admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-05-17 17:50:09 +00:00
Code Issues Packages Projects Releases Wiki Activity

chore(dependencies): upgrade rerun (#2237)

Browse Source 
* chore(dependencies): upgrade rerun

Co-authored-by: Ben Zhang <benzhangniu@gmail.com>

* test(utils): fix rerun scalars

---------

Co-authored-by: Ben Zhang <benzhangniu@gmail.com>
This commit is contained in:
Steven Palma
2025-10-18 01:35:02 +02:00
committed by GitHub
parent d6ea3bbce0
commit da5d2f3e91
4 changed files with 18 additions and 18 deletions
Download Patch File Download Diff File Expand all files Collapse all files
src/lerobot/scripts/lerobot_dataset_viz.py
+8 -8
View File
@@ -141,15 +141,15 @@ def visualize_dataset(
gc.collect()
if mode == "distant":
rr.serve(open_browser=False, web_port=web_port, ws_port=ws_port)
rr.serve_web_viewer(open_browser=False, web_port=web_port)
logging.info("Logging to Rerun")
for batch in tqdm.tqdm(dataloader, total=len(dataloader)):
# iterate over the batch
for i in range(len(batch["index"])):
rr.set_time_sequence("frame_index", batch["frame_index"][i].item())
rr.set_time_seconds("timestamp", batch["timestamp"][i].item())
rr.set_time("frame_index", sequence=batch["frame_index"][i].item())
rr.set_time("timestamp", timestamp=batch["timestamp"][i].item())
# display each camera image
for key in dataset.meta.camera_keys:
@@ -159,21 +159,21 @@ def visualize_dataset(
# display each dimension of action space (e.g. actuators command)
if ACTION in batch:
for dim_idx, val in enumerate(batch[ACTION][i]):
rr.log(f"{ACTION}/{dim_idx}", rr.Scalar(val.item()))
rr.log(f"{ACTION}/{dim_idx}", rr.Scalars(val.item()))
# display each dimension of observed state space (e.g. agent position in joint space)
if OBS_STATE in batch:
for dim_idx, val in enumerate(batch[OBS_STATE][i]):
rr.log(f"state/{dim_idx}", rr.Scalar(val.item()))
rr.log(f"state/{dim_idx}", rr.Scalars(val.item()))
if DONE in batch:
rr.log(DONE, rr.Scalar(batch[DONE][i].item()))
rr.log(DONE, rr.Scalars(batch[DONE][i].item()))
if REWARD in batch:
rr.log(REWARD, rr.Scalar(batch[REWARD][i].item()))
rr.log(REWARD, rr.Scalars(batch[REWARD][i].item()))
if "next.success" in batch:
rr.log("next.success", rr.Scalar(batch["next.success"][i].item()))
rr.log("next.success", rr.Scalars(batch["next.success"][i].item()))
if mode == "local" and save:
# save .rrd locally
src/lerobot/utils/visualization_utils.py
+6 -6
View File
@@ -46,7 +46,7 @@ def log_rerun_data(
This function iterates through the provided observation and action dictionaries and sends their contents
to the Rerun viewer. It handles different data types appropriately:
- Scalar values (floats, ints) are logged as `rr.Scalar`.
- Scalars values (floats, ints) are logged as `rr.Scalars`.
- 3D NumPy arrays that resemble images (e.g., with 1, 3, or 4 channels first) are transposed
from CHW to HWC format and logged as `rr.Image`.
- 1D NumPy arrays are logged as a series of individual scalars, with each element indexed.
@@ -65,7 +65,7 @@ def log_rerun_data(
key = k if str(k).startswith(OBS_PREFIX) else f"{OBS_STR}.{k}"
if _is_scalar(v):
rr.log(key, rr.Scalar(float(v)))
rr.log(key, rr.Scalars(float(v)))
elif isinstance(v, np.ndarray):
arr = v
# Convert CHW -> HWC when needed
@@ -73,7 +73,7 @@ def log_rerun_data(
arr = np.transpose(arr, (1, 2, 0))
if arr.ndim == 1:
for i, vi in enumerate(arr):
rr.log(f"{key}_{i}", rr.Scalar(float(vi)))
rr.log(f"{key}_{i}", rr.Scalars(float(vi)))
else:
rr.log(key, rr.Image(arr), static=True)
@@ -84,13 +84,13 @@ def log_rerun_data(
key = k if str(k).startswith("action.") else f"action.{k}"
if _is_scalar(v):
rr.log(key, rr.Scalar(float(v)))
rr.log(key, rr.Scalars(float(v)))
elif isinstance(v, np.ndarray):
if v.ndim == 1:
for i, vi in enumerate(v):
rr.log(f"{key}_{i}", rr.Scalar(float(vi)))
rr.log(f"{key}_{i}", rr.Scalars(float(vi)))
else:
# Fall back to flattening higher-dimensional arrays
flat = v.flatten()
for i, vi in enumerate(flat):
rr.log(f"{key}_{i}", rr.Scalar(float(vi)))
rr.log(f"{key}_{i}", rr.Scalars(float(vi)))