admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-07-25 02:36:11 +00:00
Code Issues Packages Projects Releases Wiki Activity

refactoring into using pre and post processor

Browse Source
This commit is contained in:
Maxime Ellerbach
2026-05-21 11:15:52 +00:00
committed by Maximellerbach
parent 999cc625d6
commit 553c217ee2
7 changed files with 268 additions and 114 deletions
Download Patch File Download Diff File Expand all files Collapse all files
tests/policies/vla_jepa/test_world_model.py
+14 -9
View File
@@ -15,10 +15,15 @@ _ACTION_EMBED_DIM = 8
def _make_predictor(
embed_dim: int = 8,
action_embed_dim: int = _ACTION_EMBED_DIM,
predictor_embed_dim: int = 16,
predictor_embed_dim: int = 24,
num_action_tokens: int = 2,
tokens_per_frame: int = 1,
) -> ActionConditionedVideoPredictor:
return ActionConditionedVideoPredictor(
num_frames=1,
img_size=(1, tokens_per_frame),
patch_size=1,
tubelet_size=1,
embed_dim=embed_dim,
action_embed_dim=action_embed_dim,
predictor_embed_dim=predictor_embed_dim,
@@ -38,16 +43,16 @@ def _make_predictor(
],
)
def test_predictor_output_shape(batch: int, num_steps: int, tokens_per_frame: int, embed_dim: int) -> None:
predictor = _make_predictor(embed_dim=embed_dim, action_embed_dim=_ACTION_EMBED_DIM)
frame_tokens = torch.randn(batch, num_steps, tokens_per_frame, embed_dim)
action_tokens = torch.randn(batch, num_steps, 2, _ACTION_EMBED_DIM)
predictor = _make_predictor(embed_dim=embed_dim, action_embed_dim=_ACTION_EMBED_DIM, tokens_per_frame=tokens_per_frame)
frame_tokens = torch.randn(batch, num_steps * tokens_per_frame, embed_dim)
action_tokens = torch.randn(batch, num_steps * 2, _ACTION_EMBED_DIM)
out = predictor(frame_tokens, action_tokens)
assert tuple(out.shape) == (batch, num_steps, tokens_per_frame, embed_dim)
assert tuple(out.shape) == (batch, num_steps * tokens_per_frame, embed_dim)
assert torch.isfinite(out).all()
def test_predictor_step_mismatch_raises() -> None:
predictor = _make_predictor()
frame_tokens = torch.randn(2, 3, 4, 8)
with pytest.raises(ValueError, match="Expected 3 action steps"):
predictor(frame_tokens, torch.randn(2, 2, 2, 8))
predictor = _make_predictor(tokens_per_frame=4)
frame_tokens = torch.randn(2, 3 * 4, 8) # 3 steps, 4 tokens each
with pytest.raises(RuntimeError):
predictor(frame_tokens, torch.randn(2, 2 * 2, 8)) # 2 steps → mismatch