admin/lerobot
1
0
Fork 0
mirror of https://github.com/huggingface/lerobot.git synced 2026-07-08 18:41:54 +00:00
Code Issues Packages Projects Releases Wiki Activity

chore(streaming): drop benchmark and SLURM scaffolding from the PR

Browse Source 
The benchmarks/streaming harness (matrix submitter, summarizer, decode
diagnostic) and the robocasa SLURM scripts are cluster-specific tooling,
not part of the streaming feature. The example's --dummy mode covers
throughput measurement for reviewers. Recoverable from git history
(894fc6bfb) for cluster runs. Example docstring de-personalized.

Co-Authored-By: Claude Fable 5 <noreply@anthropic.com>
This commit is contained in:
Pepijn
2026-06-11 21:46:43 +02:00
parent 894fc6bfb5
commit 38106ea6b4
8 changed files with 2 additions and 802 deletions
Download Patch File Download Diff File Expand all files Collapse all files
examples/scaling/train_streaming_multinode.py
+2 -2
View File
@@ -27,9 +27,9 @@ streaming features of :class:`StreamingLeRobotDataset`:
Launch with Accelerate (single node, N GPUs):
accelerate launch --num_processes=8 examples/scaling/train_streaming_multinode.py \
--repo_id=pepijn223/robocasa_pretrain_human300_v4 --batch_size=64
--repo_id=lerobot/droid_1.0.1 --batch_size=64
Multinode runs use the same script under SLURM; see ``slurm/train_streaming_robocasa.sh``.
Multinode runs launch the same script with your cluster's accelerate/SLURM setup.
Pass ``--dummy`` to skip the model entirely and measure pure dataloading throughput.
"""