Commit Graph

18 Commits

Author SHA1 Message Date
Pepijn a6dd28e8b4 fix(profiling): tolerate groot dep-install failure
groot's only policy-specific dependency is flash-attn, which has no
prebuilt wheel for torch 2.10 and requires nvcc to build from source.
The CI image is based on nvidia/cuda:12.4.1-base, which ships the
CUDA runtime but not the compiler toolkit, so the source build fails
with `/usr/local/cuda/bin/nvcc: No such file or directory`. The
repo's own pyproject.toml already carries a TODO acknowledging this:
gr00t needs bespoke flash-attn install steps.

Treat this as an environmental limitation rather than a regression:
dep-install failures for groot are logged via `::warning::` and skip
the policy without failing the job. Dep-install failures for any
other policy remain fatal, so real regressions still surface.

Made-with: Cursor
2026-04-16 21:15:14 +02:00
Pepijn 00e9defb80 fix(profiling): build flash-attn without isolation for groot
groot depends on flash-attn, which fails to build in uv's default
isolated build env because it doesn't declare torch as a build-time
dependency. Torch is a core lerobot dep and is already present in
the target venv when groot is synced, so we can safely disable
build isolation just for flash-attn. The flag is a no-op for
policies that don't pull in flash-attn.

Made-with: Cursor
2026-04-16 20:21:58 +02:00
Pepijn d483dd4c4b feat(profiling): profile groot, xvla, diffusion, wall_x on PRs
Add groot, xvla, diffusion and wall_x (wall-oss-flow) to the smoke
profiling filter and switch the runner to per-policy dependency
resolution. Each policy now gets its own `uv sync --extra <policy>`
pass followed by a profiling run, so heavy or conflicting extras
(flash-attn, peft, diffusers, etc.) can never block another policy's
profiling. A failure in one policy is logged and surfaces a non-zero
exit at the end instead of aborting the matrix.

Made-with: Cursor
2026-04-16 19:04:27 +02:00
Pepijn 8ece10e484 feat(ci): profile more models in pr smoke runs 2026-04-16 14:49:37 +02:00
Pepijn ddeb216ab9 fix(ci): skip hub publish for pr profiling runs 2026-04-16 14:38:43 +02:00
Pepijn d46d67f75d fix(profiling): forward GIT_REF + PR_NUMBER into Docker container
The previous commit moved these expressions from inline shell expansion
to job-level env: vars, but the profiling script runs inside a Docker
container. Job-level env vars are only visible in the runner, not inside
the container — they need explicit -e flags on the docker run command
(same pattern as HOST_GIT_COMMIT which was already forwarded).

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-16 13:38:13 +02:00
Pepijn b746cd3c61 fix(profiling): sort import + move expressions to env vars for zizmor
Pre-commit Quality gate flagged two issues:

1. ruff/isort: `from numbers import Real` must sort after
   `from collections.abc import Callable` (stdlib alphabetical order).

2. zizmor (high): `github.head_ref`, `github.ref_name`,
   `github.event.inputs.git_ref`, and `github.event.pull_request.head.sha`
   were expanded directly in `run:` shell blocks, which zizmor flags as
   attacker-controllable. Move all four into job-level `env:` vars
   (GIT_REF, PR_NUMBER, HOST_GIT_COMMIT) so the shell only sees env-var
   references — the same pattern the workflow already uses for
   PROFILE_MODE, POLICY_FILTER, etc.

Co-Authored-By: Claude Opus 4.6 (1M context) <noreply@anthropic.com>
2026-04-16 13:30:13 +02:00
Pepijn b27e838376 fix(profiling): publish preview rows to existing dataset 2026-04-16 11:54:35 +02:00
Pepijn 40470648d1 feat(profiling): publish preview runs for dashboard debugging 2026-04-16 10:54:34 +02:00
Pepijn 9dc38d9993 fix(ci): isolate torch cache in profiling job 2026-04-16 09:32:16 +02:00
Pepijn 3922f81791 fix(ci): set HF_LEROBOT_HOME in profiling job 2026-04-15 23:35:27 +02:00
Pepijn e1b22ed1c4 fix(ci): set torchinductor cache dir in profiling job 2026-04-15 22:55:31 +02:00
Pepijn f2d0f04dd0 fix(ci): isolate profiling container home dirs 2026-04-15 22:51:22 +02:00
Pepijn 3ea722c6c0 fix(ci): run profiling container as runner user 2026-04-15 22:47:29 +02:00
Pepijn 48660e7a7c fix(ci): avoid host shell expansion in policy error 2026-04-15 22:42:34 +02:00
Pepijn c94fe868c9 fix(ci): install only profiling policy extras 2026-04-15 22:38:37 +02:00
Pepijn d4f27cfb6e fix(ci): restore docker env line continuation 2026-04-15 22:33:14 +02:00
Pepijn 1a2aec1b04 feat(profiling): add weekly model profiling 2026-04-15 22:31:44 +02:00