GenDGRLExperienceReplay¶
- class torchrl.data.datasets.GenDGRLExperienceReplay(dataset_id: str, batch_size: int = None, *, download: bool = True, root: str | None = None, **kwargs)[source]¶
Gen-DGRL 经验回放数据集。
此数据集与论文“离线强化学习中的泛化差距”一起提供。
Arxiv:https://arxiv.org/abs/2312.05742
GitHub:https://github.com/facebookresearch/gen_dgrl
数据格式遵循TED 约定。
此类使您可以访问 ProcGen 数据集。在GenDGRLExperienceReplay.available_datasets中注册的每个dataset_id都包含一个特定任务(“bigfish”、“bossfight”、…),它由一个逗号(“bigfish-1M_E”、…)与类别(“1M_E”、“1M_S”、…)隔开。
在下载和准备过程中,数据将作为 .tar 文件下载,其中每条轨迹都独立存储在一个 .npy 文件中。这些文件中的每一个都被提取、写入连续的内存映射张量,然后被清除。此过程可能每个数据集需要几分钟。在集群上,建议首先在不同的工作程序或进程上为不同的数据集分别运行下载和预处理,然后在第二次运行训练脚本。
- 参数:
dataset_id (str) – 要下载的数据集。必须是
GenDGRLExperienceReplay.available_datasets
的一部分。batch_size (int, 可选) – 采样期间使用的批次大小。如果需要,可以通过data.sample(batch_size) 覆盖。
- 关键字参数:
root (Path 或 str, 可选) –
GenDGRLExperienceReplay
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。如果没有提供,则默认为 ``~/.cache/torchrl/gen_dgrl`。download (bool 或 str, 可选) – 如果找不到数据集,是否应下载数据集。默认为
True
。下载也可以作为"force"
传递,在这种情况下,下载的数据将被覆盖。sampler (Sampler, 可选) – 要使用的采样器。如果没有提供,将使用默认的 RandomSampler()。
writer (Writer, 可选) – 要使用的写入器。如果没有提供,将使用默认的 RoundRobinWriter()。
collate_fn (callable, 可选) – 将样本列表合并以形成张量(s)/输出的小批量。在使用来自映射样式数据集的批量加载时使用。
pin_memory (bool) – 是否应在 rb 样本上调用 pin_memory()。
prefetch (int, 可选) – 要使用多线程预取的下一个批次的数目。
transform (Transform, 可选) – 调用 sample() 时要执行的转换。要链接转换,请使用
Compose
类。
- 变量:
available_datasets – 可以下载的接受项列表。这些名称对应于 huggingface 数据集存储库中的目录路径。如果可能,该列表将从 huggingface 动态检索。如果无法连接互联网,将使用缓存版本。
示例
>>> import torch >>> torch.manual_seed(0) >>> from torchrl.data.datasets import GenDGRLExperienceReplay >>> d = GenDGRLExperienceReplay("bigfish-1M_E", batch_size=32) >>> for batch in d: ... break >>> print(batch)
- add(data: TensorDictBase) int ¶
将单个元素添加到回放缓冲区。
- 参数:
data (Any) – 要添加到回放缓冲区的数据
- 返回值:
数据在回放缓冲区中的位置索引。
- append_transform(transform: Transform, *, invert: bool = False) ReplayBuffer ¶
将转换附加到末尾。
调用sample时,将按顺序应用转换。
- 参数:
transform (Transform) – 要附加的转换
- 关键字参数:
invert (bool, 可选) – 如果为
True
,则转换将被反转(在写入期间将调用正向调用,在读取期间将调用反向调用)。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¶
数据集的路径,包括拆分。
- property data_path_root¶
数据集根路径。
- delete()¶
从磁盘删除数据集存储。
- dumps(path)¶
将重播缓冲区保存在指定路径的磁盘上。
- 参数:
path (Path 或 str) – 保存重播缓冲区的路径。
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()¶
清空重播缓冲区并将游标重置为 0。
- extend(tensordicts: TensorDictBase) Tensor ¶
使用可迭代对象中包含的一个或多个元素扩展重播缓冲区。
如果存在,将调用逆变换。
- 参数:
data (iterable) – 要添加到重播缓冲区的数据集合。
- 返回值:
添加到重播缓冲区的数据的索引。
警告
extend()
处理值列表时可能具有模棱两可的签名,这些值列表应解释为 PyTree(在这种情况下,列表中的所有元素都将被放入存储的 PyTree 中的切片中)或要逐个添加的值列表。为了解决这个问题,TorchRL 在列表和元组之间做出了明确的区别:元组将被视为 PyTree,列表(在根级别)将被解释为要逐个添加到缓冲区的值的堆栈。对于ListStorage
实例,只能提供未绑定的元素(没有 PyTree)。
- insert_transform(index: int, transform: Transform, *, invert: bool = False) ReplayBuffer ¶
插入变换。
当调用 sample 时,变换按顺序执行。
- 参数:
index (int) – 插入变换的位置。
transform (Transform) – 要附加的转换
- 关键字参数:
invert (bool, 可选) – 如果为
True
,则转换将被反转(在写入期间将调用正向调用,在读取期间将调用反向调用)。默认为False
。
- loads(path)¶
加载给定路径的重播缓冲区状态。
缓冲区应具有匹配的组件,并使用
dumps()
保存。- 参数:
path (Path 或 str) – 保存重播缓冲区的路径。
有关更多信息,请参见
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一个包含格式化数据的新的存储。
数据变换必须是单一的(作用于数据集的单个样本)。
参数和关键字参数将转发到
map()
.随后可以使用
delete()
删除数据集。- 关键字参数:
dest (path 或 equivalent) – 新数据集位置的路径。
num_frames (int, 可选) – 如果提供,将只变换前 num_frames 个。这对于首先调试变换很有用。
返回值:一个要在
ReplayBuffer
实例中使用的新的存储。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:每次创建缓冲区时都必须手动重新初始化它们。
- register_save_hook(hook: Callable[[Any], Any])¶
为存储注册保存钩子。
注意
目前保存回放缓冲区时不会序列化钩子:每次创建缓冲区时都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 中检索它们。
- 参数:
batch_size (int, optional) – 要收集的数据的大小。如果没有提供,此方法将根据采样器采样一批大小。
return_info (bool) – 是否返回信息。如果为 True,则结果为一个元组 (data, info)。如果为 False,则结果为数据。
- 返回值:
包含在回放缓冲区中选择的批数据的 tensordict。如果 return_info 标志设置为 True,则包含此 tensordict 和信息的元组。