i4h-workflow-dataset-replay

작성자: nvidia

Isaac Sim 내에서 기록된 HDF5 에피소드를 재생하여 시각적으로 검증합니다. 사용자가 HDF5 녹화를 재생, 플레이백, 또는 단계별로 실행하도록 요청할 때 사용하세요.

npx skills add https://github.com/nvidia/skills --skill i4h-workflow-dataset-replay

i4h Workflow — Replay Dataset

Purpose

Replay a recorded HDF5 episode inside Isaac Sim for visual verification. Use when the user asks to replay, play back, or step through an HDF5 recording.

Base Code

These steps drive the i4h-workflows base code (the workflows/agentic/ tree). To reuse an existing checkout, set I4H_WORKFLOWS to its path (no clone happens). Otherwise this resolves the current repo, or clones to ~/i4h-workflows — pick that default without prompting. Run every command below from the resolved root:

# Resolve the i4h-workflows base code (provides workflows/agentic/).
ROOT="${I4H_WORKFLOWS:-$(git rev-parse --show-toplevel 2>/dev/null)}"
if [ ! -d "$ROOT/workflows/agentic" ]; then
  ROOT="${I4H_WORKFLOWS:-$HOME/i4h-workflows}"
  [ -d "$ROOT/workflows/agentic" ] || git clone https://github.com/isaac-for-healthcare/i4h-workflows "$ROOT"
fi
export I4H_WORKFLOWS="$ROOT"; cd "$ROOT"

Basics

  • Env config (source of truth): workflows/agentic/config/environments/<env>.yaml — the <env> scene, robot, and cameras Arena replays against.
  • Replay runs arena/run.sh --replay against the env that produced the HDF5.
  • Use it to verify visual correctness before conversion or training.
  • Interpret ordinal wording as zero-based episode indices: "first episode" -> 0, "second episode" -> 1, etc.

Run

Run the steps below in order. Each step is a separate bash call; variables persist in the local agent's tmux session.

Step 1 — setup and resolve HDF5

REPO_ROOT="${I4H_WORKFLOWS:-$(git rev-parse --show-toplevel 2>/dev/null)}"; [ -d "$REPO_ROOT/workflows/agentic" ] || REPO_ROOT="$HOME/i4h-workflows"
ENV_ID=scissor_pick_and_place
RUNS_ROOT="${REPO_ROOT}/workflows/agentic/runs"
EPISODE_INDEX="${EPISODE_INDEX:-0}"  # For "Replay second episode", set EPISODE_INDEX=1.

# Point HDF5_PATH at a real recording (absolute path). Recordings come from teleop, mimic, or
# validate (which writes data/verify.hdf5 under each runs/eval_* dir). List candidates newest-first:
#   find "${RUNS_ROOT}" -name '*.hdf5' -printf '%TY-%Tm-%Td %TH:%TM  %p\n' | sort -r | head
HDF5_PATH="${HDF5_PATH:-}"
if [ ! -f "${HDF5_PATH}" ]; then
  echo "replay: set HDF5_PATH to an existing .hdf5 (got '${HDF5_PATH:-<unset>}'). Candidates:" >&2
  find "${RUNS_ROOT}" -name '*.hdf5' -printf '%TY-%Tm-%Td %TH:%TM  %p\n' 2>/dev/null | sort -r | head
  exit 1
fi

RUN_DIR="${RUNS_ROOT}/replay_${ENV_ID}_$(date +%Y%m%d_%H%M%S)"
mkdir -p "${RUN_DIR}/logs"
ln -sfn "${RUN_DIR}" "${RUNS_ROOT}/.latest"

Step 2 — replay

"${REPO_ROOT}/workflows/agentic/arena/run.sh" \
  --env "${ENV_ID}" \
  --replay "${HDF5_PATH}" \
  --episode-index "${EPISODE_INDEX}" \
  2>&1 | tee "${RUN_DIR}/logs/replay.log"

Notes

  • HDF5_PATH must be an absolute path to an existing recording — --replay resolves a relative path against runs/<env>/, not your cwd, so a bare/relative path silently fails to load. The block lists real candidates if it's unset or wrong.
  • Recordings come from [[i4h-workflow-dataset-teleop]], [[i4h-workflow-dataset-mimic]], or [[i4h-workflow-validate]] (validate writes data/verify.hdf5 under each runs/eval_* dir). There is no default demo.hdf5.
  • --episode-index selects the episode within the HDF5 (zero-based).
  • For "Replay second episode", use --episode-index 1.
  • Use the same env id as the env that produced the recording.

Prerequisites

  • Workflow set up via [[i4h-workflow-setup]] (the .venv must exist).
  • An existing HDF5 recording to replay.
  • The env id that produced the recording.

Limitations

  • Visual verification only; replay does not modify or expand the recording.
  • Replays one episode per invocation, selected by --episode-index.
  • Runs inside Isaac Sim; the env id must match the one that produced the HDF5.

Troubleshooting

  • Error: .venv not found / replay fails to launch - Cause: workflow not set up. Fix: run [[i4h-workflow-setup]] first.
  • Error: replay: set HDF5_PATH to an existing .hdf5 (or recording fails to load) - Cause: HDF5_PATH unset or not a real file. Fix: pick an absolute path from the printed candidates (e.g. a verify.hdf5 under runs/eval_*/data/).
  • Error: episode index out of range - Cause: --episode-index exceeds the episodes in the HDF5. Fix: use a valid zero-based index.
  • Error: mismatched/garbled playback - Cause: --env differs from the env that produced the recording. Fix: use the same env id.

Final Response

Report env, HDF5 path, episode index, launch outcome, visible mismatches.

nvidia의 다른 스킬

compileiq-debug
nvidia
무언가 잘못되었을 때 사용: Search()가 멈추거나, 모든 평가가 INVALID_SCORE를 반환하거나, 점수가 개선되지 않거나, 모든 설정이 동일한 숫자를 반환하거나, ptxas 오류 등이 발생할 때
create-github-pr
nvidia
gh CLI를 사용하여 GitHub 풀 리퀘스트를 생성합니다. 사용자가 새 PR을 만들거나, 코드 리뷰를 제출하거나, 풀 리퀘스트를 열고자 할 때 사용합니다. 트리거 키워드 -…
nemoclaw-maintainer-cross-issue-sweep
nvidia
다른 열린 이슈들을 스캔하여 주어진 PR이 함께 수정하거나 실수로 망가뜨릴 수 있는 이슈를 찾습니다. 인접 수정 기회와 모순 위험을 file:line…과 함께 출력합니다.
fhir-basics
nvidia
에이전트에게 FHIR R4 API의 작동 방식, 사용 가능한 리소스, 검색 매개변수를 사용한 쿼리 방법, 모든 응답 형식을 올바르게 파싱하는 방법을 가르칩니다…
compileiq-validate-result
nvidia
검색이 완료된 후, 속도 향상을 청구하거나 ACF를 발송하기 전에 사용합니다. dump_results CSV를 로드하고, 상위 K개 후보(단일 목표)를 추출합니다…
changelog-audit
nvidia
릴리스 전에 Warp CHANGELOG.md를 감사합니다: 누락된 항목 복구, 사용자 영향별 정렬, 항목 언어 다듬기, 줄 바꿈, (릴리스 브랜치 모드) 비교 업데이트…
maintain-dynamic-plugins
nvidia
NeMo Relay 동적 플러그인 로더, 매니페스트, Rust 네이티브 SDK, gRPC 워커 프로토콜, Python 워커 SDK, 문서, 테스트 및 릴리스 워크플로 커버리지를 유지 관리합니다.
dgx-diagnose
nvidia
일반적인 DGX Station GB300 문제 진단 — CUDA 충돌, 잘못된 GPU 타겟팅, vLLM/SGLang 컨테이너 버그, MIG 상태 문제, NVLink/Fabric Manager 오류,…