squash commit

lerobot/common/datasets/factory.py

@@ -1,15 +1,12 @@
 import logging
 
 import torch
-from omegaconf import OmegaConf
+from omegaconf import DictConfig, OmegaConf
 
 from lerobot.common.datasets.lerobot_dataset import LeRobotDataset
 
 
-def make_dataset(
-    cfg,
-    split="train",
-):
+def make_dataset(cfg: DictConfig, split="train") -> LeRobotDataset:
     if cfg.env.name not in cfg.dataset_repo_id:
         logging.warning(
             f"There might be a mismatch between your training dataset ({cfg.dataset_repo_id=}) and your "
@@ -28,6 +25,7 @@ def make_dataset(
         cfg.dataset_repo_id,
         split=split,
         delta_timestamps=delta_timestamps,
+        n_end_keyframes_dropped=eval(cfg.training.get("n_end_keyframes_dropped", "0")),
     )
 
     if cfg.get("override_dataset_stats"):

lerobot/common/datasets/lerobot_dataset.py

@@ -27,7 +27,26 @@ class LeRobotDataset(torch.utils.data.Dataset):
         split: str = "train",
         transform: callable = None,
         delta_timestamps: dict[list[float]] | None = None,
+        n_end_keyframes_dropped: int = 0,
     ):
+        """
+        Args:
+            delta_timestamps: A dictionary mapping lists of relative times (Δt) to data keys. When a frame is
+                sampled from the underlying dataset, we treat it as a "keyframe" and load multiple frames
+                according to the list of Δt's. For example {"action": [-0.05, 0, 0.05]} indicates
+                that we want to load the current keyframe's action, as well as one from 50 ms ago, and one
+                50 ms into the future. The action key then contains a (3, action_dim) tensor (whereas without
+                `delta_timestamps` there would just be a (action_dim,) tensor. When the Δt's demand that
+                frames outside of an episode boundary are retrieved, a copy padding strategy is used. See
+                `load_previous_and_future_frames` for more details.
+            n_end_keyframes_dropped: Don't sample the last n items in each episode. This option is handy when
+                used in combination with `delta_timestamps` when, for example, the Δt's demand multiple future
+                frames, but we want to avoid introducing too much copy padding into the data distribution.
+                For example if `delta_timestamps = {"action": [0, 0.05, 0.10, 0.15, 0.20, 0.25, 0.30]}`
+                and we sample the last frame in the episode, we would end up padding with 6 frames worth of
+                copies. Instead, we might want no padding (in which case we need n=6), or we might be okay
+                with up to 2 frames of padding (in which case we need n=4).
+        """
         super().__init__()
         self.repo_id = repo_id
         self.version = version
@@ -44,6 +63,12 @@ class LeRobotDataset(torch.utils.data.Dataset):
         self.info = load_info(repo_id, version, root)
         if self.video:
             self.videos_dir = load_videos(repo_id, version, root)
+        # If `n_end_keyframes_dropped == 0`, `self.index` contains exactly the indices of the hf_dataset. If
+        # `n_end_keyframes_dropped > 0`, `self.index` contains a subset of the indices of the hf_dataset where
+        # we drop those indices pertaining to the last n frames of each episode.
+        self.index = []
+        for from_ix, to_ix in zip(*self.episode_data_index.values(), strict=True):
+            self.index.extend(list(range(from_ix, to_ix - n_end_keyframes_dropped)))
 
     @property
     def fps(self) -> int:
@@ -78,7 +103,7 @@ class LeRobotDataset(torch.utils.data.Dataset):
 
     @property
     def num_samples(self) -> int:
-        return len(self.hf_dataset)
+        return len(self.index)
 
     @property
     def num_episodes(self) -> int:
@@ -97,7 +122,7 @@ class LeRobotDataset(torch.utils.data.Dataset):
         return self.num_samples
 
     def __getitem__(self, idx):
-        item = self.hf_dataset[idx]
+        item = self.hf_dataset[self.index[idx]]
 
         if self.delta_timestamps is not None:
             item = load_previous_and_future_frames(

lerobot/configs/policy/diffusion.yaml

@@ -25,11 +25,21 @@ training:
   adam_weight_decay: 1.0e-6
   online_steps_between_rollouts: 1
 
+  # For each training batch we want (consider n_obs_steps=2, horizon=16):
+  # t           | -1,  0,  1,  2,  3,  4,  5,  6,  7,  8,  9,  10, 11, 12, 13, 14
+  # action      |  a,  a,  a,  a,  a,  a,  a,  a,  a,  a,  a,   a,  a,  a,  a,  a
+  # observation |  o,  o,   ,   ,   ,   ,   ,   ,   ,   ,   ,    ,   ,   ,   ,
+  # Note that at rollout we only use some of the actions (consider n_action_steps=8):
+  # action used |   ,  a,  a,  a,  a,  a,  a,  a,  a,   ,   ,    ,   ,   ,   ,
   delta_timestamps:
     observation.image: "[i / ${fps} for i in range(1 - ${policy.n_obs_steps}, 1)]"
     observation.state: "[i / ${fps} for i in range(1 - ${policy.n_obs_steps}, 1)]"
     action: "[i / ${fps} for i in range(1 - ${policy.n_obs_steps}, 1 - ${policy.n_obs_steps} + ${policy.horizon})]"
 
+  # The original implementation doesn't sample keyframes for the last 7 steps. This is because, as described
+  # above, the last 7 actions from the diffusion model are not used.
+  n_end_keyframes_dropped: ${policy.horizon} - ${policy.n_action_steps} - ${policy.n_obs_steps} + 1
+
 eval:
   n_episodes: 50
   batch_size: 50

tests/test_datasets.py

@@ -100,6 +100,7 @@ def test_compute_stats_on_xarm():
 
     # reduce size of dataset sample on which stats compute is tested to 10 frames
     dataset.hf_dataset = dataset.hf_dataset.select(range(10))
+    dataset.index = [i for i in dataset.index if i < 10]
 
     # Note: we set the batch size to be smaller than the whole dataset to make sure we are testing batched
     # computation of the statistics. While doing this, we also make sure it works when we don't divide the