VD4RLExperienceReplay¶
- class torchrl.data.datasets.VD4RLExperienceReplay(dataset_id, batch_size: int, *, root: str | Path | None = None, download: bool | str = True, sampler: Sampler | None = None, writer: Writer | None = None, collate_fn: Callable | None =None, pin_memory: bool =False, prefetch: int | None =None, transform: 'torchrl.envs.Transform' | None =None, split_trajs: bool =False, totensor: bool =True, image_size: int | List[int] | None = None, num_workers: int =0, **env_kwargs)[source]¶
V-D4RL 經驗回放資料集。
此類從 V-D4RL 下載 H5/npz 資料,並將其處理為 mmap 格式,從而加快索引(及取樣)速度。
在此處瞭解更多關於 V-D4RL 的資訊:https://arxiv.org/abs/2206.04779
“pixels” 條目位於資料的根部,所有非獎勵、完成狀態、動作或畫素的資料都移動到 “state” 節點下。
資料格式遵循 TED 約定。
- 引數:
dataset_id (str) – 要下載的資料集。必須是 VD4RLExperienceReplay.available_datasets 的一部分。
batch_size (int) – 取樣時使用的批大小。如有必要,可透過 data.sample(batch_size) 覆蓋。
- 關鍵字引數:
root (Path or str, optional) – V-D4RL 資料集根目錄。實際的資料集記憶體對映檔案將儲存在 <root>/<dataset_id> 下。如果未提供,預設為 ~/.cache/torchrl/atari.vd4rl`。
download (bool or str, optional) – 如果未找到資料集是否應該下載。預設為
True。下載也可傳遞為"force",在此情況下,下載的資料將被覆蓋。sampler (Sampler, optional) – 要使用的取樣器。如果未提供,將使用預設的 RandomSampler()。
writer (Writer, optional) – 要使用的寫入器。如果未提供,將使用預設的
ImmutableDatasetWriter。collate_fn (callable, optional) – 合併樣本列表以形成 Tensor(s)/輸出的小批次。在從 map 風格資料集進行批次載入時使用。
pin_memory (bool) – 是否應在回放緩衝區樣本上呼叫 pin_memory()。
prefetch (int, optional) – 使用多執行緒預取的下一批次數量。
transform (Transform, optional) – 呼叫 sample() 時要執行的 Transform。要鏈式應用 transforms,請使用
Compose類。split_trajs (bool, optional) – 如果為
True,軌跡將沿第一個維度分割並填充以具有匹配的形狀。要分割軌跡,將使用"done"訊號,該訊號透過done = truncated | terminated恢復。換句話說,假定任何truncated或terminated訊號等同於軌跡的結束。對於D4RL中的某些資料集,這可能不成立。使用者需就split_trajs的此用法做出準確選擇。預設為False。totensor (bool, optional) – 如果為
True,ToTensorImage轉換將包含在轉換列表(如果未自動檢測)中。預設為True。image_size (int, list of ints or None) – 如果不是
None,此引數將用於建立一個Resize轉換,該轉換將附加到轉換列表。支援 int 型別(方形縮放)或 int 的列表/元組(矩形縮放)。預設為None(不縮放)。num_workers (int, optional) – 下載檔案的 worker 數量。預設為
0(無多程序)。
- 變數:
available_datasets – 可接受下載的條目列表。這些名稱對應於 huggingface 資料集倉庫中的目錄路徑。如果可能,列表將從 huggingface 動態檢索。如果無網際網路連線,將使用快取版本。
注意
由於並非所有經驗回放都包含開始和停止訊號,因此我們在檢索到的資料集中不標記 episode。
示例
>>> import torch >>> torch.manual_seed(0) >>> from torchrl.data.datasets import VD4RLExperienceReplay >>> d = VD4RLExperienceReplay("main/walker_walk/random/64px", batch_size=32, ... image_size=50) >>> for batch in d: ... break >>> print(batch) TensorDict( fields={ action: Tensor(shape=torch.Size([32, 6]), device=cpu, dtype=torch.float32, is_shared=False), done: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.bool, is_shared=False), index: Tensor(shape=torch.Size([32]), device=cpu, dtype=torch.int64, is_shared=False), is_init: Tensor(shape=torch.Size([32]), device=cpu, dtype=torch.bool, is_shared=False), next: TensorDict( fields={ done: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.bool, is_shared=False), observation: TensorDict( fields={ height: Tensor(shape=torch.Size([32]), device=cpu, dtype=torch.float32, is_shared=False), orientations: Tensor(shape=torch.Size([32, 14]), device=cpu, dtype=torch.float32, is_shared=False), velocity: Tensor(shape=torch.Size([32, 9]), device=cpu, dtype=torch.float32, is_shared=False)}, batch_size=torch.Size([32]), device=cpu, is_shared=False), pixels: Tensor(shape=torch.Size([32, 3, 50, 50]), device=cpu, dtype=torch.float32, is_shared=False), reward: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.float32, is_shared=False), terminated: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.bool, is_shared=False), truncated: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.bool, is_shared=False)}, batch_size=torch.Size([32]), device=cpu, is_shared=False), observation: TensorDict( fields={ height: Tensor(shape=torch.Size([32]), device=cpu, dtype=torch.float32, is_shared=False), orientations: Tensor(shape=torch.Size([32, 14]), device=cpu, dtype=torch.float32, is_shared=False), velocity: Tensor(shape=torch.Size([32, 9]), device=cpu, dtype=torch.float32, is_shared=False)}, batch_size=torch.Size([32]), device=cpu, is_shared=False), pixels: Tensor(shape=torch.Size([32, 3, 50, 50]), device=cpu, dtype=torch.float32, is_shared=False), terminated: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.bool, is_shared=False), truncated: Tensor(shape=torch.Size([32, 1]), device=cpu, dtype=torch.bool, is_shared=False)}, batch_size=torch.Size([32]), device=cpu, is_shared=False)
- add(data: TensorDictBase) int¶
向回放緩衝區新增單個元素。
- 引數:
data (Any) – 要新增到回放緩衝區的資料
- 返回:
資料在回放緩衝區中的索引。
- append_transform(transform: Transform, *, invert: bool =False) ReplayBuffer¶
在末尾追加 transform。
呼叫 sample 時按順序應用 transforms。
- 引數:
transform (Transform) – 要追加的 transform
- 關鍵字引數:
invert (bool, optional) – 如果為
True,transform 將被反轉(寫入時呼叫 forward,讀取時呼叫 inverse)。預設為False。
示例
>>> rb = ReplayBuffer(storage=LazyMemmapStorage(10), batch_size=4) >>> data = TensorDict({"a": torch.zeros(10)}, [10]) >>> def t(data): ... data += 1 ... return data >>> rb.append_transform(t, invert=True) >>> rb.extend(data) >>> assert (data == 1).all()
- property data_path¶
資料集路徑,包含分割資訊。
- property data_path_root¶
資料集根路徑。
- delete()¶
從磁碟刪除資料集儲存。
- dumps(path)¶
將回放緩衝區儲存到磁碟上的指定路徑。
- 引數:
path (Path or str) – 儲存回放緩衝區的路徑。
示例
>>> import tempfile >>> import tqdm >>> from torchrl.data import LazyMemmapStorage, TensorDictReplayBuffer >>> from torchrl.data.replay_buffers.samplers import PrioritizedSampler, RandomSampler >>> import torch >>> from tensordict import TensorDict >>> # Build and populate the replay buffer >>> S = 1_000_000 >>> sampler = PrioritizedSampler(S, 1.1, 1.0) >>> # sampler = RandomSampler() >>> storage = LazyMemmapStorage(S) >>> rb = TensorDictReplayBuffer(storage=storage, sampler=sampler) >>> >>> for _ in tqdm.tqdm(range(100)): ... td = TensorDict({"obs": torch.randn(100, 3, 4), "next": {"obs": torch.randn(100, 3, 4)}, "td_error": torch.rand(100)}, [100]) ... rb.extend(td) ... sample = rb.sample(32) ... rb.update_tensordict_priority(sample) >>> # save and load the buffer >>> with tempfile.TemporaryDirectory() as tmpdir: ... rb.dumps(tmpdir) ... ... sampler = PrioritizedSampler(S, 1.1, 1.0) ... # sampler = RandomSampler() ... storage = LazyMemmapStorage(S) ... rb_load = TensorDictReplayBuffer(storage=storage, sampler=sampler) ... rb_load.loads(tmpdir) ... assert len(rb) == len(rb_load)
- empty()¶
清空回放緩衝區並將游標重置為 0。
- extend(tensordicts: TensorDictBase) Tensor¶
使用 iterable 中包含的一個或多個元素擴充套件回放緩衝區。
如果存在,將呼叫 inverse transforms。`
- 引數:
data (iterable) – 要新增到回放緩衝區的資料集合。
- 返回:
新增到回放緩衝區的資料索引。
警告
extend()在處理值列表時簽名可能存在歧義,列表應被解釋為 PyTree(在這種情況下,列表中的所有元素將被放入儲存中 PyTree 的一個切片中),或者被解釋為要逐個新增的值列表。為了解決這個問題,TorchRL 明確區分了 list 和 tuple:tuple 將被視為一個 PyTree,list(在根級別)將被解釋為要逐個新增到緩衝區的值堆疊。對於ListStorage例項,只能提供未繫結的元素(不能是 PyTrees)。
- insert_transform(index: int, transform: Transform, *, invert: bool =False) ReplayBuffer¶
插入 transform。
呼叫 sample 時按順序執行 transforms。
- 引數:
index (int) – 插入 transform 的位置。
transform (Transform) – 要追加的 transform
- 關鍵字引數:
invert (bool, optional) – 如果為
True,transform 將被反轉(寫入時呼叫 forward,讀取時呼叫 inverse)。預設為False。
- loads(path)¶
載入給定路徑的回放緩衝區狀態。
緩衝區應具有匹配的元件,並使用
dumps()儲存。- 引數:
path (Path or str) – 回放緩衝區儲存的路徑。
有關更多資訊,請參閱
dumps()。
- preprocess(fn: Callable[[TensorDictBase], TensorDictBase], dim: int =0, num_workers: int | None =None, *, chunksize: int | None =None, num_chunks: int | None =None, pool: mp.Pool | None =None, generator: torch.Generator | None =None, max_tasks_per_child: int | None =None, worker_threads: int =1, index_with_generator: bool =False, pbar: bool =False, mp_start_method: str | None =None, num_frames: int | None =None, dest: str | Path) TensorStorage¶
預處理資料集並返回帶有格式化資料的新儲存。
資料轉換必須是單元的(作用於資料集的單個樣本)。
Args 和 Keyword Args 將轉發給
map()。隨後可以使用
delete()刪除資料集。- 關鍵字引數:
dest (path or equivalent) – 新資料集位置的路徑。
num_frames (int, optional) – 如果提供,將僅轉換前 num_frames。這對於初步除錯 transform 很有用。
返回:一個可用於
ReplayBuffer例項的新儲存。示例
>>> from torchrl.data.datasets import MinariExperienceReplay >>> >>> data = MinariExperienceReplay( ... list(MinariExperienceReplay.available_datasets)[0], ... batch_size=32 ... ) >>> print(data) MinariExperienceReplay( storages=TensorStorage(TensorDict( fields={ action: MemoryMappedTensor(shape=torch.Size([1000000, 8]), device=cpu, dtype=torch.float32, is_shared=True), episode: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.int64, is_shared=True), info: TensorDict( fields={ distance_from_origin: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), forward_reward: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), goal: MemoryMappedTensor(shape=torch.Size([1000000, 2]), device=cpu, dtype=torch.float64, is_shared=True), qpos: MemoryMappedTensor(shape=torch.Size([1000000, 15]), device=cpu, dtype=torch.float64, is_shared=True), qvel: MemoryMappedTensor(shape=torch.Size([1000000, 14]), device=cpu, dtype=torch.float64, is_shared=True), reward_ctrl: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), reward_forward: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), reward_survive: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), success: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.bool, is_shared=True), x_position: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), x_velocity: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), y_position: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), y_velocity: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), next: TensorDict( fields={ done: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.bool, is_shared=True), info: TensorDict( fields={ distance_from_origin: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), forward_reward: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), goal: MemoryMappedTensor(shape=torch.Size([1000000, 2]), device=cpu, dtype=torch.float64, is_shared=True), qpos: MemoryMappedTensor(shape=torch.Size([1000000, 15]), device=cpu, dtype=torch.float64, is_shared=True), qvel: MemoryMappedTensor(shape=torch.Size([1000000, 14]), device=cpu, dtype=torch.float64, is_shared=True), reward_ctrl: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), reward_forward: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), reward_survive: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), success: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.bool, is_shared=True), x_position: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), x_velocity: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), y_position: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True), y_velocity: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.float64, is_shared=True)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), observation: TensorDict( fields={ achieved_goal: MemoryMappedTensor(shape=torch.Size([1000000, 2]), device=cpu, dtype=torch.float64, is_shared=True), desired_goal: MemoryMappedTensor(shape=torch.Size([1000000, 2]), device=cpu, dtype=torch.float64, is_shared=True), observation: MemoryMappedTensor(shape=torch.Size([1000000, 27]), device=cpu, dtype=torch.float64, is_shared=True)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), reward: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.float64, is_shared=True), terminated: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.bool, is_shared=True), truncated: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.bool, is_shared=True)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), observation: TensorDict( fields={ achieved_goal: MemoryMappedTensor(shape=torch.Size([1000000, 2]), device=cpu, dtype=torch.float64, is_shared=True), desired_goal: MemoryMappedTensor(shape=torch.Size([1000000, 2]), device=cpu, dtype=torch.float64, is_shared=True), observation: MemoryMappedTensor(shape=torch.Size([1000000, 27]), device=cpu, dtype=torch.float64, is_shared=True)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False)), samplers=RandomSampler, writers=ImmutableDatasetWriter(), batch_size=32, transform=Compose( ), collate_fn=<function _collate_id at 0x120e21dc0>) >>> from torchrl.envs import CatTensors, Compose >>> from tempfile import TemporaryDirectory >>> >>> cat_tensors = CatTensors( ... in_keys=[("observation", "observation"), ("observation", "achieved_goal"), ... ("observation", "desired_goal")], ... out_key="obs" ... ) >>> cat_next_tensors = CatTensors( ... in_keys=[("next", "observation", "observation"), ... ("next", "observation", "achieved_goal"), ... ("next", "observation", "desired_goal")], ... out_key=("next", "obs") ... ) >>> t = Compose(cat_tensors, cat_next_tensors) >>> >>> def func(td): ... td = td.select( ... "action", ... "episode", ... ("next", "done"), ... ("next", "observation"), ... ("next", "reward"), ... ("next", "terminated"), ... ("next", "truncated"), ... "observation" ... ) ... td = t(td) ... return td >>> with TemporaryDirectory() as tmpdir: ... new_storage = data.preprocess(func, num_workers=4, pbar=True, mp_start_method="fork", dest=tmpdir) ... rb = ReplayBuffer(storage=new_storage) ... print(rb) ReplayBuffer( storage=TensorStorage( data=TensorDict( fields={ action: MemoryMappedTensor(shape=torch.Size([1000000, 8]), device=cpu, dtype=torch.float32, is_shared=True), episode: MemoryMappedTensor(shape=torch.Size([1000000]), device=cpu, dtype=torch.int64, is_shared=True), next: TensorDict( fields={ done: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.bool, is_shared=True), obs: MemoryMappedTensor(shape=torch.Size([1000000, 31]), device=cpu, dtype=torch.float64, is_shared=True), observation: TensorDict( fields={ }, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), reward: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.float64, is_shared=True), terminated: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.bool, is_shared=True), truncated: MemoryMappedTensor(shape=torch.Size([1000000, 1]), device=cpu, dtype=torch.bool, is_shared=True)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), obs: MemoryMappedTensor(shape=torch.Size([1000000, 31]), device=cpu, dtype=torch.float64, is_shared=True), observation: TensorDict( fields={ }, batch_size=torch.Size([1000000]), device=cpu, is_shared=False)}, batch_size=torch.Size([1000000]), device=cpu, is_shared=False), shape=torch.Size([1000000]), len=1000000, max_size=1000000), sampler=RandomSampler(), writer=RoundRobinWriter(cursor=0, full_storage=True), batch_size=None, collate_fn=<function _collate_id at 0x168406fc0>)
- register_load_hook(hook: Callable[[Any], Any])¶
為儲存註冊一個載入鉤子。
注意
目前,儲存回放緩衝區時不會序列化鉤子:每次建立緩衝區時,必須手動重新初始化它們。
- register_save_hook(hook: Callable[[Any], Any])¶
為儲存註冊一個儲存鉤子。
注意
目前,儲存回放緩衝區時不會序列化鉤子:每次建立緩衝區時,必須手動重新初始化它們。
- sample(batch_size: int | None = None, return_info: bool = False, include_info: bool = None) TensorDictBase¶
從回放緩衝區中取樣一批資料。
使用 Sampler 取樣索引,並從 Storage 中檢索資料。
- 引數:
batch_size (int, 可選) – 要收集的資料大小。如果未提供,此方法將按照 Sampler 指定的批大小進行取樣。
return_info (bool) – 是否返回附加資訊(info)。如果為 True,結果是一個元組 (data, info)。如果為 False,結果僅為 data。
- 返回:
一個包含從回放緩衝區中選擇的一批資料的 tensordict。如果設定了 return_info 標誌為 True,則返回一個包含此 tensordict 和 info 的元組。
- set_storage(storage: Storage, collate_fn: Callable | None = None)¶
在回放緩衝區中設定新的儲存,並返回之前的儲存。
- 引數:
storage (Storage) – 緩衝區的新儲存。
collate_fn (callable, 可選) – 如果提供,collate_fn 將設定為此值。否則,它將重置為預設值。
- property write_count¶
透過 add 和 extend 方法寫入緩衝區的資料項總數。