Mj@&ddlZddlZddlmZddlmZddlmZmZm Z ddl m Z m Z m Z mZddlmZgdZed Gd d ZGd d ej&ZGddej&Zy)N) dataclass)Any)Metadata MetadataIndex StorageMeta)LoadPlan LoadPlannerSavePlan SavePlanner)Future) WriteResult StorageWriter StorageReaderT)frozenc,eZdZUeed<eed<eed<y)r index size_in_bytes storage_dataN)__name__ __module__ __qualname__r__annotations__intrc/var/www/html/ia-enhanced/venv/lib/python3.12/site-packages/torch/distributed/checkpoint/storage.pyr r s rr c eZdZdZej ddeejzdzddfdZ ej de de de ddfd Z ej d e de fd Zej d ee dee fd Zej d e dedeeefdZej dedeeeddfdZeej deejzde fdZdedzfdZy)ra8 Interface used by ``save_state_dict`` to write to storage. One StorageWriter instance acts as both the coordinator and the follower in a distributed checkpoint. As part of initialization, each instance is told its role. A subclass should expect the following sequence of calls. 0) (all ranks) set checkpoint_id if users pass a valid checkpoint_id. 1) (all ranks) set_up_storage_writer() 2) (all ranks) prepare_local_plan() 3) (coordinator) prepare_global_plan() 4) (all ranks) write_data() 5) (coordinator) finish() N checkpoint_idreturncy)a Calls to indicates a brand new checkpoint write is going to happen. A checkpoint_id may be present if users set the checkpoint_id for this checkpoint write. The meaning of the checkpiont_id is storage-dependent. It can be a path to a folder/file or a key for a key-value storage. Args: checkpoint_id (Union[str, os.PathLike, None]): The ID of this checkpoint instance. The meaning of the checkpoint_id depends on the storage. It can be a path to a folder or to a file. It can also be a key if the storage is a key-value store. (Default: ``None``) Nrselfrs rresetzStorageWriter.reset- ris_coordinatorargskwargscy)z Initialize this instance. Args: is_coordinator (bool): Whether this instance is responsible for coordinating the checkpoint. Nr)r"r%r&r's rset_up_storage_writerz#StorageWriter.set_up_storage_writer?rplancy)a Perform storage-specific local planning. While this method can produce a completely different plan, the recommended way is to store storage specific data in SavePlan::storage_data. Args: plan (SavePlan): The local plan from the ``SavePlanner`` in use. Returns: A transformed ``SavePlan`` after storage local planning Nrr"r+s rprepare_local_planz StorageWriter.prepare_local_planKr*rplanscy)a Perform centralized planning of storage. This method is only called on the coordinator instance. While this method can produce a completely different plan, the preferred way is to store storage specific data in SavePlan::storage_data. Args: plans: A list of ``SavePlan`` instances, one for each rank. Returns: A list of transformed ``SavePlan`` after storage global planning Nrr"r/s rprepare_global_planz!StorageWriter.prepare_global_planZr*rplannercy)a Write all items from ``plan`` using ``planner`` to resolve the data. A subclass should call ``SavePlanner::resolve_data`` on each item from the plan to get access to the underlying object to write. Subclasses should lazily call `resolve_data` as it can allocate memory. In case of tensors, make following assumptions: - They might be on any device, including not matching the one on ``WriteItem::tensor_data`` - They might be views or not contiguous. Only the projection needs to be saved. Args: plan (SavePlan): The save plan to execute. planner (SavePlanner): Planner object to be used to resolve items to data. Returns: A future that completes to a list of WriteResult Nrr"r+r3s r write_datazStorageWriter.write_datakr*rmetadataresultscy)a Write the metadata and marks the current checkpoint as successful. The actual format/schema used for serializing `metadata` is an implementation detail. The only requirement is that it's recoverable in to the same object graph. Args: metadata (Metadata): metadata for the new checkpoint results: A list of WriteResults from all ranks. Returns: None Nr)r"r7r8s rfinishzStorageWriter.finishr*rcyz Check if the given checkpoint_id is supported by the storage. This allow us to enable automatic storage selection. Nrclsrs rvalidate_checkpoint_idz$StorageWriter.validate_checkpoint_id rcy)a= Return the storage-specific metadata. This is used to store additional information in a checkpoint that can be useful for providing request-level observability. StorageMeta is passed to the ``SavePlanner`` during save calls. Returns None by default. TODO: provide an example Nr)r"s r storage_metazStorageWriter.storage_metasrN)rrr__doc__abcabstractmethodstrosPathLiker#boolrr)r r.listr2r r r r6rr: classmethodr?rrBrrrrrs"  3#4t#; t  "   "  +.  :=         x  H      h DN     '2 [! "  .  x $tK7H2I d   33D   kD0rrc eZdZdZej ddeejzdzddfdZ ej de de de fdZ ej d e d e de de ddf d Zej d edefd Zej deedeefdZej d edededfdZeej deejzde fdZy)raV Interface used by ``load_state_dict`` to read from storage. One StorageReader instance acts as both the coordinator and the follower in a distributed checkpoint. As part of initialization, each instance is told its role. A subclass should expected the following sequence of calls by ``load_state_dict``: 0) (all ranks) set checkpoint_id if users pass a valid checkpoint_id. 1) (all ranks) read_metadata() 2) (all ranks) set_up_storage_reader() 3) (all ranks) prepare_local_plan() 4) (coordinator) prepare_global_plan() 5) (all ranks) read_data() Nrrcy)a Calls to indicates a brand new checkpoint read is going to happen. A checkpoint_id may be present if users set the checkpoint_id for this checkpoint read. The meaning of the checkpiont_id is storage-dependent. It can be a path to a folder/file or a key for a key-value storage. Args: checkpoint_id (Union[str, os.PathLike, None]): The ID of this checkpoint instance. The meaning of the checkpoint_id depends on the storage. It can be a path to a folder or to a file. It can also be a key if the storage is more like a key-value store. (Default: ``None``) Nrr!s rr#zStorageReader.resetr$rr&r'cy)z Read the checkpoint metadata. Returns: The metadata object associated with the checkpoint being loaded. Nr)r"r&r's r read_metadatazStorageReader.read_metadatar*rr7r%cy)z Initialize this instance. Args: metadata (Metadata): The metadata schema to use. is_coordinator (bool): Whether this instance is responsible for coordinating the checkpoint. Nr)r"r7r%r&r's rset_up_storage_readerz#StorageReader.set_up_storage_readerr*rr+cy)a Perform storage-specific local planning. While this method can produce a completely different plan, the recommended way is to store storage specific data in LoadPlan::storage_data. Args: plan (LoadPlan): The local plan from the ``LoadPlan`` in use. Returns: A transformed ``LoadPlan`` after storage local planning Nrr-s rr.z StorageReader.prepare_local_planr*rr/cy)a Perform centralized planning of storage loading. This method is only called on the coordinator instance. While this method can produce a completely different plan, the preferred way is to store storage specific data in LoadPlan::storage_data. Args: plans: A list of ``LoadPlan`` instances, one for each rank. Returns: A list of transformed ``LoadPlan`` after storage global planning Nrr1s rr2z!StorageReader.prepare_global_planr*rr3cy)a Read all items from ``plan`` using ``planner`` to resolve the data. A subclass should call ``LoadPlanner::load_bytes`` to deserialize a BytesIO object into the right place. A subclass should call ``LoadPlanner::resolve_tensor`` to get access to the tensors that in should load data into. It's the StorageLayer responsibility to properly schedule any cross device copies required. Args: plan (LoadPlan): The local plan to execute on planner (LoadPlanner): The planner object to use to resolve items. Returns: A future that completes once all reads are finished. Nrr5s r read_datazStorageReader.read_datar*rcyr<rr=s rr?z$StorageReader.validate_checkpoint_idr@rrC)rrrrDrErFrGrHrIr#rrrPrJrRrr.rKr2r r rVrLr?rrrrrsr"  3#4t#; t  "  3 # (        26  ?B  NQ         x  H      h DN    h    * 33D   rr)rErH dataclassesrtypingr%torch.distributed.checkpoint.metadatarrr$torch.distributed.checkpoint.plannerrr r r torch.futuresr __all__r ABCrrrrrr_sp !VV ! < $JCGGJZx CGGx r