fix(ci): run VLABench tasks 5-at-a-time in parallel

The eval script already supports running multiple tasks concurrently via
a ThreadPoolExecutor (env.max_parallel_tasks). Set it to 5 so the 10
VLABench tasks finish in ~2 waves instead of running sequentially.

Co-Authored-By: Claude Opus 4.7 (1M context) <noreply@anthropic.com>

.github/workflows/benchmark_tests.yml

@@ -382,7 +382,6 @@ jobs:
                 --policy.path=\"\$ROBOTWIN_POLICY\" \
                 --env.type=robotwin \
                 --env.task=\"\$ROBOTWIN_TASKS\" \
-                --env.max_parallel_tasks=5 \
                 --eval.batch_size=1 \
                 --eval.n_episodes=1 \
                 --eval.use_async_envs=false \
@@ -483,7 +482,6 @@ jobs:
                 --policy.path=lerobot/smolvla_robocasa \
                 --env.type=robocasa \
                 --env.task=CloseFridge,OpenCabinet,OpenDrawer,TurnOnMicrowave,TurnOffStove,CloseToasterOvenDoor,SlideDishwasherRack,TurnOnSinkFaucet,NavigateKitchen,TurnOnElectricKettle \
-                --env.max_parallel_tasks=5 \
                 --eval.batch_size=1 \
                 --eval.n_episodes=1 \
                 --eval.use_async_envs=false \
@@ -695,7 +693,6 @@ jobs:
                 --env.task=\"\$ROBOMME_TASKS\" \
                 --env.dataset_split=test \
                 --env.task_ids=[0] \
-                --env.max_parallel_tasks=5 \
                 --eval.batch_size=1 \
                 --eval.n_episodes=1 \
                 --eval.use_async_envs=false \
@@ -803,7 +800,6 @@ jobs:
                 --env.type=libero_plus \
                 --env.task=\"\$LIBERO_PLUS_SUITE\" \
                 --env.task_ids=\"\$LIBERO_PLUS_TASK_IDS\" \
-                --env.max_parallel_tasks=5 \
                 --eval.batch_size=1 \
                 --eval.n_episodes=1 \
                 --eval.use_async_envs=false \

src/lerobot/configs/train.py

@@ -256,9 +256,7 @@ class TrainPipelineConfig(HubMixin):
                 ) from e
 
         cli_args = kwargs.pop("cli_args", [])
-        # Legacy RA-BC migration only applies to framework-saved checkpoints (always JSON).
-        # Hand-written YAML/TOML configs are expected to use the current sample_weighting schema.
-        if config_file is not None and config_file.endswith(".json"):
+        if config_file is not None:
             with open(config_file) as f:
                 config = json.load(f)
             migrated_config = _migrate_legacy_rabc_fields(config)

src/lerobot/policies/diffusion/configuration_diffusion.py

@@ -100,8 +100,8 @@ class DiffusionConfig(PreTrainedConfig):
 
     # Inputs / output structure.
     n_obs_steps: int = 2
-    horizon: int = 64
-    n_action_steps: int = 32
+    horizon: int = 16
+    n_action_steps: int = 8
 
     normalization_mapping: dict[str, NormalizationMode] = field(
         default_factory=lambda: {
@@ -122,10 +122,10 @@ class DiffusionConfig(PreTrainedConfig):
     crop_ratio: float = 1.0
     crop_shape: tuple[int, int] | None = None
     crop_is_random: bool = True
-    pretrained_backbone_weights: str | None = "ResNet18_Weights.IMAGENET1K_V1"
-    use_group_norm: bool = False
+    pretrained_backbone_weights: str | None = None
+    use_group_norm: bool = True
     spatial_softmax_num_keypoints: int = 32
-    use_separate_rgb_encoder_per_camera: bool = True
+    use_separate_rgb_encoder_per_camera: bool = False
     # Unet.
     down_dims: tuple[int, ...] = (512, 1024, 2048)
     kernel_size: int = 5

src/lerobot/policies/vqbet/configuration_vqbet.py

@@ -97,8 +97,8 @@ class VQBeTConfig(PreTrainedConfig):
     vision_backbone: str = "resnet18"
     crop_shape: tuple[int, int] | None = (84, 84)
     crop_is_random: bool = True
-    pretrained_backbone_weights: str | None = "ResNet18_Weights.IMAGENET1K_V1"
-    use_group_norm: bool = False
+    pretrained_backbone_weights: str | None = None
+    use_group_norm: bool = True
     spatial_softmax_num_keypoints: int = 32
     # VQ-VAE
     n_vqvae_training_steps: int = 20000

src/lerobot/robots/lekiwi/config_lekiwi.py

@@ -46,7 +46,7 @@ class LeKiwiConfig(RobotConfig):
     cameras: dict[str, CameraConfig] = field(default_factory=lekiwi_cameras_config)
 
     # Set to `True` for backward compatibility with previous policies/dataset
-    use_degrees: bool = True
+    use_degrees: bool = False
 
 
 @dataclass

tests/artifacts/policies/pusht_diffusion_/actions.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:51effd76b73e972f10d31f5084ab906386134b600c87b2668767d30232a902bd
+oid sha256:54aecbc1af72a4cd5e9261492f5e7601890517516257aacdf2a0ffb3ce281f1b
 size 992

tests/artifacts/policies/pusht_diffusion_/grad_stats.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:d4d7a16ca67f9adefac0e0620a7b2e9c822f2db42faaaced7a89fbad60e5ead4
-size 47680
+oid sha256:88a9c3775a2aa1e90a08850521970070a4fcf0f6b82aab43cd8ccc5cf77e0013
+size 47424

tests/artifacts/policies/pusht_diffusion_/output_dict.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:796c439ee8a64bf9901ff8325e7419bda8bd316360ee95e6304e8e1ae0f4c36c
+oid sha256:91a2635e05a75fe187a5081504c5f35ce3417378813fa2deaf9ca4e8200e1819
 size 68

tests/artifacts/policies/pusht_diffusion_/param_stats.safetensors

@@ -1,3 +1,3 @@
 version https://git-lfs.github.com/spec/v1
-oid sha256:ad33a8b47c39c2e1374567ff9da43cdb95e2dbe904c1b02a35051346d3043095
-size 47680
+oid sha256:645bff922ac7bea63ad018ebf77c303c0e4cd2c1c0dc5ef3192865281bef3dc6
+size 47424