ copy.copy(full_osd_state[unflat_param_name]) unflat_param_names_to_flat_param_id[unflat_param_name] = flat_param_id # Handle the "param_groups" part of the optimizer state dict sharded_osd_param_groups: List[Dict[str, Any]] = [] for unflat_param_group in full_osd["param_groups"]: flat_param_group = copy.deepcopy(unflat_param_group) # Map from unflattened parameter names to flattened parameter IDs flat_param_ids = sorted(set( unflat_param_names_to_flat_param_id[unflat_param_name] for unflat_param_name in unflat_param_group["params"] )) flat_param_group["params"] = flat_param_ids sharded_osd_param_groups.append(flat_param_group) optim_state_dict = { "state": flat_osd_state, "param_groups": sharded_osd_param_groups, } return optim_state_dict, fsdp_flat_param_ids def _flatten_optim_state( unflat_osd_state: Dict[str, Dict[str, Any]], unflat_param_names: List[str], fsdp_module, flat_param: FlatParameter, shard_state: bool, ) -> Dict[str, Any]: """ Flattens the optimizer state in ``full_optim_state_dict`` for a single flattened parameter ``flat_param`` in ``fsdp_module`` corresponding to the unflattened parameter names in ``unflat_param_names``. Args: unflat_osd_state (Dict[str, Dict[str, Any]]): The "state" part of the optimizer state dict corresponding to the unflattened parameters. unflat_param_names (List[str]): A :class:`list` of unflattened parameter names corresponding to the flattened parameter ``flat_param``. fsdp_module (FullyShardedDataParallel): FSDP module owning the flattened parameter. flat_param (FlatParameter): The flattened parameter. shard_state (bool): Whether to shard flattened positive-dimension tensor state; if ``False``, then the full flattened tensor is kept in the returned :class:`dict. Returns: Dict[str, Any]: A :class:`dict` mapping state names to their values for a particular flattened parameter. The sharded optimizer state dict's "state" part will map the flattened parameter ID to this returned value. """ num_unflat_params = len(unflat_param_names) assert num_unflat_params > 0, \ "Expects at least one unflattened parameter corresponding to the " \ "flattened parameter" unflat_param_shapes = flat_param._param_shapes num_unflat_param_shapes = len(unflat_param_shapes) assert num_unflat_params == num_unflat_param_shapes, \ f"Expects {num_unflat_params} shapes but got {num_unflat_param_shapes}" # Check if these unflattened parameters have any optimizer state has_state = [ bool(unflat_param_name in unflat_osd_state) for unflat_param_name in unflat_param_names ] # If none of the unflattened parameters comprising this flattened parameter # have any state, then we do not want an entry in the optimizer state dict if not any(has_state): return {} # no need to flatten any state # There may still be some unflattened parameters with state and some # without unflat_param_states = [ unflat_osd_state[unflat_param_name] if unflat_param_name in unflat_osd_state else None for unflat_param_name in unflat_param_names ] # Check that the unflattened parameters have the same state names state_names = None for unflat_param_state in unflat_param_states: if unflat_param_state is None: continue if state_names is None: state_names = set(unflat_param_state.keys()) else: if state_names != set(unflat_param_state.keys()): raise ValueError( "Differing optimizer state names for the unflattened " f"parameters: {unflat_param_names}" ) assert state_names is not None # Flatten the state flat_state: Dict[str, Any] = {} for state_name in state_names: state_values = [ unflat_param_state[state_name] if unflat_param_state is not None else None for unflat_param_state in unflat_param_states ] non_none_state_values = [v for v in state_values if v is not None] are_pos_dim_tensors = are_zero_dim_tensors = are_non_tensors = True for v in non_none_state_values: are_pos_dim_tensors &= torch.is_tensor(v) and v.dim() > 0 are_zero_dim_tensors &= _is_zero_dim_tensor(v) are_non_tensors &= not torch.is_tensor(v) types = set(type(v) for v in non_none_state_values) if len(types) != 1 or not ( are_pos_dim_tensors or are_zero_dim_tensors or are_non_tensors ): raise ValueError( f"Differing optimizer state types for state {state_name}, " f"values {non_none_state_values}, and unflattened parameter " f"names {unflat_param_names}" ) if are_pos_dim_tensors: flat_tensor = _flatten_tensor_optim_state( state_name, state_values, unflat_param_names, unflat_param_shapes, flat_param, ) if shard_state: # Shard the flattened tensor immediately to minimize max memory # usage sharded_flat_tensor, _ = fsdp_module._get_shard(flat_tensor) flat_state[state_name] = sharded_flat_tensor else: flat_state[state_name] = flat_tensor elif are_zero_dim_tensors: flat_state[state_name] = _flatten_zero_dim_tensor_optim_state( state_name, state_values, unflat_param_names, ) else: assert are_non_tensors flat_state[state_name] = _flatten_non_tensor_optim_state( state_name, state_values, unflat_param_names, ) return flat_state def _flatten_tensor_optim_state( state_name: str, pos_dim_tensors: List[torch.Tensor], unflat_param_names: List[str], unflat_param_shapes: List[torch.Size], flat_param: FlatParameter, ) -> torch.Tensor: """ Flattens the positive-dimension tensor optimizer state given by the values ``tensors`` for the state ``state_name`` for a single flattened parameter ``flat_param`` corresponding to the unflattened parameter names ``unflat_param_names`` and unflatted parameter shapes ``unflat_param_shapes``. This flattens each unflattened parameter's tensor state into one tensor. NOTE: We use zero tensors for any unflattened parameters without state since some value is required to fill those entries. This assumes that the zero tensor is mathematically equivalent to having no state, which is true for Adam's ``exp_avg`` and ``exp_avg_sq`` but may not be true for all optimizers. Args: state_name (str): Optimizer state name. pos_dim_tensors (List[torch.Tensor]): Positive-dimension tensor optimizer state values for the unflattened parameters corresponding to the single flattened parameter. unflat_param_names (List[str]): A :class:`list` of unflattened parameter names corresponding to the single flattened parameter. unflat_param_shapes (List[torch.Size]): Unflattened parameter shapes corresponding to the single flattened parameter. flat_param (FlatParameter): The flattened parameter. Returns: torch.Tensor: A flattened tensor containing the optimizer state corresponding to ``state_name`` constructed by concatenating the unflattened parameter tensor states in ``pos_dim_tensors`` (using zero tensors for any unflattened parameters without the state). """ non_none_tensors = [t for t in pos_dim_tensors if t is not None] # Check that all are tensors with the same dtype dtypes = set(t.dtype for t in non_none_tensors) if len(dtypes) != 1: raise ValueError( "All unflattened parameters comprising a single flattened " "parameter must have positive-dimension tensor state with the " f"same dtype but got dtypes {dtypes} for state {state_name} and " f"unflattened parameter names {unflat_param_names}" ) dtype = next(iter(dtypes)) # Check that each tensor state matches its parameter's shape for tensor, shape in zip(pos_dim_tensors, unflat_param_shapes): if tensor is None and len(shape) == 0: raise ValueError( "Flattening a zero-dimension parameter is not supported" ) elif tensor is not None and tensor.shape != shape: raise ValueError( "Tensor optimizer state does not have same shape as its " f"parameter: {tensor.shape} {shape}" ) # Flatten the tensor states cpu_device = torch.device("cpu") tensors = [ torch.flatten(state_value.to(cpu_device)) if state_value is not None else torch.flatten(torch.zeros( size=shape, dtype=dtype, device=cpu_device, )) for state_value, shape in zip(pos_dim_tensors, unflat_param_shapes) ] padding = flat_param.num_padded if padding > 0: tensors.append(torch.zeros(padding, dtype=dtype, device=cpu_device)) flat_tensor = torch.cat(tensors) # `flat_tensor`'s shape should be 1D and less than or equal to the # flattened parameter's shape (where the inequality is strict for positive # padding) if not flat_param._is_sharded: # currently, only when world size is 1 # If the parameter is not sharded, then `_full_param_padded` is not # used, so we skip the shape check return flat_tensor full_padded_dim = flat_param._full_param_padded.dim() # type: ignore[attr-defined] full_padded_shape = flat_param._full_param_padded.shape # type: ignore[attr-defined] assert flat_tensor.dim() == 1, \ f"`flat_tensor` should be 1D but got {flat_tensor.dim()} dims" assert full_padded_dim == 1, \ f"`_full_param_padded` should be 1D but got {full_padded_dim} dims" assert flat_tensor.shape[0] <= full_padded_shape[0], \ f"tensor optim state: {flat_tensor.shape} " \ f"parameter: {full_padded_shape}" return flat_tensor def _flatten_zero_dim_tensor_optim_state( state_name: str, zero_dim_tensors: List[torch.Tensor], unflat_param_names: List[str], ) -> torch.Tensor: """ Flattens the zero-dimension tensor optimizer state given by the values ``zero_dim_tensors`` for the state ``state_name`` for a single flattened parameter corresponding to the unflattened parameter names ``unflat_param_names`` by enforcing that all tensors are the same and using that common value. NOTE: The requirement that the tensors are the same across all unflattened parameters comprising the flattened parameter is needed to maintain the invariant that FSDP performs the same computation as its non-sharded equivalent. This means that none of the unflattened parameters can be missing this state since imposing a value may differ from having no value. For example, for Adam's "step", no value means maximum bias correction, while having some positive value means less bias correction. Args: state_name (str): Optimizer state name. zero_dim_tensors (List[torch.Tensor]): Zero-dimension optimizer state for the unflattened parameters corresponding to the single flattened parameter. unflat_param_names (List[str]): A :class:`list` of unflattened parameter names corresponding to the single flattened parameter. Returns: torch.Tensor: A zero-dimensional tensor giving the value of the state ``state_name`` for all unflattened parameters corresponding to the names ``unflat_param_names``. """ non_none_tensors = [t for t in zero_dim_tensors if t is not None] # Enforce that all have the same value and dtype values_set = set(t.item() if t is not None else None for t in zero_dim_tensors) dtypes = set(t.dtype if t is not None else None for t in zero_dim_tensors) if len(non_none_tensors) != len(zero_dim_tensors) or \ len(values_set) != 1 or len(dtypes) != 1: raise ValueError( "All unflattened parameters comprising a single flattened " "parameter must have scalar state with the same value and dtype " f"but got values {values_set} and dtypes {dtypes} for state " f"{state_name} and unflattened parameter names " f"{unflat_param_names}" ) value = next(iter(values_set)) dtype = next(iter(dtypes)) return torch.tensor(value, dtype=dtype, device=torch.device("cpu")) def _flatten_non_tensor_optim_state( state_name: str, non_tensors: List[Any], unflat_param_names: List[str], ) -> Any: """ Flattens the non-tensor optimizer state given by the values ``non_tensors`` for the state ``state_name`` for a single flattened parameter corresponding to the unflattened parameter names ``unflat_param_names`` by enforcing that all values are the same and using that common value. See the note in :func:`_flatten_zero_dim_tensor_optim_state`. Args: state_name (str): Optimizer state name. non_tensors (List[Any]): Non-tensor optimizer state for the unflattened parameters corresponding to the single flattened parameter. unflat_param_names (List[str]): A :class:`list` of unflattened parameter names corresponding to the single flattened parameter. Returns: Any: A non-tensor giving the value of the state ``state_name`` for all unflattened parameters corresponding to the names ``unflat_param_names``. """ non_none_non_tensors = [nt for nt in non_tensors if nt is not None] # Enforce that all have the same value (same type already checked) non_tensor_set = set(non_tensors) if len(non_none_non_tensors) != len(non_tensors) or \ len(non_tensor_set) != 1: raise ValueError( "All unflattened parameters comprising a single flattened " "parameter must have scalar state with the same value and dtype " f"but got values {non_tensor_set} for state {state_name} and " f"unflattened parameter names {unflat_param_names}" ) non_tensor = next(iter(non_tensor_set)) return non_tensor def _process_pos_dim_tensor_state( flat_optim_state_dict: Dict[str, Any], fsdp_flat_param_ids: Set[int], world_size: int, ) -> Dict[str, Any]: """ Processes positive-dimension tensor states in ``flat_optim_state_dict`` by replacing them with metadata. This is done so the processed optimizer state dict can be broadcast from rank 0 to all ranks without copying those tensor states, and thus, this is meant to only be called on rank 0. Args: flat_optim_state_dict (Dict[str, Any]): Flattened optimizer state dict with the positive-dimension tensor states unsharded; this should be returned by :meth:`_flatten_optim_state` with ``shard_state=False``. fsdp_flat_param_ids (Set[int]): Parameter IDs corresponding to FSDP parameters. Returns: Dict[str, Any]: The flattened optimizer state dict with positive- dimension tensor states replaced by metadata. """ flat_osd = flat_optim_state_dict # alias no_tensor_osd: Dict[str, Any] = {"state": {}} cpu_device = torch.device("cpu") for param_id, param_state in flat_osd["state"].items(): no_tensor_osd["state"][param_id] = {} for state_name, state_value in param_state.items(): is_pos_dim_tensor_state = torch.is_tensor(state_value) and \ state_value.dim() > 0 if not is_pos_dim_tensor_state: no_tensor_osd["state"][param_id][state_name] = state_value continue if param_id in fsdp_flat_param_ids: # FSDP parameter # Use `_get_chunk()` to get a view and avoid allocating any new # tensor storage via either `clone()` or `pad()`; each rank's # chunk has the same padded shape, so we can pass rank 0 chunk, num_to_pad = FSDP.FullyShardedDataParallel._get_chunk( state_value, 0, world_size, ) assert len(chunk.shape) == 1, \ f"Chunk should be 1D but got {chunk.shape}" # Include the padding to get the final shard shape info = _PosDimTensorInfo( shape=torch.Size([chunk.shape[0] + num_to_pad]), dtype=chunk.dtype, ) else: # non-FSDP parameter info = _PosDimTensorInfo( shape=state_value.shape, dtype=state_value.dtype, ) no_tensor_osd["state"][param_id][state_name] = info no_tensor_osd["param_groups"] = copy.deepcopy(flat_osd["param_groups"]) return no_tensor_osd def _broadcast_processed_optim_state_dict( processed_optim_state_dict: Optional[Dict[str, Any]], fsdp_flat_param_ids: Optional[Set[int]], rank: int, group, ) -> Tuple[Dict[str, Any], Set[int]]: """ Broadcasts the processed optimizer state dict and the accompanying FSDP parameter IDs from rank 0 to all ranks. Args: processed_optim_state_dict (Optional[Dict[str, Any]]): The full optimizer state dict with positive-dimension tensor states replaced with metadata if on rank 0; ignored otherwise. fsdp_flat_param_ids (Optional[Set[int]]): Parameter IDs corresponding to FSDP parameters if on rank 0; ignored otherwise. Returns: Tuple[Dict[str, Any], Set[int]]: The processed optimizer state dict and the parameter IDs corresponding to FSDP parameters. """ # Broadcast the two data structures rank 0 to all ranks obj_list = [processed_optim_state_dict, fsdp_flat_param_ids] if rank == 0 \ else [None, None] dist.broadcast_object_list(obj_list, src=0, group=group) processed_optim_state_dict, fsdp_flat_param_ids = obj_list # type: ignore[assignment] assert processed_optim_state_dict is not None assert fsdp_flat_param_ids is not None # Keep zero-dimension tensors on CPU return processed_optim_state_dict, fsdp_flat_param_ids def _broadcast_pos_dim_tensor_states( processed_optim_state_dict: Dict[str, Any], fsdp_flat_param_ids: Set[int], flat_optim_state_dict: Optional[Dict[str, Any]], rank: int, world_size: int, group, broadcast_device: torch.device, ) -> Dict[str, Any]: """ Takes ``processed_optim_state_dict``, which has metadata in place of positive-dimension tensor states, and broadcasts those tensor states from rank 0 to all ranks. For tensor states corresponding to FSDP parameters, rank 0 shards the tensor and broadcasts shard-by-shard, and for tensor states corresponding to non-FSDP parameters, rank 0 broadcasts the full tensor. Args: processed_optim_state_dict (Dict[str, Any]): The full optimizer state dict with positive-dimension tensor states replaced with metadata; should be returned by :meth:`_process_pos_dim_tensor_state` and non-empty on all ranks (e.g. via a ``broadcast()`` from rank 0). fsdp_flat_param_ids (Set[int]): Parameter IDs corresponding to FSDP parameters. flat_optim_state_dict (Optional[Dict[str, Any]]): Flattened optimizer state dict if on rank 0; ignored on nonzero ranks. Returns: Dict[str, Any]: The optimizer state dict with the positive-dimension tensor state correctly populated via ``broadcast()`` s from rank 0. """ assert rank != 0 or flat_optim_state_dict is not None, \ "Expects rank 0 to pass in the flattened optimizer state dict" no_tensor_osd = processed_optim_state_dict # alias flat_osd = flat_optim_state_dict # alias for param_id, param_state in no_tensor_osd["state"].items(): for state_name, value in param_state.items(): is_pos_dim_tensor_state = isinstance(value, _PosDimTensorInfo) if not is_pos_dim_tensor_state: continue if rank == 0: assert flat_osd is not None unsharded_tensor = flat_osd["state"][param_id][state_name] else: unsharded_tensor = None shape, dtype = value.shape, value.dtype if param_id in fsdp_flat_param_ids: # FSDP parameter _broadcast_sharded_pos_dim_tensor_state( unsharded_tensor, param_state, state_name, shape, dtype, broadcast_device, rank, world_size, group, ) # modify `param_state` destructively else: # non-FSDP parameter _broadcast_unsharded_pos_dim_tensor_state( unsharded_tensor, param_state, state_name, shape, dtype, broadcast_device, rank, group, ) # modify `param_state` destructively return no_tensor_osd def _broadcast_sharded_pos_dim_tensor_state( unsharded_tensor: Optional[torch.Tensor], param_state: Dict[str, Any], state_name: str, shape: torch.Size, dtype: torch.dtype, broadcast_device: torch.device, rank: int, world_size: int, group, ) -> None: """ Broadcasts positive-dimension tensor state for the state ``state_name`` corresponding to an FSDP parameter shard-by-shard, only to be saved on the relevant rank. This modifies ``param_state`` destructively. Args: unsharded_tensor (Optional[torch.Tensor]): Unsharded tensor from which to broadcast shards if on rank 0; ignored otherwise. shape (torch.Size): Shape of the sharded tensor; same on all ranks. """ get_shard: Optional[functools.partial[Tuple[torch.Tensor, int]]] = None if rank == 0: assert unsharded_tensor is not None, \ "Expects rank 0 to pass in the unsharded tensor" get_shard = functools.partial( FSDP.FullyShardedDataParallel._get_shard_functional, unsharded_tensor, ) for target_rank in range(1, world_size): if rank == 0: assert get_shard is not None sharded_tensor = get_shard(target_rank, world_size)[0].to(broadcast_device) else: sharded_tensor = torch.zeros( shape, requires_grad=False, dtype=dtype, device=broadcast_device, ) dist.broadcast(sharded_tensor, src=0, group=group) # Only keep the shard on the target rank and keep it on the broadcast # device, which is typically GPU if rank == target_rank: param_state[state_name] = sharded_tensor else: del sharded_tensor # Lastly, shard on rank 0 if rank != 0: return param_state[state_name] = get_shard(0, world_size)[0].to(broadcast_device) # type: ignore[misc] def _broadcast_unsharded_pos_dim_tensor_state( unsharded_tensor: Optional[torch.Tensor], param_state: Dict[str, Any], state_name: str, shape: torch.Size, dtype: torch.dtype, broadcast_device: torch.device, rank: int, group, ) -> None: """ Broadcasts positive-dimension tensor state for the state ``state_name`` corresponding to an unsharded non-FSDP parameter from rank 0 to all ranks. This modifies ``param_state`` destructively. Args: unsharded_tensor (Optional[torch.Tensor]): Unsharded tensor to broadcast if on rank 0; ignored otherwise. """ if rank == 0: assert unsharded_tensor is not None, \ "Expects rank 0 to pass in the unsharded tensor" assert shape == unsharded_tensor.shape, \ f"Shape mismatch: {shape} {unsharded_tensor.shape}" assert dtype == unsharded_tensor.dtype, \ f"dtype mismatch: {dtype} {unsharded_tensor.dtype}" unsharded_tensor = unsharded_tensor.to(broadcast_device) else: unsharded_tensor = torch.zeros( shape, requires_grad=False, dtype=dtype, device=broadcast_device, ) dist.broadcast(unsharded_tensor, src=0, group=group) # Keep the tensor on the broadcast device, which is typically GPU param_state[state_name] = unsharded_tensor def _get_flat_param_to_fsdp_module(model: torch.nn.Module): """ Constructs a mapping from FSDP flattened parameters to their owning FSDP modules and ensures that all FSDP modules are initialized. Args: model (torch.nn.model): Root module (which may or may not be a :class:`FullyShardedDataParallel` instance). Returns: Dict[FlatParameter, FullyShardedDataParallel]: Mapping from FSDP flattened parameters to their owning FSDP modules. """ flat_param_to_fsdp_module = {} for module in model.modules(): if isinstance(module, FSDP.FullyShardedDataParallel): module._lazy_init() for param in module.params: # may have none flat_param_to_fsdp_module[param] = module return flat_param_to_fsdp_module def _get_param_id_to_param( model: torch.nn.Module, optim_input: Optional[Union[ List[Dict[str, Any]], Iterable[torch.nn.Parameter], ]] = None, ) -> List[torch.nn.Parameter]: """ Constructs a mapping from parameter IDs to parameters. This may be used both for models with ``FlatParameter`` s and without. NOTE: We critically assume that, whether the optimizer input is a list of parameters or a list of parameter groups, :class:`torch.optim.Optimizer` enumerates the parameter IDs in order. In other words, for a parameter list input, the parameter IDs should be in that list order, and for a parameter groups input, the parameter IDs should be in order within each parameter group and in order across parameter groups. Args: model (torch.nn.Module): Model whose parameters are passed into the optimizer. optim_input (Optional[Union[List[Dict[str, Any]], Iterable[torch.nn.Parameter]]]): Input passed into the optimizer representing either a :class:`list` of parameter groups or an iterable of parameters; if ``None``, then this method assumes the input was ``model.parameters()``. (Default: ``None``) Returns: List[torch.nn.Parameter]: Mapping from parameter IDs to parameters, where the parameter ID is implicitly the index in the :class:`list`. """ # Assume the standard case of passing `model.parameters()` to the optimizer # if `optim_input` is not specified if optim_input is None: return list(model.parameters()) try: params = list(optim_input) except TypeError: raise TypeError( "Optimizer input should be an iterable of Tensors or dicts, " f"but got {optim_input}" ) if len(params) == 0: raise ValueError("Optimizer input should not be empty") # Check if the optimizer input represents tensors or parameter groups all_tensors = True all_dicts = True for param in params: all_tensors &= isinstance(param, torch.Tensor) all_dicts &= isinstance(param, dict) if not all_tensors and not all_dicts: raise TypeError( "Optimizer input should be an iterable of Tensors or dicts" ) if all_tensors: return params # type: ignore[return-value] assert all_dicts param_id_to_param = [] for param_group in params: has_params_key = "params" in param_group # type: ignore[operator] assert has_params_key, \ "A parameter group should map \"params\" to a list of the " \ "parameters in the group" for param in param_group["params"]: # type: ignore[index] # Implicitly map `flat_param_id` (current length of the list) to # `param` param_id_to_param.append(param) return param_id_to_param # type: ignore[return-value] def _get_param_to_param_id( model: torch.nn.Module, optim_input: Optional[Union[ List[Dict[str, Any]], Iterable[torch.nn.Parameter], ]] = None, ) -> Dict[torch.nn.Parameter, int]: """Constructs the inverse mapping of :func:`_get_param_id_to_param`.""" param_id_to_param = _get_param_id_to_param(model, optim_input) return { param: param_id for param_id, param in enumerate(param_id_to_param) } def _get_unflat_to_flat_param_ids( flat_to_unflat_param_ids: Dict[int, List[int]], ) -> List[int]: """ Inverts the mapping ``flat_to_unflat_param_ids`` to be from unflattened parameter ID to flattened parameter ID, where the unflattened parameter ID is the index in the returned :class:`list`. There may be multiple unflattened parameter IDs mapping to the same flattened parameter ID. Args: flat_to_unflat_param_ids (Dict[int, List[int]]): A mapping from flattened parameter ID to a :class:`list` of corresponding unflattened parameter IDs. Returns: List[int]: A mapping from unflattened parameter ID to flattened parameter ID, where the unflattened parameter ID is the index in the :class:`list`. """ # Construct as a dict and then convert to list unflat_to_flat_param_ids = {} for flat_param_id, unflat_param_ids in flat_to_unflat_param_ids.items(): for unflat_param_id in unflat_param_ids: assert unflat_param_id not in unflat_to_flat_param_ids, \ "`flat_to_unflat_param_ids` has the unflattened parameter " \ f"ID {unflat_param_id} mapped to multiple flattened " \ "parameter IDs" unflat_to_flat_param_ids[unflat_param_id] = flat_param_id num_unflat_param_ids = len(unflat_to_flat_param_ids) unflat_param_ids_set = set(unflat_to_flat_param_ids.keys()) assert unflat_param_ids_set == set(range(num_unflat_param_ids)), \ "The set of unflattened parameter IDs should be {0, ..., " + \ str(num_unflat_param_ids - 1) + "} but got " + \ f"{unflat_param_ids_set}" return [ unflat_to_flat_param_ids[unflat_param_id] for unflat_param_id in range(num_unflat_param_ids) ] def _is_zero_dim_tensor(x: Any) -> bool: return torch.is_tensor(x) and x.dim() == 0