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 经验回放数据集。

此类从 V-D4RL 下载 H5/npz 数据，并将其处理为 mmap 格式，从而加快索引（及采样）速度。

在此处了解更多关于 V-D4RL 的信息：https://arxiv.org/abs/2206.04779

“pixels” 条目位于数据的根部，所有非奖励、完成状态、动作或像素的数据都移动到 “state” 节点下。

数据格式遵循 TED 约定。

参数:

dataset_id (str) – 要下载的数据集。必须是 VD4RLExperienceReplay.available_datasets 的一部分。
batch_size (int) – 采样时使用的批大小。如有必要，可通过 data.sample(batch_size) 覆盖。

关键字参数:

root (Path or str, optional) – V-D4RL 数据集根目录。实际的数据集内存映射文件将保存在 <root>/<dataset_id> 下。如果未提供，默认为 ~/.cache/torchrl/atari.vd4rl`。
download (bool or str, optional) – 如果未找到数据集是否应该下载。默认为 True。下载也可传递为 "force"，在此情况下，下载的数据将被覆盖。
sampler (Sampler, optional) – 要使用的采样器。如果未提供，将使用默认的 RandomSampler()。
writer (Writer, optional) – 要使用的写入器。如果未提供，将使用默认的 ImmutableDatasetWriter。
collate_fn (callable, optional) – 合并样本列表以形成 Tensor(s)/输出的小批量。在从 map 风格数据集进行批量加载时使用。
pin_memory (bool) – 是否应在回放缓冲区样本上调用 pin_memory()。
prefetch (int, optional) – 使用多线程预取的下一批次数量。
transform (Transform, optional) – 调用 sample() 时要执行的 Transform。要链式应用 transforms，请使用 Compose 类。
split_trajs (bool, optional) – 如果为 True，轨迹将沿第一个维度分割并填充以具有匹配的形状。要分割轨迹，将使用 "done" 信号，该信号通过 done = truncated | terminated 恢复。换句话说，假定任何 truncated 或 terminated 信号等同于轨迹的结束。对于 D4RL 中的某些数据集，这可能不成立。用户需就 split_trajs 的此用法做出准确选择。默认为 False。
totensor (bool, optional) – 如果为 True，ToTensorImage 转换将包含在转换列表（如果未自动检测）中。默认为 True。
image_size (int, list of ints or None) – 如果不是 None，此参数将用于创建一个 Resize 转换，该转换将附加到转换列表。支持 int 类型（方形缩放）或 int 的列表/元组（矩形缩放）。默认为 None（不缩放）。
num_workers (int, optional) – 下载文件的 worker 数量。默认为 0（无多进程）。

变量:

available_datasets – 可接受下载的条目列表。这些名称对应于 huggingface 数据集仓库中的目录路径。如果可能，列表将从 huggingface 动态检索。如果无互联网连接，将使用缓存版本。

注意

由于并非所有经验回放都包含开始和停止信号，因此我们在检索到的数据集中不标记 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¶

向回放缓冲区添加单个元素。

参数:: data (Any) – 要添加到回放缓冲区的数据
返回:: 数据在回放缓冲区中的索引。

append_transform(transform: Transform, *, invert: bool =False) → ReplayBuffer¶

在末尾追加 transform。

调用 sample 时按顺序应用 transforms。

参数:: transform (Transform) – 要追加的 transform
关键字参数:: invert (bool, optional) – 如果为 True，transform 将被反转（写入时调用 forward，读取时调用 inverse）。默认为 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¶: 数据集路径，包含分割信息。

property data_path_root¶: 数据集根路径。

delete()¶: 从磁盘删除数据集存储。

dump(*args, **kwargs)¶: dumps() 的别名。

dumps(path)¶

将回放缓冲区保存到磁盘上的指定路径。

参数:: path (Path or str) – 保存回放缓冲区的路径。

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()¶: 清空回放缓冲区并将光标重置为 0。

extend(tensordicts: TensorDictBase) → Tensor¶

使用 iterable 中包含的一个或多个元素扩展回放缓冲区。

如果存在，将调用 inverse transforms。`

参数:: data (iterable) – 要添加到回放缓冲区的数据集合。
返回:: 添加到回放缓冲区的数据索引。

警告

extend() 在处理值列表时签名可能存在歧义，列表应被解释为 PyTree（在这种情况下，列表中的所有元素将被放入存储中 PyTree 的一个切片中），或者被解释为要逐个添加的值列表。为了解决这个问题，TorchRL 明确区分了 list 和 tuple：tuple 将被视为一个 PyTree，list（在根级别）将被解释为要逐个添加到缓冲区的值堆栈。对于 ListStorage 实例，只能提供未绑定的元素（不能是 PyTrees）。

insert_transform(index: int, transform: Transform, *, invert: bool =False) → ReplayBuffer¶

插入 transform。

调用 sample 时按顺序执行 transforms。

参数:

index (int) – 插入 transform 的位置。
transform (Transform) – 要追加的 transform

关键字参数:

invert (bool, optional) – 如果为 True，transform 将被反转（写入时调用 forward，读取时调用 inverse）。默认为 False。

load(*args, **kwargs)¶: loads() 的别名。

loads(path)¶

加载给定路径的回放缓冲区状态。

缓冲区应具有匹配的组件，并使用 dumps() 保存。

参数:: path (Path or str) – 回放缓冲区保存的路径。

有关更多信息，请参阅 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¶

预处理数据集并返回带有格式化数据的新存储。

数据转换必须是单元的（作用于数据集的单个样本）。

Args 和 Keyword Args 将转发给 map()。

随后可以使用 delete() 删除数据集。

关键字参数:

dest (path or equivalent) – 新数据集位置的路径。
num_frames (int, optional) – 如果提供，将仅转换前 num_frames。这对于初步调试 transform 很有用。

返回：一个可用于 ReplayBuffer 实例的新存储。

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])¶: 为存储注册一个加载钩子。

注意

目前，保存回放缓冲区时不会序列化钩子：每次创建缓冲区时，必须手动重新初始化它们。

register_save_hook(hook: Callable[[Any], Any])¶: 为存储注册一个保存钩子。

注意

目前，保存回放缓冲区时不会序列化钩子：每次创建缓冲区时，必须手动重新初始化它们。

sample(batch_size: int | None = None, return_info: bool = False, include_info: bool = None) → TensorDictBase¶

从回放缓冲区中采样一批数据。

使用 Sampler 采样索引，并从 Storage 中检索数据。

参数:

batch_size (int, 可选) – 要收集的数据大小。如果未提供，此方法将按照 Sampler 指定的批大小进行采样。
return_info (bool) – 是否返回附加信息（info）。如果为 True，结果是一个元组 (data, info)。如果为 False，结果仅为 data。

返回:

一个包含从回放缓冲区中选择的一批数据的 tensordict。如果设置了 return_info 标志为 True，则返回一个包含此 tensordict 和 info 的元组。

property sampler¶

回放缓冲区的采样器。

采样器必须是 Sampler 的实例。

save(*args, **kwargs)¶: dumps() 的别名。

set_sampler(sampler: Sampler)¶: 在回放缓冲区中设置新的采样器，并返回之前的采样器。

set_storage(storage: Storage, collate_fn: Callable | None = None)¶

在回放缓冲区中设置新的存储，并返回之前的存储。

参数:

storage (Storage) – 缓冲区的新存储。
collate_fn (callable, 可选) – 如果提供，collate_fn 将设置为此值。否则，它将重置为默认值。

set_writer(writer: Writer)¶: 在回放缓冲区中设置新的写入器，并返回之前的写入器。

property storage¶

回放缓冲区的存储。

存储必须是 Storage 的实例。

property write_count¶: 通过 add 和 extend 方法写入缓冲区的数据项总数。

property writer¶

回放缓冲区的写入器。

写入器必须是 Writer 的实例。

VD4RLExperienceReplay¶

文档

教程

资源