diff --git a/tasks/task_083/instruction.md b/tasks/task_083/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..70ea17eedef8b0be9c8d8b890f0a40a90af36d74 --- /dev/null +++ b/tasks/task_083/instruction.md @@ -0,0 +1,22 @@ +# Repair bond-order species reporting + +You are working with a local snapshot of ReacNetGenerator, a tool for building +reaction networks from reactive molecular-dynamics trajectories. + +Run the public workbench: + +```bash +python reproduce.py +``` + +The bundled bond-file fixture exercises a small ring-like structure under +several bond-order assignments. Inspect the report and repair the source so the +molecule, species, and reaction outputs are chemically consistent for the +fixture. + +Keep changes inside `source/`. Do not edit the public fixtures or the +reproduction script. + +Do not hard-code the supplied fixture or report values. Your fix should work for +other bond-order variants that preserve atom membership and bonding +connectivity. diff --git a/tasks/task_083/metadata.json b/tasks/task_083/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..23da113ce784b472bf72f3ec25763e8df5d2804c --- /dev/null +++ b/tasks/task_083/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "083", + "title": "Repair bond-order species reporting", + "domain": "computational-chemistry", + "language": "python", + "source_repository": "https://github.com/deepmodeling/reacnetgenerator", + "source_commit": "37b0bef6fd9d5945f1b752ac82d4f8564728cd8b", + "source_license": "LGPL-3.0-family", + "license_source": "LICENSE", + "gpl_family": true, + "public_payload_sha256": "48f2557e6046a003e567fb8b9055d9a6f9a93401f664313f9c04cedadd30780c" +} diff --git a/tasks/task_083/task.toml b/tasks/task_083/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..e9b4b9dc6127ccac192f96c751fa6b4209361a41 --- /dev/null +++ b/tasks/task_083/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-083" +description = "Repair bond-order species reporting" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "083" +language = "python" +base_commit_hash = "37b0bef6fd9d5945f1b752ac82d4f8564728cd8b" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-083:v0.1.0@sha256:9483e6962b892e928f5f1e1c608d646d2d35c5899da15f57fa7bc8c23bd2e14f" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_083" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-083:v0.1.0@sha256:a1c0f1dabd6d4c334c815cb398eda399fbb8b7739b7b5ec60ea8abb1882b7727" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_083" + +[[verifier.collect]] +command = "cd /app/task_083 && git config --global --add safe.directory /app/task_083 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_083/tests/Dockerfile b/tasks/task_083/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..45214e07b2bf8131fcf1e1faaa39fdf6087eee04 --- /dev/null +++ b/tasks/task_083/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-083:v0.1.0@sha256:a1c0f1dabd6d4c334c815cb398eda399fbb8b7739b7b5ec60ea8abb1882b7727 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_083/tests/test.sh b/tasks/task_083/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_083/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_084/environment/Dockerfile b/tasks/task_084/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..a875626125f8f91d7f564820500993851f6f8024 --- /dev/null +++ b/tasks/task_084/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_084 +COPY repo/ /app/task_084/source/ +COPY public/ /app/task_084/ +COPY requirements.lock /app/task_084/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_084:/app/task_084/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_084/instruction.md b/tasks/task_084/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..5d43d296c9d40551a87240c6809616e2ddcf466b --- /dev/null +++ b/tasks/task_084/instruction.md @@ -0,0 +1,27 @@ +# Repair ABACUS relaxation archive/reload semantics + +The supplied ABACUS relaxation examples can be loaded, but the supported +archive/reload workflow does not preserve the trajectory in a scientifically +usable form consistently across the provided examples. + +Study the supplied paper and method notes, reproduce the issue, inspect the +source, and repair the implementation so that valid ABACUS relaxation +catalogues remain scientifically coherent through the supported +archive/reload workflow. The repair must generalize beyond the supplied +examples. Do not replace scientific objects with fixed outputs, +special-case the supplied data, or change the public reproduction to hide the +failure. + +Work only under this task directory. The evaluation environment is Linux and +has no network access. You may modify files under `source/`; do not modify the +task statement, reproduction, fixtures, workflow, or evaluation +infrastructure. + +Run the public smoke test with: + +```bash +python reproduce.py +``` + +The command returns `1` when the untouched source reproduces the issue, `0` +after a successful repair, and `2` for a runner or environment failure. diff --git a/tasks/task_084/metadata.json b/tasks/task_084/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..b25ae28a8ee13de527f9a1a7e1841125bf1904ba --- /dev/null +++ b/tasks/task_084/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "084", + "title": "Repair ABACUS relaxation archive/reload semantics", + "domain": "materials-science-scientific-data-semantics", + "language": "python", + "source_repository": "https://github.com/deepmodeling/dpdata", + "source_commit": "80abfc5037876c0e883dcb58cd96521397e3fe36", + "source_license": "LGPL-3.0-family", + "license_source": "LICENSE", + "gpl_family": true, + "public_payload_sha256": "0b14a30b4b7ceb203ee7c08d2196b2c8b6a4efa48f85b12627a87a49f5f8b83d" +} diff --git a/tasks/task_084/task.toml b/tasks/task_084/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..032a8f5f4804f100fd5868fcdce3bacc0b05ce99 --- /dev/null +++ b/tasks/task_084/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-084" +description = "Repair ABACUS relaxation archive/reload semantics" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "084" +language = "python" +base_commit_hash = "80abfc5037876c0e883dcb58cd96521397e3fe36" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-084:v0.1.0@sha256:735e7ece88fa49a77257cfb2a59b435ed6ef511e99cc1739e24b366ea129836b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_084" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-084:v0.1.0@sha256:8cbe4554c727f25c1584f79c48f574a3d08e1fcf6d3c4cb32df36a7d8482216a" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_084" + +[[verifier.collect]] +command = "cd /app/task_084 && git config --global --add safe.directory /app/task_084 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_084/tests/Dockerfile b/tasks/task_084/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..5ec576948a534c02215a180e2e4525144ddeeb54 --- /dev/null +++ b/tasks/task_084/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-084:v0.1.0@sha256:8cbe4554c727f25c1584f79c48f574a3d08e1fcf6d3c4cb32df36a7d8482216a +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_084/tests/test.sh b/tasks/task_084/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_084/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_085/environment/Dockerfile b/tasks/task_085/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..0dd2895089959af08b801356679c1d3c2e5ea9a5 --- /dev/null +++ b/tasks/task_085/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=1 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_085 +COPY repo/ /app/task_085/source/ +COPY public/ /app/task_085/ +COPY requirements.lock /app/task_085/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_085:/app/task_085/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_085/instruction.md b/tasks/task_085/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..3eb0e5906a7e94678636d3133ebb580f58cbf340 --- /dev/null +++ b/tasks/task_085/instruction.md @@ -0,0 +1,13 @@ +# Inconsistent archived observables + +A molecular-modelling workflow is replaying archived DFTB+ calculations with +this historical source snapshot. Several valid archived records complete +successfully, yet some reported observables do not stay consistent across the +documented input conventions. + +Run the supplied offline workflow, inspect the raw reports, the complete source +snapshot, the official DFTB+ materials, and the documented input conventions. +Repair the historical implementation so supported calculations behave +consistently for valid inputs. The repair must generalize beyond the supplied +archives. Do not hard-code the fixtures or reports, and do not modify the +public workflow or scientific inputs. diff --git a/tasks/task_085/metadata.json b/tasks/task_085/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..fb5a868530d8cee13205aa9dd02c6f6ea152593d --- /dev/null +++ b/tasks/task_085/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "085", + "title": "Inconsistent archived observables", + "domain": "computational-chemistry-and-solvation", + "language": "fortran", + "source_repository": "https://github.com/dftbplus/dftbplus", + "source_commit": "d4da8c5027b643291f3a6295efd4340ce79d6a36", + "source_license": "AGPL-family", + "license_source": "COPYING", + "gpl_family": true, + "public_payload_sha256": "2db37177603a05c62b276b94c738f8592f40853359df72a3f25205c1d5bb5748" +} diff --git a/tasks/task_085/task.toml b/tasks/task_085/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..2890c8d5f6fb4bcc1886961320c9b780f421355b --- /dev/null +++ b/tasks/task_085/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-085" +description = "Inconsistent archived observables" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "085" +language = "fortran" +base_commit_hash = "d4da8c5027b643291f3a6295efd4340ce79d6a36" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-fortran-task-085:v0.1.0@sha256:c4613c3e94fefc6cd721075d314d4fc463269cdd92cf7f6d0b656078a838a7cd" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_085" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-fortran-task-085:v0.1.0@sha256:e68ac55c7cfb257756103351c305b97789d72f1e79203305d2aa7fa39e657eb1" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_085" + +[[verifier.collect]] +command = "cd /app/task_085 && git config --global --add safe.directory /app/task_085 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_085/tests/Dockerfile b/tasks/task_085/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..5d4ebe5af55fa09da80da5db86438c16fc3e8718 --- /dev/null +++ b/tasks/task_085/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-fortran-task-085:v0.1.0@sha256:e68ac55c7cfb257756103351c305b97789d72f1e79203305d2aa7fa39e657eb1 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_085/tests/test.sh b/tasks/task_085/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_085/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_086/environment/Dockerfile b/tasks/task_086/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..5349bab548d7b4f485be6576f0e704e9cfe82241 --- /dev/null +++ b/tasks/task_086/environment/Dockerfile @@ -0,0 +1,63 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_086 +COPY repo/ /app/task_086/source/ +COPY public/ /app/task_086/ +COPY requirements.lock /app/task_086/requirements.lock +COPY wheels/torch-2.6.0+cpu-cp311-cp311-linux_x86_64.whl /app/task_086/torch-2.6.0+cpu-cp311-cp311-linux_x86_64.whl + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir /app/task_086/torch-2.6.0+cpu-cp311-cp311-linux_x86_64.whl \ + && rm -f /app/task_086/torch-2.6.0+cpu-cp311-cp311-linux_x86_64.whl \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_086:/app/task_086/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_086/instruction.md b/tasks/task_086/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..7616a5eac681731aa76f23400fbde8804b2cbdbc --- /dev/null +++ b/tasks/task_086/instruction.md @@ -0,0 +1,10 @@ +Run `python reproduce.py` from this task directory and inspect the bundled +source snapshot and offline workbench. + +The current preflight is small on purpose, but it exercises a boundary case +that should not be treated as valid. Repair the package code in `source/` so +the workbench reaches the expected post-fix status without changing the +fixture, the public runner, or the task metadata. + +Keep the repair general. The corrected code should remain sensible for other +valid materials inputs, not only for the supplied local case. diff --git a/tasks/task_086/metadata.json b/tasks/task_086/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..9a8115b97e225ef668bed2db7b50f6ba6adf4cfe --- /dev/null +++ b/tasks/task_086/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "086", + "title": "Repair a silent MatGL fine-tuning incompatibility", + "domain": "atomistic-machine-learning-potentials", + "language": "python", + "source_repository": "https://github.com/materialyzeai/matgl", + "source_commit": "02b415ba8bb25244bc5e3ad970491f0fbbea0292", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "dd83643eed59e22c0cf5fa9d29e4de990944b8add3edd5058391df75a585f537" +} diff --git a/tasks/task_086/task.toml b/tasks/task_086/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..7df7419802bb6e2ef678cf36e8b1b5e777dfee33 --- /dev/null +++ b/tasks/task_086/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-086" +description = "Repair a silent MatGL fine-tuning incompatibility" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "086" +language = "python" +base_commit_hash = "02b415ba8bb25244bc5e3ad970491f0fbbea0292" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-086:v0.1.0@sha256:7a54bd7e811a51abdd7571abbc03213eece0310016b174bfa52dcbe4ad171476" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_086" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-086:v0.1.0@sha256:fb47312ac7443db59115807334c5d2345e3a2fb7a8f83d89bc8bfb48f7ce4bac" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_086" + +[[verifier.collect]] +command = "cd /app/task_086 && git config --global --add safe.directory /app/task_086 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_086/tests/Dockerfile b/tasks/task_086/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..d76a9e9823ab863ee00f43613a3906ab629bd2ee --- /dev/null +++ b/tasks/task_086/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-086:v0.1.0@sha256:fb47312ac7443db59115807334c5d2345e3a2fb7a8f83d89bc8bfb48f7ce4bac +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_086/tests/test.sh b/tasks/task_086/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_086/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_087/environment/Dockerfile b/tasks/task_087/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..920adc67769014d197a02203c2097bc9f0c02854 --- /dev/null +++ b/tasks/task_087/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_087 +COPY repo/ /app/task_087/source/ +COPY public/ /app/task_087/ +COPY requirements.lock /app/task_087/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_087:/app/task_087/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_087/instruction.md b/tasks/task_087/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..79dbc6df169fcf0b4c3d7fbfc24d48cf1d632fd3 --- /dev/null +++ b/tasks/task_087/instruction.md @@ -0,0 +1,10 @@ +Run `python reproduce.py` from this task directory and inspect the bundled +source snapshot and offline workbench. + +The public workflow should expose a finite, inspectable diffusion-MRI +observation on the supplied local case, but the reported output is irregular +before repair. Repair the implementation in `source/` so the same workflow +produces a coherent result on the public case and on broader valid inputs. + +Keep the repair general. Do not hard-code the supplied observation or change +the public runner, public metadata, or task directory layout. diff --git a/tasks/task_087/metadata.json b/tasks/task_087/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..9d6b927a8d0b78649fedcc41e71e6c0fe1c69852 --- /dev/null +++ b/tasks/task_087/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "087", + "title": "Repair a silent diffusion-MRI denoising inconsistency", + "domain": "diffusion-mri-denoising", + "language": "python", + "source_repository": "https://github.com/dipy/dipy", + "source_commit": "ee6016ab81dd649ab40439d75212f35669302810", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "442792643e44ca3c36c70b3d63c80708e9f58b5972e5bdf43e90834720834814" +} diff --git a/tasks/task_087/task.toml b/tasks/task_087/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..d301ebc2807b3699a38f05a9f97fe2f319051620 --- /dev/null +++ b/tasks/task_087/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-087" +description = "Repair a silent diffusion-MRI denoising inconsistency" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "087" +language = "python" +base_commit_hash = "ee6016ab81dd649ab40439d75212f35669302810" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-087:v0.1.0@sha256:5a00e4d7a825618ba70259ef5c5e229f7ba2549ab7648ead6fc91fce31256793" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_087" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-087:v0.1.0@sha256:b1ce39c75913bccef5f09acfd2b989857c071fc64c7fdc0dbff092e14dc5270f" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_087" + +[[verifier.collect]] +command = "cd /app/task_087 && git config --global --add safe.directory /app/task_087 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_087/tests/Dockerfile b/tasks/task_087/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..ee3f64c67131e948314c8da8375a6ad86c14b4bb --- /dev/null +++ b/tasks/task_087/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-087:v0.1.0@sha256:b1ce39c75913bccef5f09acfd2b989857c071fc64c7fdc0dbff092e14dc5270f +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_087/tests/test.sh b/tasks/task_087/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_087/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_088/environment/Dockerfile b/tasks/task_088/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..45f788efe3b09c38ccab6ac671fc0be83f358bee --- /dev/null +++ b/tasks/task_088/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_088 +COPY repo/ /app/task_088/source/ +COPY public/ /app/task_088/ +COPY requirements.lock /app/task_088/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_088:/app/task_088/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_088/instruction.md b/tasks/task_088/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..0a1afffc2a6607c9eabb72e1abc5bbf7d28aa25c --- /dev/null +++ b/tasks/task_088/instruction.md @@ -0,0 +1,22 @@ +# Repair a boundary aggregation failure + +A short time series fails when it is aggregated into a coarser bin. The same +workflow should run to completion and return a finite aggregate for valid +inputs. + +Reproduce the issue from the supplied source snapshot and references, inspect +the implementation, and repair the behavior so the aggregation workflow works +for the supported coordinates. The fix must generalize beyond the public +example and must not hard-code a single span or output value. + +Work only in this task directory. The evaluation environment is Linux and has +no network access. You may modify files under `source/`. + +Use: + +```bash +python reproduce.py +``` + +The command returns `1` for the reproduced scientific inconsistency, `0` after +a successful repair, and `2` for a runner or environment failure. diff --git a/tasks/task_088/metadata.json b/tasks/task_088/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..c063fec78c78d09879de1cb913914b4ceb977007 --- /dev/null +++ b/tasks/task_088/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "088", + "title": "Repair a boundary aggregation failure", + "domain": "boundary-aggregation", + "language": "python", + "source_repository": "https://github.com/pydata/xarray", + "source_commit": "0b071e369a88f6672accc2a99c40d3b4b334f2f9", + "source_license": "Apache-2.0", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "5e7474f8aa8269a9d76e03f048cce31b30e91fadfa8bb5e3401d63b59c63ddbf" +} diff --git a/tasks/task_088/task.toml b/tasks/task_088/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..ba90d29431f98dd2ed1050eb1e2b85a1b6378260 --- /dev/null +++ b/tasks/task_088/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-088" +description = "Repair a boundary aggregation failure" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "088" +language = "python" +base_commit_hash = "0b071e369a88f6672accc2a99c40d3b4b334f2f9" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-088:v0.1.0@sha256:c98a81a135c210d7681b57574c8f39a00f1a92778e458dc5300a7f2775e2ba6f" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_088" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-088:v0.1.0@sha256:9f8a00b079fb998f0672f57af8f75628dabc8582122a294f2b20cd304503ac6e" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_088" + +[[verifier.collect]] +command = "cd /app/task_088 && git config --global --add safe.directory /app/task_088 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_088/tests/Dockerfile b/tasks/task_088/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..46ef481dd0486a64f2e96b141bcd0b2086750a33 --- /dev/null +++ b/tasks/task_088/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-088:v0.1.0@sha256:9f8a00b079fb998f0672f57af8f75628dabc8582122a294f2b20cd304503ac6e +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_088/tests/test.sh b/tasks/task_088/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_088/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_089/environment/Dockerfile b/tasks/task_089/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..c3762d789e6c6bf7a58e147f54e2608f53978b38 --- /dev/null +++ b/tasks/task_089/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_089 +COPY repo/ /app/task_089/source/ +COPY public/ /app/task_089/ +COPY requirements.lock /app/task_089/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_089:/app/task_089/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_089/instruction.md b/tasks/task_089/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..03728a8cb359bd085d7248efa1d2768463311fb8 --- /dev/null +++ b/tasks/task_089/instruction.md @@ -0,0 +1,30 @@ +# Repair inconsistent ORCA excited-state spectrum parsing + +A computational-chemistry workflow imports ORCA excited-state spectra from a +historical source snapshot. The parser completes and returns finite transition +data, but a downstream spectrum summary made from the parsed attributes is not +consistent with the supplied ORCA-style record. + +Run the public reproduction, inspect the complete source snapshot and supplied +scientific material, and repair the implementation so valid ORCA excited-state +spectrum records retain their scientific meaning in the parsed data model. The +repair must generalize to other valid ORCA spectra, output-version layouts, +additional transition-property sections, and related spectroscopy sections that +may appear in the same output. + +Do not hard-code the supplied fixture, spectrum names, line numbers, numerical +values, or generated report. Work only in this task directory. The evaluation +environment is Linux and has no network access. You may modify files under +`source/`; do not modify the public reproduction, fixtures, workflow, or +evaluation infrastructure. + +Use: + +```bash +python reproduce.py +``` + +The command is a diagnostic workflow. It returns `0` when the parser can produce +a finite nonempty spectrum observation, `1` when it cannot, and `2` for a runner +or dependency failure. The independent verifier checks scientific consistency on +broader spectra. diff --git a/tasks/task_089/metadata.json b/tasks/task_089/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..d5800ff318a6e1216c6f5b80ebc7694119f2fed4 --- /dev/null +++ b/tasks/task_089/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "089", + "title": "Repair inconsistent ORCA excited-state oscillator strengths", + "domain": "computational-chemistry-spectroscopy", + "language": "python", + "source_repository": "https://github.com/cclib/cclib", + "source_commit": "6ba33bdd48bf336a81400147d2870323b241e457", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "fe78fd81c8f3bde1750d471aec4416de4b93ff45a1490008fe50e529855c6782" +} diff --git a/tasks/task_089/task.toml b/tasks/task_089/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..bdf0169ebd3ed225c1ea7213856224ec3c2e4528 --- /dev/null +++ b/tasks/task_089/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-089" +description = "Repair inconsistent ORCA excited-state oscillator strengths" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "089" +language = "python" +base_commit_hash = "6ba33bdd48bf336a81400147d2870323b241e457" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-089:v0.1.0@sha256:87e8b70f8d192990fe5a0f5d1773dcb304ab1bf5074a187236f2e0bd041ba356" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_089" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-089:v0.1.0@sha256:66ee7d7df906e3cad4e9d27664f1c6ea9bd6673ef36d5ddd61f5aaa75c88db60" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_089" + +[[verifier.collect]] +command = "cd /app/task_089 && git config --global --add safe.directory /app/task_089 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_089/tests/Dockerfile b/tasks/task_089/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..71db30a0953d4c70daaa86c1fbf9dac931af5951 --- /dev/null +++ b/tasks/task_089/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-089:v0.1.0@sha256:66ee7d7df906e3cad4e9d27664f1c6ea9bd6673ef36d5ddd61f5aaa75c88db60 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_089/tests/test.sh b/tasks/task_089/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_089/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_090/environment/Dockerfile b/tasks/task_090/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..44a41ca86abf3fd2461f6272f4ad0ca964746f46 --- /dev/null +++ b/tasks/task_090/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_090 +COPY repo/ /app/task_090/source/ +COPY public/ /app/task_090/ +COPY requirements.lock /app/task_090/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_090:/app/task_090/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_090/instruction.md b/tasks/task_090/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..6d172dd22a74d8bedf39d152493a1f718cf6ff99 --- /dev/null +++ b/tasks/task_090/instruction.md @@ -0,0 +1,26 @@ +# Restore multidimensional NMR acquisition grids + +A multidimensional NMR acquisition can arrive as a compressed FID stream plus a +sparse coordinate schedule. Study the source snapshot and the public +reproduction. Repair the package so that valid acquisition workflows preserve +that schedule alongside the compressed stream in both eager and low-memory +reader paths, and so the public processing workflow can restore the compressed +rows into a full multidimensional acquisition grid without losing the declared +acquisition and quadrature layout. Preserve ordinary fully-sampled behavior. +The repair must generalize when valid layouts, dimensions, direct FID lengths, +dtypes, and sampling patterns change. + +Do not hard-code the supplied observations or replace the NMR data model with a +fixed answer. Work only under this task directory. The evaluation environment is +offline; all required Python dependencies and source files are provided. + +Run the public diagnostic with: + +```bash +python reproduce.py +``` + +The command returns `0` when the diagnostic report is produced and `2` for an +import, dependency, path, or other runner failure. The public report is a +workbench signal; final evaluation performs independent scientific consistency +checks on broader cases. diff --git a/tasks/task_090/metadata.json b/tasks/task_090/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..5dd67bc16aae6d2ca0ce4ee5c516a66981576e77 --- /dev/null +++ b/tasks/task_090/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "090", + "title": "Restore multidimensional NMR acquisition grids", + "domain": "nuclear-magnetic-resonance-spectroscopy", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE.txt", + "gpl_family": false, + "public_payload_sha256": "dd119886283401127edc350f562371a2bad9d92d79cfe0488afeaa26a8dde9b4" +} diff --git a/tasks/task_090/task.toml b/tasks/task_090/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..bd4b1af5fc8825c7239caa844d9019dda048577a --- /dev/null +++ b/tasks/task_090/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-090" +description = "Restore multidimensional NMR acquisition grids" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "090" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-090:v0.1.0@sha256:a2a7d3de4b603c2e170a1ea679103aeefdd43c1503f5c2210f32fec9f51e3efa" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_090" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-090:v0.1.0@sha256:8ba7b5dff3e7bd208683b7d457ebfe6277ce4bdd218bea4d2e0ea322f49a5007" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_090" + +[[verifier.collect]] +command = "cd /app/task_090 && git config --global --add safe.directory /app/task_090 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_090/tests/Dockerfile b/tasks/task_090/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..97c6dc45398af8a821c5c79176bfb1f03ac31744 --- /dev/null +++ b/tasks/task_090/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-090:v0.1.0@sha256:8ba7b5dff3e7bd208683b7d457ebfe6277ce4bdd218bea4d2e0ea322f49a5007 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_090/tests/test.sh b/tasks/task_090/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_090/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_091/environment/Dockerfile b/tasks/task_091/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..f5f290eea3d7d9f0b97dec68ada992ad35cb06c4 --- /dev/null +++ b/tasks/task_091/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_091 +COPY repo/ /app/task_091/source/ +COPY public/ /app/task_091/ +COPY requirements.lock /app/task_091/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_091:/app/task_091/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_091/instruction.md b/tasks/task_091/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..abae1b8ec57cdcb2b0fe863c69f35d4416e9d43e --- /dev/null +++ b/tasks/task_091/instruction.md @@ -0,0 +1,21 @@ +# Repair a workflow regression + +A bundled input file exercises the public workflow and currently fails the +smoke test. + +Run the public reproduction, inspect the complete source snapshot, and repair +the implementation so valid inputs are handled consistently across +representative cases. The repair must generalize to other valid input files and +similar inputs. Do not hard-code the bundled file or output. + +Work only in this task directory. The evaluation environment is Linux and has +no network access. You may modify files under `source/`. + +Use: + +```bash +python reproduce.py +``` + +The command returns `1` for the reproduced finite inconsistency, `0` after a +successful repair, and `2` for a runner or environment failure. diff --git a/tasks/task_091/metadata.json b/tasks/task_091/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..84802bbaf2f37e7ae5f8258ea138a1ff60ed7c7c --- /dev/null +++ b/tasks/task_091/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "091", + "title": "Repair a workflow regression", + "domain": "materials-science-volumetric-data-semantics", + "language": "python", + "source_repository": "https://github.com/materialsproject/pymatgen-core", + "source_commit": "c8cfee71f3ce5d92c5b5f6ca3d363f8f3a9f4495", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "ee6d2a8fa5af6c99aded10bacf8ffa2910bcb148ed68fb80707b9cebe7eb0437" +} diff --git a/tasks/task_091/task.toml b/tasks/task_091/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..15ab9805a4fe74a97549c2c423319910ac71d79d --- /dev/null +++ b/tasks/task_091/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-091" +description = "Repair a workflow regression" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "091" +language = "python" +base_commit_hash = "c8cfee71f3ce5d92c5b5f6ca3d363f8f3a9f4495" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-091:v0.1.0@sha256:e99b9784d9297b360713e12ff77833dfa629f31298f504322fa3ee98dc833ed8" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_091" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-091:v0.1.0@sha256:274078dbc1df8898a9f4d704e0e833c21b54447ff33e78314d2cc6e8c2e16435" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_091" + +[[verifier.collect]] +command = "cd /app/task_091 && git config --global --add safe.directory /app/task_091 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_091/tests/Dockerfile b/tasks/task_091/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..191decd270d8d9820bf339ca898234d7a99c91f6 --- /dev/null +++ b/tasks/task_091/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-091:v0.1.0@sha256:274078dbc1df8898a9f4d704e0e833c21b54447ff33e78314d2cc6e8c2e16435 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_091/tests/test.sh b/tasks/task_091/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_091/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_092/environment/Dockerfile b/tasks/task_092/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..17ff3f8fe204775302b156f994eea55e446fb9af --- /dev/null +++ b/tasks/task_092/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_092 +COPY repo/ /app/task_092/source/ +COPY public/ /app/task_092/ +COPY requirements.lock /app/task_092/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_092:/app/task_092/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_092/instruction.md b/tasks/task_092/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..b4ce3043afa5ad27fcf0869cda9cc16c63fa4864 --- /dev/null +++ b/tasks/task_092/instruction.md @@ -0,0 +1,11 @@ +Inspect the accompanying scientific material and the supplied source snapshot, then repair the implementation so that Chemprop's reaction workflow behaves consistently on the provided organic-reaction examples. + +A researcher is using the project on small organic reactions. The public workflow completes and writes finite observations, but the behavior is inconsistent with the accompanying material. The repair must generalize beyond the public examples and must not hard-code a molecule, reaction string, case-specific constant, or fixed output. + +Run the public workflow with: + +```bash +python reproduce.py +``` + +The command writes `outputs/public_report.json`. Infrastructure failures should be treated separately from scientific observations. diff --git a/tasks/task_092/metadata.json b/tasks/task_092/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..bd87f763ff66721acbbecad199fe53a091871cab --- /dev/null +++ b/tasks/task_092/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "092", + "title": "Repair inconsistent Chemprop reaction observations", + "domain": "ai-chemistry-reaction-property-learning", + "language": "python", + "source_repository": "https://github.com/chemprop/chemprop", + "source_commit": "1f367d55d1bac14dd6b9e0cd5746574fcfab9a59", + "source_license": "MIT", + "license_source": "LICENSE.txt", + "gpl_family": false, + "public_payload_sha256": "873bfddc79c4be961e1b653f28bf7718df73b74df7a12102e96a97ac64f51f07" +} diff --git a/tasks/task_092/task.toml b/tasks/task_092/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..9b36b35931fc4cc3848c0d5ac4ab0a7d967a48fe --- /dev/null +++ b/tasks/task_092/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-092" +description = "Repair inconsistent Chemprop reaction observations" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "092" +language = "python" +base_commit_hash = "1f367d55d1bac14dd6b9e0cd5746574fcfab9a59" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-092:v0.1.0@sha256:43ddc1b17809459d59ff0d6eaf8886368305e6164d7076d81a1660685a8dfac5" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_092" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-092:v0.1.0@sha256:6eac8072686557e2a87b94a3600cdb5e6e1a94862c97c69cdf34a916ea0d7c83" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_092" + +[[verifier.collect]] +command = "cd /app/task_092 && git config --global --add safe.directory /app/task_092 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_092/tests/Dockerfile b/tasks/task_092/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..f3901a6a16794a1f26d9d0a542d88b6efaf8df90 --- /dev/null +++ b/tasks/task_092/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-092:v0.1.0@sha256:6eac8072686557e2a87b94a3600cdb5e6e1a94862c97c69cdf34a916ea0d7c83 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_092/tests/test.sh b/tasks/task_092/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_092/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_093/environment/Dockerfile b/tasks/task_093/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..f541d76570f40389861951e047c29fc0323d5eea --- /dev/null +++ b/tasks/task_093/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_093 +COPY repo/ /app/task_093/source/ +COPY public/ /app/task_093/ +COPY requirements.lock /app/task_093/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_093:/app/task_093/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_093/instruction.md b/tasks/task_093/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..e0e59091e13a39ad4933456c45d9b224d985005c --- /dev/null +++ b/tasks/task_093/instruction.md @@ -0,0 +1,9 @@ +# Repair a patient-space round trip in MONAI + +You are working on a compact medical-imaging I/O workflow built from the supplied MONAI source snapshot. A small voxel study should keep the same physical meaning when it goes from a reader-produced `MetaTensor` into an ITK image and back through the direct reader path. + +Run the public reproduction and inspect the source and supporting notes. The command completes on the supplied inputs, but the round-tripped image no longer agrees with the direct reader route on the same study. This is a scientific coordinate-consistency problem, not an installation or syntax problem. + +Repair the implementation so that the bridge preserves the same patient-space interpretation for supported 2-D and 3-D scalar volumes, including studies whose NRRD header omits `space` metadata. Preserve the public API, the reader metadata contract, and the round-trip behavior for equivalent explicit-space and implicit-space inputs. Do not hard-code the supplied array values, affine values, or file names. The fix must generalize to new studies, spacings, and origins. + +Do not modify the public reproduction to conceal a failing result or bypass the reader, bridge, or writer implementation. diff --git a/tasks/task_093/metadata.json b/tasks/task_093/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..5120b96e47eeac76a814340a7094547a9acf6f73 --- /dev/null +++ b/tasks/task_093/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "093", + "title": "Repair a patient-space round trip in MONAI", + "domain": "medical-image-processing", + "language": "python", + "source_repository": "https://github.com/Project-MONAI/MONAI", + "source_commit": "5a644e4edcc37e5c963e52fda6fab4c85b3ff02e", + "source_license": "UNKNOWN", + "license_source": "not-detected", + "gpl_family": false, + "public_payload_sha256": "0756b99aa4bc899129a72159146d991cddcd1544c136b1d3e76eb47f583c1cf3" +} diff --git a/tasks/task_093/task.toml b/tasks/task_093/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..6545a6ae0861e72e389b25948261dd56f72c722d --- /dev/null +++ b/tasks/task_093/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-093" +description = "Repair a patient-space round trip in MONAI" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "093" +language = "python" +base_commit_hash = "5a644e4edcc37e5c963e52fda6fab4c85b3ff02e" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-093:v0.1.0@sha256:5a60042d49b60549e07475ed0d0f717e1ceddac7c1e7832cec4803a8c5647e6a" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_093" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-093:v0.1.0@sha256:1a0872591921ff43bf480d6d6daa9e99f8f30c94d74f308337235ca21fe08e26" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_093" + +[[verifier.collect]] +command = "cd /app/task_093 && git config --global --add safe.directory /app/task_093 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_093/tests/Dockerfile b/tasks/task_093/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..8aef01a63f3e50ee68a6c8da736b977ccfe18f6c --- /dev/null +++ b/tasks/task_093/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-093:v0.1.0@sha256:1a0872591921ff43bf480d6d6daa9e99f8f30c94d74f308337235ca21fe08e26 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_093/tests/test.sh b/tasks/task_093/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_093/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_094/environment/Dockerfile b/tasks/task_094/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..a27194f6be3f61df549fafa1bc6c14a1cbb0063b --- /dev/null +++ b/tasks/task_094/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_094 +COPY repo/ /app/task_094/source/ +COPY public/ /app/task_094/ +COPY requirements.lock /app/task_094/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_094:/app/task_094/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_094/instruction.md b/tasks/task_094/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..c1faa7624a77ec41562c4dbcaffc0c36a12dbd97 --- /dev/null +++ b/tasks/task_094/instruction.md @@ -0,0 +1,28 @@ +# Repair magnetic-space-group operations across archived settings + +A materials archive stores magnetic symmetry records contributed by data +producers using their own coordinate conventions. The public diagnostic record +resolves to a finite magnetic operation family, but downstream archive consumers +cannot rely on this retained source to preserve the scientific meaning of valid +records across producer conventions. + +Run the public reproduction, inspect the source and public workflow, and repair +the implementation so that archived magnetic records retain their scientific +operation semantics across supported producer conventions. The repair must +generalize beyond the supplied record. Do not replace the magnetic symmetry +calculation with fixed outputs, special-case the supplied BNS number, or change +the public reproduction to hide a failure. + +Work only under this task directory. The evaluation environment is Linux and +has no network access. You may modify files under `source/`; do not modify the +task statement, reproduction, fixtures, workflow, or evaluation infrastructure. + +Run the public smoke test with: + +```bash +python reproduce.py +``` + +The command returns `0` when the diagnostic workflow completes and writes a +finite report, and `2` for a runner or environment failure. The public report +is a workbench observation, not the final correctness oracle. diff --git a/tasks/task_094/metadata.json b/tasks/task_094/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..192d3bd794342eef12d8d44459bc815d3884278e --- /dev/null +++ b/tasks/task_094/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "094", + "title": "Repair magnetic-space-group operations across archived settings", + "domain": "condensed-matter-magnetism-and-magnetic-symmetry", + "language": "python", + "source_repository": "https://github.com/materialsproject/pymatgen", + "source_commit": "", + "source_license": "MIT", + "license_source": "LICENSE.rst", + "gpl_family": false, + "public_payload_sha256": "83caaa6fe0d259a727975a11f4606bee8784fe286c6fec89eb78769b09ba9b44" +} diff --git a/tasks/task_094/task.toml b/tasks/task_094/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..f2132e069db0301bbb44bcf8cc4799ca5af66a68 --- /dev/null +++ b/tasks/task_094/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-094" +description = "Repair magnetic-space-group operations across archived settings" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "094" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-094:v0.1.0@sha256:29e3fa554e2177b07fb0e7f1cc9b470dd310202d31612af2e75baef512ae97e4" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_094" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-094:v0.1.0@sha256:0b8980bec95351d787ad34347fde67096254dbe62b04ceb69b5d4fd0f8ea86df" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_094" + +[[verifier.collect]] +command = "cd /app/task_094 && git config --global --add safe.directory /app/task_094 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_094/tests/Dockerfile b/tasks/task_094/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..a8a4c7bd20a9aab143892c4cfbb62800427aab3d --- /dev/null +++ b/tasks/task_094/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-094:v0.1.0@sha256:0b8980bec95351d787ad34347fde67096254dbe62b04ceb69b5d4fd0f8ea86df +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_094/tests/test.sh b/tasks/task_094/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_094/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_095/environment/Dockerfile b/tasks/task_095/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..703f6f281e3c04fc70ba75201051338d352a863f --- /dev/null +++ b/tasks/task_095/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_095 +COPY repo/ /app/task_095/source/ +COPY public/ /app/task_095/ +COPY requirements.lock /app/task_095/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_095:/app/task_095/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_095/instruction.md b/tasks/task_095/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..261e8dd14f753dc2d1822d8674aa07d6fe81d76e --- /dev/null +++ b/tasks/task_095/instruction.md @@ -0,0 +1,27 @@ +# Repair a repeated porous-electrode transport workflow + +A porous-network workflow processes a compact sequence of electrode-like +stages. The inputs are finite and valid, and the calculation completes, but the +reported transport observations are not stable across the supplied control +stages. + +Run the public reproduction, inspect the scientific notes and complete source +snapshot, and repair the implementation so the staged transport observations +are scientifically consistent. The repair must generalize to other valid pore +networks, stage orders, transport algorithms, and model parameters. Do not +special-case the supplied fixture or replace the scientific calculation with a +fixed report. + +Work only under this task directory. The evaluation environment is Linux and +has no network access. You may modify files under `source/`. + +Use: + +```bash +python reproduce.py +``` + +The command returns `0` when the diagnostic workflow produces finite +observations and `2` for a runner or environment failure. The report is +diagnostic; the repair must be judged from the scientific behavior, not by +hard-coding one printed value. diff --git a/tasks/task_095/metadata.json b/tasks/task_095/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..dd0d4db335ae98897e119e4ce5aee6933efe7728 --- /dev/null +++ b/tasks/task_095/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "095", + "title": "Repair a repeated porous-electrode transport workflow", + "domain": "new-energy-materials", + "language": "python", + "source_repository": "https://github.com/PMEAL/OpenPNM", + "source_commit": "d8efaa211ea6721cc4a718890f11fb60b8b43326", + "source_license": "UNKNOWN", + "license_source": "not-detected", + "gpl_family": false, + "public_payload_sha256": "9682714aad6233c65f16201b2c737da9e0e133cf60cfc3203b375a3dd3b9487c" +} diff --git a/tasks/task_095/task.toml b/tasks/task_095/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..edb1c34985dac0056137573d2585e8b8b8867f2e --- /dev/null +++ b/tasks/task_095/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-095" +description = "Repair a repeated porous-electrode transport workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "095" +language = "python" +base_commit_hash = "d8efaa211ea6721cc4a718890f11fb60b8b43326" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-095:v0.1.0@sha256:08fb28caff95e4cc11459e4bc8d21965e2b52cfa2ad1829e98aa440eba3f47be" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_095" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-095:v0.1.0@sha256:9e1febafa3ee120cd3fdf92d8882fc81ac9c1f6215094130cc8f58de53fd82f9" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_095" + +[[verifier.collect]] +command = "cd /app/task_095 && git config --global --add safe.directory /app/task_095 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_095/tests/Dockerfile b/tasks/task_095/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..2ad30a30f67d7cfb33b08daa8447e08940a78dd7 --- /dev/null +++ b/tasks/task_095/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-095:v0.1.0@sha256:9e1febafa3ee120cd3fdf92d8882fc81ac9c1f6215094130cc8f58de53fd82f9 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_095/tests/test.sh b/tasks/task_095/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_095/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_096/environment/Dockerfile b/tasks/task_096/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..10537620f68aab7211a40f0fc26718cccd550059 --- /dev/null +++ b/tasks/task_096/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_096 +COPY repo/ /app/task_096/source/ +COPY public/ /app/task_096/ +COPY requirements.lock /app/task_096/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_096:/app/task_096/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_096/instruction.md b/tasks/task_096/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..06b6fc4bf34f41141617f04237299bd3e707c805 --- /dev/null +++ b/tasks/task_096/instruction.md @@ -0,0 +1,8 @@ +# Repair a wrapped tropical-cyclone hazard workflow + +Inspect the paper and source code, run the reproduction workflow, and repair the +implementation so the hazard calculation behaves consistently for wrapped +longitude inputs and equivalent date-line encodings. + +The repair must generalize beyond the public example. Do not hard-code one track +shape, one centroid layout, or one fixed output. diff --git a/tasks/task_096/metadata.json b/tasks/task_096/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..d1496810e0daa21b89fe40897ed0b1fafc420a6f --- /dev/null +++ b/tasks/task_096/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "096", + "title": "Repair a wrapped tropical-cyclone hazard workflow", + "domain": "tropical-cyclone-hazard-monitoring", + "language": "python", + "source_repository": "https://github.com/CLIMADA-project/climada_python", + "source_commit": "104dfdb64ed00fce6c49a5d67be371f4fca016dd", + "source_license": "UNKNOWN", + "license_source": "not-detected", + "gpl_family": false, + "public_payload_sha256": "5da44df229cac52680b35b0d4806ec17dd1e8a8851512c04cbcdf3159e88c977" +} diff --git a/tasks/task_096/task.toml b/tasks/task_096/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..08a70da8d90814cc90362f52f6985c6a48e49a74 --- /dev/null +++ b/tasks/task_096/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-096" +description = "Repair a wrapped tropical-cyclone hazard workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "096" +language = "python" +base_commit_hash = "104dfdb64ed00fce6c49a5d67be371f4fca016dd" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-096:v0.1.0@sha256:b5d83702d5a8bf8b17ce88fe6c95847eaf8b4ad5132e037fbaf6727154256a5f" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_096" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-096:v0.1.0@sha256:f83195b4ce577c453e2228e642418f22e935ebd835db2a0c168b8b7c93be28d9" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_096" + +[[verifier.collect]] +command = "cd /app/task_096 && git config --global --add safe.directory /app/task_096 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_096/tests/Dockerfile b/tasks/task_096/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..26948193c95ec5356e06c1bbbdd44e1b8e535d89 --- /dev/null +++ b/tasks/task_096/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-096:v0.1.0@sha256:f83195b4ce577c453e2228e642418f22e935ebd835db2a0c168b8b7c93be28d9 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_096/tests/test.sh b/tasks/task_096/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_096/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_097/environment/Dockerfile b/tasks/task_097/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..b7cfc337b1ec01c5e6f81690eb6dac9edeb3a1c5 --- /dev/null +++ b/tasks/task_097/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_097 +COPY repo/ /app/task_097/source/ +COPY public/ /app/task_097/ +COPY requirements.lock /app/task_097/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_097:/app/task_097/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_097/instruction.md b/tasks/task_097/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..b4cdc2572f2844ecbf00f1da8183b69772013d52 --- /dev/null +++ b/tasks/task_097/instruction.md @@ -0,0 +1,11 @@ +# Repair spin-model consistency in DeePMD-kit + +Inspect the paper and source code, run the reproduction workflow, and repair the +implementation so the public spin calculation remains consistent across the +dense and lower layouts exercised by the fixture. + +The public workflow uses a few small padded spin systems. It reports aggregate +responses for changes applied to padded batch entries and for control changes +applied to ordinary spin-carrying entries, across both NumPy and Torch +backends. The repair should generalize beyond the particular atom ordering, +padding position, layout, backend, and output mode used by the public run. diff --git a/tasks/task_097/metadata.json b/tasks/task_097/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..73e07b27ac4c7c3b7848f88de0ab3e8fb5b2b960 --- /dev/null +++ b/tasks/task_097/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "097", + "title": "Repair spin-model consistency in DeePMD-kit", + "domain": "magnetism-and-spin-models", + "language": "python", + "source_repository": "https://github.com/deepmodeling/deepmd-kit", + "source_commit": "6c3b985cbc6e2dec51976a20ae51214186c693b5", + "source_license": "UNKNOWN", + "license_source": "not-detected", + "gpl_family": false, + "public_payload_sha256": "0be2c62ea59a56d67afb0a87a5e2d6699169c73794c50b096a0d77f1f9474945" +} diff --git a/tasks/task_097/task.toml b/tasks/task_097/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..eeca86c16a67ef0385b1bb1c118f5987d892eec0 --- /dev/null +++ b/tasks/task_097/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-097" +description = "Repair spin-model consistency in DeePMD-kit" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "097" +language = "python" +base_commit_hash = "6c3b985cbc6e2dec51976a20ae51214186c693b5" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-097:v0.1.0@sha256:d568b716103abb844e7d5f8e92693fc9bda65f9329fa82628da0433f52660de2" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_097" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-097:v0.1.0@sha256:f3cb6aea78fe4974c5f9fa31a55c02cec54773666d27857bfa0c1d9b79856ba1" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_097" + +[[verifier.collect]] +command = "cd /app/task_097 && git config --global --add safe.directory /app/task_097 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_097/tests/Dockerfile b/tasks/task_097/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..20dcd6f93e77130e1cba824264fb266045b1bf88 --- /dev/null +++ b/tasks/task_097/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-097:v0.1.0@sha256:f3cb6aea78fe4974c5f9fa31a55c02cec54773666d27857bfa0c1d9b79856ba1 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_097/tests/test.sh b/tasks/task_097/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_097/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_098/environment/Dockerfile b/tasks/task_098/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..7219b8655ca6936df5c1c3d6d69fb66ffeb26f47 --- /dev/null +++ b/tasks/task_098/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_098 +COPY repo/ /app/task_098/source/ +COPY public/ /app/task_098/ +COPY requirements.lock /app/task_098/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_098:/app/task_098/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_098/instruction.md b/tasks/task_098/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..24ba7ed7f06be4a8794233cda5e8505bd2e2a24b --- /dev/null +++ b/tasks/task_098/instruction.md @@ -0,0 +1,15 @@ +# Repair a satellite geolocation workflow + +The supplied source is a historical slice of a remote-sensing package. A small +geostationary observation product and a cloud-top-height product can be read, +but the resulting geolocation is not physically consistent for supported +product inputs. + +Run the offline reproduction, inspect the supplied scientific documentation, +and trace the product data flow. Repair the implementation so that the +documented satellite geolocation behavior holds for supported inputs. Keep the +workflow offline and preserve the package API used by the supplied runner. + +Do not replace the workflow with a fixed report or a special case for the +public fixture. The implementation must work for other valid satellite +positions, cloud heights, masks, and target grids. diff --git a/tasks/task_098/metadata.json b/tasks/task_098/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..5adbdb05b8ea128a8e945bcf90b4ada630de904c --- /dev/null +++ b/tasks/task_098/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "098", + "title": "Repair inconsistent satellite parallax geolocation", + "domain": "remote-sensing-and-geostationary-geolocation", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "UNKNOWN", + "license_source": "not-detected", + "gpl_family": false, + "public_payload_sha256": "551331a7c8d9ec732af39574c53fd2f2139520ba8eb5a322aed7d412101fe61d" +} diff --git a/tasks/task_098/task.toml b/tasks/task_098/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..9cb6a00c151d01192d099acd0bccc618720001a4 --- /dev/null +++ b/tasks/task_098/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-098" +description = "Repair inconsistent satellite parallax geolocation" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "098" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-098:v0.1.0@sha256:7749c5504674fd868d0c9f318caa3bddbc2eb7beaa77fc4b92822b0c19ced514" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_098" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-098:v0.1.0@sha256:60c92aee990b8ff7809aa094f99a6ada0935ac7fe605bfa6f90345912e56e598" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_098" + +[[verifier.collect]] +command = "cd /app/task_098 && git config --global --add safe.directory /app/task_098 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_098/tests/Dockerfile b/tasks/task_098/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..bb3bd426ab7cc2b4ae0a26890ece86dc8c9d098a --- /dev/null +++ b/tasks/task_098/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-098:v0.1.0@sha256:60c92aee990b8ff7809aa094f99a6ada0935ac7fe605bfa6f90345912e56e598 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_098/tests/test.sh b/tasks/task_098/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_098/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_099/environment/Dockerfile b/tasks/task_099/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..8081095d9951de5cc92eb660467eeaf2f444d7a8 --- /dev/null +++ b/tasks/task_099/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_099 +COPY repo/ /app/task_099/source/ +COPY public/ /app/task_099/ +COPY requirements.lock /app/task_099/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_099:/app/task_099/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_099/instruction.md b/tasks/task_099/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..41bd5ad72bcf2e171ecf465868dad0465ce2c8cb --- /dev/null +++ b/tasks/task_099/instruction.md @@ -0,0 +1,26 @@ +# Repair an inconsistent GNSS point solution + +A GNSS navigation workflow uses the supplied scientific package to solve for a +receiver state from satellite positions and corrected pseudoranges. The +calculation completes and returns finite coordinates, but the recovered +receiver state is materially offset from the supplied control state. + +Run the public reproduction, inspect the complete source snapshot, and repair +the implementation so that the point-positioning workflow is scientifically +consistent. The repair must generalize to other valid measurement tables, +satellite constellations, epoch groupings, and receiver states. Do not +special-case the supplied fixture or replace the scientific calculation with +fixed output. + +Work only in this task directory. The evaluation environment is Linux and has +no network access. You may modify files under `source/`. + +Use: + +```bash +python reproduce.py +``` + +The command returns `1` for the reproduced finite scientific inconsistency, +`0` after a successful repair, and `2` for a runner or environment failure. + diff --git a/tasks/task_099/metadata.json b/tasks/task_099/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..434e0bd1dc5f0e4592d316685e390c4bfd3b523c --- /dev/null +++ b/tasks/task_099/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "099", + "title": "Repair an inconsistent GNSS point solution", + "domain": "satellite-navigation-and-gnss-positioning", + "language": "python", + "source_repository": "https://github.com/Stanford-NavLab/gnss_lib_py", + "source_commit": "9a4e467ba06e636ffa4de6ceee302cefee8b2a34", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "fff992d5304213d729a9c9bac0a2bd157c983e49cf4ca9f4d49da3d4fef14c0f" +} diff --git a/tasks/task_099/task.toml b/tasks/task_099/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..b1677a362012459d8c4c1eeb6ab0ac516201281a --- /dev/null +++ b/tasks/task_099/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-099" +description = "Repair an inconsistent GNSS point solution" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "099" +language = "python" +base_commit_hash = "9a4e467ba06e636ffa4de6ceee302cefee8b2a34" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-099:v0.1.0@sha256:99a563d1d49d0ae4fcff3a709bf6eadb46e5c1514ec902545d7c4fa7b11c529a" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_099" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-099:v0.1.0@sha256:821dfafb1035f4ffe52e58226ae63be3d779edcdd077cd1e2890b9916b67a97b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_099" + +[[verifier.collect]] +command = "cd /app/task_099 && git config --global --add safe.directory /app/task_099 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_099/tests/Dockerfile b/tasks/task_099/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..a89d5fe2e9428902e9c6501de4993db46abefa56 --- /dev/null +++ b/tasks/task_099/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-099:v0.1.0@sha256:821dfafb1035f4ffe52e58226ae63be3d779edcdd077cd1e2890b9916b67a97b +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_099/tests/test.sh b/tasks/task_099/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_099/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_100/environment/Dockerfile b/tasks/task_100/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..037d9219a345d1283c0db44cab5b4f06670cbdf6 --- /dev/null +++ b/tasks/task_100/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_100 +COPY repo/ /app/task_100/source/ +COPY public/ /app/task_100/ +COPY requirements.lock /app/task_100/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_100:/app/task_100/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_100/instruction.md b/tasks/task_100/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..4cfea203a717ae72013f17dd8115e666c40e48ed --- /dev/null +++ b/tasks/task_100/instruction.md @@ -0,0 +1,29 @@ +# Repair calibrated framework embedding relaxation + +A framework-generation workflow uses the relaxed embedding mode to let slot +positions move during optimization while preserving the topology. On the +supplied lower-symmetry panel, the relaxed build currently completes but still +behaves inconsistently: some cases regress bond closure, some cases inflate the +cell, and a pinned control should remain unchanged. + +Repair the implementation so the relaxed workflow behaves consistently on the +whole visible panel and generalizes beyond it. The public comparison checks +bond-closure drift, cell-volume blow-up, and a pinned no-op control. + +Run the public reproduction, inspect the source and documentation, and repair +the implementation so the relaxed workflow behaves consistently on supported +inputs. The repair must generalize beyond the public panel and must not +hard-code the supplied topologies, building units, or numbers. + +Work only in this task directory. The evaluation environment is offline. You +may modify files under `source/`. + +Use: + +```bash +python reproduce.py +``` + +The command returns `1` when the public comparison still shows an +inconsistency, `0` after a successful repair, and `2` for a runner or +environment failure. diff --git a/tasks/task_100/metadata.json b/tasks/task_100/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..f82d7e784f89ed84aff213580fb70ef5733258f2 --- /dev/null +++ b/tasks/task_100/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "100", + "title": "Repair calibrated framework embedding relaxation", + "domain": "reticular-chemistry-and-framework-geometry", + "language": "python", + "source_repository": "https://github.com/DCoupry/autografs", + "source_commit": "c2f22341f2b1eb2f619b76534cf4abe2e5e6b3a3", + "source_license": "LGPL-2.1-family", + "license_source": "COPYING", + "gpl_family": true, + "public_payload_sha256": "1e14eacadfc8bee10ae95178b7335c5fa6d923d9c528037c54d0d52c051c65c5" +} diff --git a/tasks/task_100/task.toml b/tasks/task_100/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..0e3577657ba32677c980e1c18cd649d3e57c34da --- /dev/null +++ b/tasks/task_100/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-100" +description = "Repair calibrated framework embedding relaxation" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "100" +language = "python" +base_commit_hash = "c2f22341f2b1eb2f619b76534cf4abe2e5e6b3a3" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-100:v0.1.0@sha256:9bfa2b87a911734bd13a90d76db6e6fc482793b851cfe9e37e2616368eb4ea8b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_100" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-100:v0.1.0@sha256:24d7057aa54c241cb71822a31b12711ad36c6d13c53d58af790290752df9ef84" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_100" + +[[verifier.collect]] +command = "cd /app/task_100 && git config --global --add safe.directory /app/task_100 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_100/tests/Dockerfile b/tasks/task_100/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..2fcc76f30b65988cfd2e8f6bc603b47d865c733f --- /dev/null +++ b/tasks/task_100/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-100:v0.1.0@sha256:24d7057aa54c241cb71822a31b12711ad36c6d13c53d58af790290752df9ef84 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_100/tests/test.sh b/tasks/task_100/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_100/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_101/environment/Dockerfile b/tasks/task_101/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..b634475cf692d198897756cbfd6d3828a0b6567a --- /dev/null +++ b/tasks/task_101/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_101 +COPY repo/ /app/task_101/source/ +COPY public/ /app/task_101/ +COPY requirements.lock /app/task_101/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_101:/app/task_101/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_101/instruction.md b/tasks/task_101/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..bf41e2deef126b28e02377c3382ca7c45877bd7d --- /dev/null +++ b/tasks/task_101/instruction.md @@ -0,0 +1,22 @@ +# Repair an inconsistent battery-model reproduction + +A battery-model workflow in the supplied package does not complete under one of +the documented model configurations exercised by the public reproduction. Run the +public reproduction, inspect the source snapshot and the background material, and +repair the implementation so the documented workflows produce finite, physically +plausible results. + +The repair must generalize beyond the supplied example. Do not hard-code the +fixture, a single parameter set, or a fixed output trace. + +Work only in this task directory. + +Use: + +```bash +python reproduce.py +``` + +The script reports whether the public workflow completed, whether it hit the +expected scientific failure before repair, or whether an infrastructure problem +prevented the check from running. diff --git a/tasks/task_101/metadata.json b/tasks/task_101/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..5cf637a5df703a02b9f6ffabe963a7030ec4f6b9 --- /dev/null +++ b/tasks/task_101/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "101", + "title": "Repair an inconsistent battery-model reproduction", + "domain": "lithium-ion-battery-modeling", + "language": "python", + "source_repository": "https://github.com/pybamm-team/PyBaMM", + "source_commit": "1e9fbcea25b78a0a633face62b077dae1f92bf62", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE.txt", + "gpl_family": false, + "public_payload_sha256": "7ca613711ba2cbfcb078328bc5fe977dda892c408974f1d7140f7f1d7d0228b3" +} diff --git a/tasks/task_101/task.toml b/tasks/task_101/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..0a2f6af81695865c43d4f7460e1fc281aa0a12dc --- /dev/null +++ b/tasks/task_101/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-101" +description = "Repair an inconsistent battery-model reproduction" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "101" +language = "python" +base_commit_hash = "1e9fbcea25b78a0a633face62b077dae1f92bf62" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-101:v0.1.0@sha256:fd4bfc3ad2bcdaa12f1e2c7f8451646b14175f70bfd76a5066b06fbc8fa775b6" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_101" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-101:v0.1.0@sha256:d87a3417f48574040b26cf616241f4f448f766aac2c04b1945ad05392ec04840" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_101" + +[[verifier.collect]] +command = "cd /app/task_101 && git config --global --add safe.directory /app/task_101 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_101/tests/Dockerfile b/tasks/task_101/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..1f74a47758ca22a117ab97c68f05b470982cae80 --- /dev/null +++ b/tasks/task_101/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-101:v0.1.0@sha256:d87a3417f48574040b26cf616241f4f448f766aac2c04b1945ad05392ec04840 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_101/tests/test.sh b/tasks/task_101/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_101/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_102/environment/Dockerfile b/tasks/task_102/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..50ecbb34a39bdd977f89bdf930a28ca770d1598a --- /dev/null +++ b/tasks/task_102/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_102 +COPY repo/ /app/task_102/source/ +COPY public/ /app/task_102/ +COPY requirements.lock /app/task_102/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_102:/app/task_102/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_102/instruction.md b/tasks/task_102/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..07efe4b9af4b87b3f310b5fbee8471920f96940b --- /dev/null +++ b/tasks/task_102/instruction.md @@ -0,0 +1,25 @@ +# Repair an inconsistent polymer coordinate workflow + +A valid Polyply workflow using the supplied polymer fixture does not finish +cleanly on the provided source snapshot. The command can read the input files, +but the coordinate-generation run halts before it produces the expected output +coordinates. + +Inspect the paper, the documentation, and the source snapshot; run the public +reproduction; and repair the implementation so the workflow completes on the +supplied fixture and remains correct for other valid inputs of the same +kind. Do not hard-code the supplied fixture or replace the scientific +workflow with a fixed output. + +Work only in this task directory. The evaluation environment is offline. You +may modify files under `source/`. + +Use: + +```bash +python reproduce.py +``` + +The command returns `1` when the public workflow hits the reproduced scientific +failure, `0` after a successful repair, and `2` for a runner or environment +failure. diff --git a/tasks/task_102/metadata.json b/tasks/task_102/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..28658c0d9e3ef2e70b0338177e1f1c1f1211e811 --- /dev/null +++ b/tasks/task_102/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "102", + "title": "Repair an inconsistent polymer coordinate workflow", + "domain": "polymer-coordinate-generation", + "language": "python", + "source_repository": "https://github.com/marrink-lab/polyply_1.0", + "source_commit": "160e9707d844ad1b108e1d8184152050d16088a0", + "source_license": "Apache-2.0", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "2152837000fbb33fc8d41e69096794de10a4c9bfe7e3ca4f6237362e5d33a560" +} diff --git a/tasks/task_102/task.toml b/tasks/task_102/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..fe8fe62f5594072f01c8cb17868f8cf8e1e976e9 --- /dev/null +++ b/tasks/task_102/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-102" +description = "Repair an inconsistent polymer coordinate workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "102" +language = "python" +base_commit_hash = "160e9707d844ad1b108e1d8184152050d16088a0" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-102:v0.1.0@sha256:961828d6d0051e381646dc093bc531e80659fb62b7bda3181e6d624aa381efba" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_102" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-102:v0.1.0@sha256:efc0d2d6b049ecc34b2272455392e8eaec4e82f52d340a1b944597cbbb00342b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_102" + +[[verifier.collect]] +command = "cd /app/task_102 && git config --global --add safe.directory /app/task_102 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_102/tests/Dockerfile b/tasks/task_102/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..3c8e5241a0b5aa5d63cb755c5a2ede30f524df68 --- /dev/null +++ b/tasks/task_102/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-102:v0.1.0@sha256:efc0d2d6b049ecc34b2272455392e8eaec4e82f52d340a1b944597cbbb00342b +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_102/tests/test.sh b/tasks/task_102/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_102/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_103/environment/Dockerfile b/tasks/task_103/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..e935f5b9ce79eec471bb3410ed8db2c823108376 --- /dev/null +++ b/tasks/task_103/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_103 +COPY repo/ /app/task_103/source/ +COPY public/ /app/task_103/ +COPY requirements.lock /app/task_103/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_103:/app/task_103/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_103/instruction.md b/tasks/task_103/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..f6556ae1c88c1d81a77ffe35128f87c171bb3d01 --- /dev/null +++ b/tasks/task_103/instruction.md @@ -0,0 +1,26 @@ +# Repair a symmetric composite laminate ABD calculation + +A composite-shell property workflow computes ABD matrices for laminated +plates. A symmetric laminate and the same ply stack written explicitly should +carry the same physical stiffness meaning through the supported workflow, but +the supplied package does not preserve that behavior. + +Reproduce the issue, inspect the source and public workflow, and repair the +implementation so valid laminate descriptions retain their scientific meaning +across supported symmetric and explicit layup calculations. The repair must +generalize beyond the supplied case. Do not special-case the fixture, hardcode +matrices, or change the public reproduction to hide a failure. + +Work only under this task directory. The evaluation environment is Linux and +has no network access. You may modify files under `source/`; do not modify +the task statement, reproduction, fixtures, or evaluation infrastructure. + +Run the public smoke test with: + +```bash +python reproduce.py +``` + +The command returns `1` when the untouched source reproduces the laminate +consistency failure, `0` after a successful repair, and `2` for a runner or +environment failure. diff --git a/tasks/task_103/metadata.json b/tasks/task_103/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..cd8b1cc8f002554b0a0ad739b35253ebe971dc4f --- /dev/null +++ b/tasks/task_103/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "103", + "title": "Repair symmetric composite laminate ABD calculations", + "domain": "high-performance-fibers-and-composites", + "language": "python", + "source_repository": "https://github.com/SteveDoyle2/pyNastran", + "source_commit": "bdff2bb616847a058960c56328b68e36ab3764e3", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE.md", + "gpl_family": false, + "public_payload_sha256": "637633ba5229cbf0bf969d215196618fbee0a970d14266a1c9942dbef3eabc86" +} diff --git a/tasks/task_103/task.toml b/tasks/task_103/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..97e7235c23e9953784064e5bb6c87f1d37a0c5e7 --- /dev/null +++ b/tasks/task_103/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-103" +description = "Repair symmetric composite laminate ABD calculations" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "103" +language = "python" +base_commit_hash = "bdff2bb616847a058960c56328b68e36ab3764e3" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-103:v0.1.0@sha256:8b1934f764347452d72e5a136b8925daf6002803df163fdda7285f9757831467" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_103" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-103:v0.1.0@sha256:6eee1cc75c6043694dc04e4d6de95a09eb4300c684c9901cb64fde126651ef29" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_103" + +[[verifier.collect]] +command = "cd /app/task_103 && git config --global --add safe.directory /app/task_103 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_103/tests/Dockerfile b/tasks/task_103/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..4cc3b7be71bd59c68364cc3ebb9a6b34b74d1f2e --- /dev/null +++ b/tasks/task_103/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-103:v0.1.0@sha256:6eee1cc75c6043694dc04e4d6de95a09eb4300c684c9901cb64fde126651ef29 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_103/tests/test.sh b/tasks/task_103/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_103/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_104/environment/Dockerfile b/tasks/task_104/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..d900e7355a9744483fcd739b90489b6efff1e586 --- /dev/null +++ b/tasks/task_104/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_104 +COPY repo/ /app/task_104/source/ +COPY public/ /app/task_104/ +COPY requirements.lock /app/task_104/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_104:/app/task_104/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_104/instruction.md b/tasks/task_104/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..d83805a11cc230af6bcb1913e9026696002d6649 --- /dev/null +++ b/tasks/task_104/instruction.md @@ -0,0 +1,33 @@ +# Repair inconsistent finite-volume advection edge states + +An energy-materials transport workflow uses the supplied scientific package to +build one-dimensional finite-volume advection models. The models discretise a +cell-centered state, reconstruct edge states for the advective flux, and then +advance a transient source problem. The calculations finish with finite values, +but the edge-state observations and the transported profiles are not +consistent with the finite-volume semantics documented for this package. + +Study the source snapshot, public documentation assets, and public +reproduction. Trace how symbolic spatial operators become edge-centered +quantities during discretisation, and repair the source so that valid +upwind/downwind advection workflows are internally consistent with the package's +finite-volume boundary representation. The repair must generalize beyond the +provided mesh, tracer profile, boundary value, velocity direction, and source +term. + +Do not hard-code the supplied observations or replace the model with fixed +outputs. Work only under this task directory. You may modify files under +`source/`; do not modify the task statement, fixtures, workflow, reproduction, +or evaluation infrastructure. The evaluation environment is offline. + +Run the public diagnostic with: + +```bash +python reproduce.py +``` + +The command returns `0` when the finite-volume workflows produce finite, +nonempty scientific observations, `1` only when the workflow cannot produce +finite observations, and `2` for an import, dependency, path, or runner +failure. The public diagnostic is a workbench; the verifier performs +independent scientific consistency checks on broader cases. diff --git a/tasks/task_104/metadata.json b/tasks/task_104/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..9ce6fc619e3af8f67dab1ca0bb75ba06a2a9ca73 --- /dev/null +++ b/tasks/task_104/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "104", + "title": "Repair inconsistent finite-volume advection edge states", + "domain": "energy-materials-and-finite-volume-transport", + "language": "python", + "source_repository": "https://github.com/pybamm-team/PyBaMM", + "source_commit": "4cf09f36bb4afb42c661e8765e14b4527581eff7", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE.txt", + "gpl_family": false, + "public_payload_sha256": "b9b443a00cc7b5c3a4e14a66c2c1462517649ad3dda55c2db9c549ecf8a07e67" +} diff --git a/tasks/task_104/task.toml b/tasks/task_104/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..e3a8c158eb6a89980e1413138898775aefeb7287 --- /dev/null +++ b/tasks/task_104/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-104" +description = "Repair inconsistent finite-volume advection edge states" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "104" +language = "python" +base_commit_hash = "4cf09f36bb4afb42c661e8765e14b4527581eff7" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-104:v0.1.0@sha256:0d857f3eb3f235360133e7ac9f179871788984cc44ad877c33592de25b9a6d55" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_104" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-104:v0.1.0@sha256:255bdf4c710228c19cbdb1ae9add8b9740edc29edc129361dc83e13bd1c07e6e" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_104" + +[[verifier.collect]] +command = "cd /app/task_104 && git config --global --add safe.directory /app/task_104 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_104/tests/Dockerfile b/tasks/task_104/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..757b3b8800f31cfd36fe14831c59dbe95012834c --- /dev/null +++ b/tasks/task_104/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-104:v0.1.0@sha256:255bdf4c710228c19cbdb1ae9add8b9740edc29edc129361dc83e13bd1c07e6e +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_104/tests/test.sh b/tasks/task_104/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_104/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_105/environment/Dockerfile b/tasks/task_105/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..8e2145e8f870eaaac176fc7fc04feccb3b3eb571 --- /dev/null +++ b/tasks/task_105/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=1 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_105 +COPY repo/ /app/task_105/source/ +COPY public/ /app/task_105/ +COPY requirements.lock /app/task_105/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_105:/app/task_105/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_105/instruction.md b/tasks/task_105/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..823608d66eda2a906e777ddf6aa350b4b682359d --- /dev/null +++ b/tasks/task_105/instruction.md @@ -0,0 +1,21 @@ +# Restore a stereo-preserving tautomer workflow + +A cheminformatics workflow uses the supplied RDKit C++ source snapshot to +canonicalize tautomeric forms on a molecule that already carries stereochemical +annotations. One compact V3000 input loads successfully, but the +canonicalization step does not complete. + +Run the public reproduction, inspect the source snapshot and the bundled +references, and repair the implementation so that this workflow completes on +the supplied input and remains coherent on other stereo-bearing molecules. Do +not hard-code the fixture, the generated probe, or the public report. + +Work only under this task directory. The evaluation environment is Linux, +offline, and provides a C++ compiler, CMake, and Boost headers. You may modify +files under `source/`. + +Run the public diagnostic with: + +```bash +python reproduce.py +``` diff --git a/tasks/task_105/metadata.json b/tasks/task_105/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..60153957547daec58f8f531f60f7b2f976180324 --- /dev/null +++ b/tasks/task_105/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "105", + "title": "Restore a stereo-preserving tautomer workflow", + "domain": "drug-discovery-cheminformatics", + "language": "python-cpp", + "source_repository": "", + "source_commit": "", + "source_license": "BSD-3-Clause", + "license_source": "license.txt", + "gpl_family": false, + "public_payload_sha256": "d496a6d39aae48fa19d84c6a4abeb31ffdcb47c89055e7f92e5cd9cf6309608c" +} diff --git a/tasks/task_105/task.toml b/tasks/task_105/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..c1037202eb408b19d4ec23febac4c870377f3eda --- /dev/null +++ b/tasks/task_105/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-105" +description = "Restore a stereo-preserving tautomer workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "105" +language = "python-cpp" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-cpp-task-105:v0.1.0@sha256:c300da3c9e6851ebd49f1aaee9d30b11edfb3ff0d116687fd6d648d1dc237f9b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_105" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-cpp-task-105:v0.1.0@sha256:f1db6d0edacb4e211696f3e954c59bccbe2beb5ef28ea730dc73ee2bbb041d69" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_105" + +[[verifier.collect]] +command = "cd /app/task_105 && git config --global --add safe.directory /app/task_105 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_105/tests/Dockerfile b/tasks/task_105/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..ec6a884c018eaff305f1e21b1fdbc3c91f688b44 --- /dev/null +++ b/tasks/task_105/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-cpp-task-105:v0.1.0@sha256:f1db6d0edacb4e211696f3e954c59bccbe2beb5ef28ea730dc73ee2bbb041d69 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_105/tests/test.sh b/tasks/task_105/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_105/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_106/environment/Dockerfile b/tasks/task_106/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..e40aa891b1a65d77473da68496a099bfc2aeb506 --- /dev/null +++ b/tasks/task_106/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_106 +COPY repo/ /app/task_106/source/ +COPY public/ /app/task_106/ +COPY requirements.lock /app/task_106/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_106:/app/task_106/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_106/instruction.md b/tasks/task_106/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..4fe53c4c7246e9b142a25cd9f06c042eca53795d --- /dev/null +++ b/tasks/task_106/instruction.md @@ -0,0 +1,25 @@ +# Reproduce and extend a version-pinned MAP4/MAPC retrieval protocol + +You are working in a chemoinformatics codebase used to reproduce and extend a +published molecular-space retrieval benchmark. A methods-validation group has +re-run a small part of the workflow using an upstream peptide benchmark input +and a version-pinned MAP4/MAPC method contract. The current source snapshot +does not complete the requested offline observation. + +Run the public workbench first: + +```text +python reproduce.py +``` + +Treat its output as experimental evidence from a failed methods reproduction. +Use it together with the complete source snapshot, the raw benchmark input, +and the unmodified versioned method documents in `paper_assets/` to determine +which scientific behavior is not being preserved and repair the implementation. +The reference panel and query panel must remain comparable under the documented +protocol when the workflow is repeated or extended. + +Do not use the network or add a dependency. Do not special-case the benchmark +rows, expected output values, or report file. The implementation should +generalize to other molecules, widths, seeds, input orders, storage modes, and +stereochemical inputs covered by the public API. diff --git a/tasks/task_106/metadata.json b/tasks/task_106/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..712d2a66e3fa1c31c1093d18ca876f631888f70a --- /dev/null +++ b/tasks/task_106/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "106", + "title": "Reproduce and extend a version-pinned MAP4/MAPC retrieval protocol", + "domain": "chemoinformatics-molecular-fingerprints", + "language": "python", + "source_repository": "https://github.com/MLCIL/scikit-fingerprints", + "source_commit": "839001468ba8f8da64ba9d54b33fa33c795ed3c5", + "source_license": "MIT", + "license_source": "LICENSE.md", + "gpl_family": false, + "public_payload_sha256": "e94e60a0ebbd19494438edfdc820c664f7f9d4a72a7a4e9f4a5158e65019d996" +} diff --git a/tasks/task_106/task.toml b/tasks/task_106/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..ed56c025e31c7e46193d80b28457f551c75b91c5 --- /dev/null +++ b/tasks/task_106/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-106" +description = "Reproduce and extend a version-pinned MAP4/MAPC retrieval protocol" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "106" +language = "python" +base_commit_hash = "839001468ba8f8da64ba9d54b33fa33c795ed3c5" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-106:v0.1.0@sha256:06c0ff508ca79a201a203fbdba027bc4325e83a9d66ea6da91dc6edb5a3bfc70" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_106" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-106:v0.1.0@sha256:1cd49e20a456955be7217fb7d096e81562c9bbc081931055fd3e719ee9eaf111" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_106" + +[[verifier.collect]] +command = "cd /app/task_106 && git config --global --add safe.directory /app/task_106 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_106/tests/Dockerfile b/tasks/task_106/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..327e7c1673280eb6ca2872760f26fce574589910 --- /dev/null +++ b/tasks/task_106/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-106:v0.1.0@sha256:1cd49e20a456955be7217fb7d096e81562c9bbc081931055fd3e719ee9eaf111 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_106/tests/test.sh b/tasks/task_106/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_106/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_107/environment/Dockerfile b/tasks/task_107/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..acd2d882c29b2a814bf0501983f9c33c9a228f4d --- /dev/null +++ b/tasks/task_107/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_107 +COPY repo/ /app/task_107/source/ +COPY public/ /app/task_107/ +COPY requirements.lock /app/task_107/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_107:/app/task_107/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_107/instruction.md b/tasks/task_107/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..f64545502ed7c460c8c6c4c4e8eee7ddb5d3ed37 --- /dev/null +++ b/tasks/task_107/instruction.md @@ -0,0 +1,11 @@ +# Repair a public symbolic workflow + +A user is running a legal symbolic workflow through the public SymPy API. +The command completes, but the documented public workflow fails on the +supplied workbench cases. + +Run the public reproduction, inspect the provided source snapshot and public +documentation, and repair the implementation so the symbolic integration +workflow is consistent on the supplied workbench cases. The repair must +generalize beyond the supplied cases and must not hard-code the fixture or any +fixed output. diff --git a/tasks/task_107/metadata.json b/tasks/task_107/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..0f8bced295b7b33e92eeea4d46d839d3283ae05f --- /dev/null +++ b/tasks/task_107/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "107", + "title": "Repair a public symbolic workflow", + "domain": "computer-algebra", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "9938aee4065bf2710315908c516fc18766d810d19943994ca58551f7b178794a" +} diff --git a/tasks/task_107/task.toml b/tasks/task_107/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..be364fe4178c19e58604a7d70ea1721c78fccc3a --- /dev/null +++ b/tasks/task_107/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-107" +description = "Repair a public symbolic workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "107" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-107:v0.1.0@sha256:0d36982015454e5d48e5d6f24a8964813386dfc86bd7d2bb6b89c5183b18a289" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_107" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-107:v0.1.0@sha256:dfc6b902fd30de1a96bf83fc929da8a7fd05d2bdbbc9ca9b0a4dce53107363f9" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_107" + +[[verifier.collect]] +command = "cd /app/task_107 && git config --global --add safe.directory /app/task_107 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_107/tests/Dockerfile b/tasks/task_107/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..b83cc5af9f755df3158e82dc73713479cedc4dc6 --- /dev/null +++ b/tasks/task_107/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-107:v0.1.0@sha256:dfc6b902fd30de1a96bf83fc929da8a7fd05d2bdbbc9ca9b0a4dce53107363f9 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_107/tests/test.sh b/tasks/task_107/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_107/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_108/environment/Dockerfile b/tasks/task_108/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..dbf4e0e31b9abfc36da6c1362cd95ee2dd90a3c6 --- /dev/null +++ b/tasks/task_108/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_108 +COPY repo/ /app/task_108/source/ +COPY public/ /app/task_108/ +COPY requirements.lock /app/task_108/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_108:/app/task_108/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_108/instruction.md b/tasks/task_108/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..a51756a1fbcb4312fefd42b8f6cb0ca4a616fa8f --- /dev/null +++ b/tasks/task_108/instruction.md @@ -0,0 +1,15 @@ +# Restore stable Jacobi theta evaluations across difficult regimes + +The task-local `source/` tree contains a snapshot of mpmath. Work only inside that source tree. + +An offline reproduction script reports Jacobi theta observations for valid complex inputs. Some calls either fail to produce a value for legal inputs, or become unstable when the same mathematical quantity is repeated at higher working precision. + +Repair the implementation so these theta evaluations are stable for legal inputs and continue to support all four Jacobi theta kinds and derivatives with respect to `z`. + +Run: + +```bash +python reproduce.py +``` + +The public workflow is diagnostic and intentionally small. It should keep running offline and producing a finite JSON report after your repair. diff --git a/tasks/task_108/metadata.json b/tasks/task_108/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..38a3d91433065e9704c5de3de94b80874d234064 --- /dev/null +++ b/tasks/task_108/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "108", + "title": "Restore stable Jacobi theta evaluations across difficult regimes", + "domain": "computational-mathematics-special-functions", + "language": "python", + "source_repository": "https://github.com/mpmath/mpmath", + "source_commit": "5e683847c782a4c38ec71b8a88e6b1fbd903548f", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "444409e8fe5c224b3165318b5a161d72cb6c9781565eabad46a5c08fe59685dc" +} diff --git a/tasks/task_108/task.toml b/tasks/task_108/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..ee6f6743b5497bab366bb3ed5dbb7fb7ce68ef74 --- /dev/null +++ b/tasks/task_108/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-108" +description = "Restore stable Jacobi theta evaluations across difficult regimes" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "108" +language = "python" +base_commit_hash = "5e683847c782a4c38ec71b8a88e6b1fbd903548f" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-108:v0.1.0@sha256:27ee5433771c81d7c81fe8dd421f6be77525f2970022969db930273d2c361cc0" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_108" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-108:v0.1.0@sha256:c7d2a5a47ab0b41c8da2154568fc37a12cd8da33569e1e73c41ddd2556550bf5" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_108" + +[[verifier.collect]] +command = "cd /app/task_108 && git config --global --add safe.directory /app/task_108 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_108/tests/Dockerfile b/tasks/task_108/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..02a06f976ca498a4ac48974727f97c5232785bc6 --- /dev/null +++ b/tasks/task_108/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-108:v0.1.0@sha256:c7d2a5a47ab0b41c8da2154568fc37a12cd8da33569e1e73c41ddd2556550bf5 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_108/tests/test.sh b/tasks/task_108/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_108/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_109/environment/Dockerfile b/tasks/task_109/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..48b2d11abcc0228841a4db4911368455fd06314f --- /dev/null +++ b/tasks/task_109/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_109 +COPY repo/ /app/task_109/source/ +COPY public/ /app/task_109/ +COPY requirements.lock /app/task_109/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_109:/app/task_109/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_109/instruction.md b/tasks/task_109/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..1777dc2c1db1c183b48e6872d93b437d0f25ef09 --- /dev/null +++ b/tasks/task_109/instruction.md @@ -0,0 +1,12 @@ +# Repair paired wrinkle FE analysis + +Inspect the provided `wrinkleFE` source snapshot and run the public +reproduction. The current source does not complete the paired-wrinkle FE +observation. + +Repair the implementation so the documented public workflow runs end-to-end on +the supplied source snapshot and the reported FE observables remain finite for +the paired layout. + +Keep the task offline and deterministic. Do not change the public reproduction +fixture or hard-code the reported values. diff --git a/tasks/task_109/metadata.json b/tasks/task_109/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..451b0b301f0b8512a0c0b6027f93623314116ad7 --- /dev/null +++ b/tasks/task_109/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "109", + "title": "Repair paired wrinkle FE analysis", + "domain": "high-performance-fibers-and-composites", + "language": "python", + "source_repository": "https://github.com/elhajjar1/wrinkleFE", + "source_commit": "e81538d9a2ab5031559d0dd74b742ded5465406c", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "8570c909ae4c3289c01426c7f2720de40ea02ac9dd32e04d729be0a44d48ac5e" +} diff --git a/tasks/task_109/task.toml b/tasks/task_109/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..b0b6f6970889aa8ed7986555d6c9f15257461e69 --- /dev/null +++ b/tasks/task_109/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-109" +description = "Repair paired wrinkle FE analysis" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "109" +language = "python" +base_commit_hash = "e81538d9a2ab5031559d0dd74b742ded5465406c" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-109:v0.1.0@sha256:28e18df47f0a3b6a207c6c8ecfc9c13ecf92888acf3abd19c8bb172636b9934a" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_109" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-109:v0.1.0@sha256:401e19dafcba0c5e8d08b55e01eca2380927e41be82c2144d9e99e3a86968d9b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_109" + +[[verifier.collect]] +command = "cd /app/task_109 && git config --global --add safe.directory /app/task_109 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_109/tests/Dockerfile b/tasks/task_109/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..6ce757a32914ae6dd8ef043a67788e9c3a1da745 --- /dev/null +++ b/tasks/task_109/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-109:v0.1.0@sha256:401e19dafcba0c5e8d08b55e01eca2380927e41be82c2144d9e99e3a86968d9b +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_109/tests/test.sh b/tasks/task_109/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_109/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_110/environment/Dockerfile b/tasks/task_110/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..7fb31c0d3ed98348e0e4a7c1188068731f5bbf5e --- /dev/null +++ b/tasks/task_110/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_110 +COPY repo/ /app/task_110/source/ +COPY public/ /app/task_110/ +COPY requirements.lock /app/task_110/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_110:/app/task_110/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_110/instruction.md b/tasks/task_110/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..949de555b214f310af980c2058ac54e3cead33eb --- /dev/null +++ b/tasks/task_110/instruction.md @@ -0,0 +1,36 @@ +# Restore consistent cross-section analysis results + +A structural-analysis team is reviewing several valid connected cross-section +models. Normal meshing and engineering analyses complete with finite reports. +The supplied laboratory also derives each model's ordinary first moments of +area directly from its constructed outline. For material-bearing models, the +package's ordinary first-moment result is finite but incompatible with that +construction observation, without clearly reporting that the requested result +is unavailable. Results carried through ordinary workflows therefore cannot be +interpreted consistently in downstream beam calculations and research records. + +Reproduce the laboratory, inspect the supplied source and complete scientific +material, and repair the implementation so that ordinary result requests do +not silently return a physically incompatible quantity. Preserve valid analysis +and reporting behavior throughout supported workflows. The repair must +generalize beyond the supplied laboratory: do not special-case its geometry, +values, case labels, or printed output, and do not replace calculations with +fixed results. + +Work only under this task directory. The evaluation environment is Linux and has no network access. You may modify files under source/; do not modify the task statement, public evidence, or reproduction workflow. + +Run the public laboratory with: + +~~~bash +python reproduce.py +~~~ + +A finite first-moment result that is incompatible with its construction observation +reports `pre_fix_expected_failure` and exit code 1. After every model either reports a +matching first moment or explicitly reports that the ordinary result is unavailable, the +laboratory reports `post_fix_success` and exit code 0. `runner_failure` and exit code 2 +identify an execution problem rather than a scientific conclusion. + +For this public laboratory, an explicitly unavailable ordinary result is reported by a +`RuntimeError` or `ValueError` with a nonempty explanation. Other unexpected exceptions +remain execution failures. diff --git a/tasks/task_110/metadata.json b/tasks/task_110/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..4c557e2f20ba12962bebc1cd71855b1f888bfd14 --- /dev/null +++ b/tasks/task_110/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "110", + "title": "Restore consistent cross-section analysis results", + "domain": "high-performance-fibers-and-composites", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "5da464679f638b67d9409b9d680c9a4e5f7bb314f46ae0fc14a693500f22d5bb" +} diff --git a/tasks/task_110/task.toml b/tasks/task_110/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..e5a8afef6ad5517848daf0145c26d9c17539948c --- /dev/null +++ b/tasks/task_110/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-110" +description = "Restore consistent cross-section analysis results" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "110" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-110:v0.1.0@sha256:31221c81a3b4eecab2d2a678620d800acdb02f804093d809656ad695b4e667d6" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_110" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-110:v0.1.0@sha256:f73746797050455f54de4c5aef6c2d50ad43601623072479fda97d9d80bbece7" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_110" + +[[verifier.collect]] +command = "cd /app/task_110 && git config --global --add safe.directory /app/task_110 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_110/tests/Dockerfile b/tasks/task_110/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..ac6eed8808e4ca2f4e4eb6af49b5552ad0dadb5b --- /dev/null +++ b/tasks/task_110/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-110:v0.1.0@sha256:f73746797050455f54de4c5aef6c2d50ad43601623072479fda97d9d80bbece7 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_110/tests/test.sh b/tasks/task_110/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_110/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_111/environment/Dockerfile b/tasks/task_111/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..e31252d46932b96d545cb8302608c5a9c307d944 --- /dev/null +++ b/tasks/task_111/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_111 +COPY repo/ /app/task_111/source/ +COPY public/ /app/task_111/ +COPY requirements.lock /app/task_111/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_111:/app/task_111/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_111/instruction.md b/tasks/task_111/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..369d79fbf73b8586aa1881dcaa8255c8f134c930 --- /dev/null +++ b/tasks/task_111/instruction.md @@ -0,0 +1,47 @@ +# Restore consistent tapered-filament free response + +A filament-mechanics workflow uses the supplied scientific package to study +slender specimens whose radius changes along their length. Each specimen is +initially straight, free of imposed loads, and given a uniform velocity. A +single material response setting is then used to characterize its free +response. + +For tapered specimens, the recorded local response changes markedly with the +cross-section even though the material setting is common to the specimen. This +makes results from different manufactured taper profiles difficult to compare +and is not consistent with the intended material interpretation of that +setting. + +Use SI units for the supplied workbench: lengths and radii are in metres, +density is in kg/m^3, velocity is in m/s, and the package's material-response +setting is an inverse-time rate in s^-1. This task's rate convention is the one +used by the rod API; the paper also discusses other viscous coefficients with +different units, which are not substituted for the workbench setting. The +supplied workbench directly exercises one translational free-response path. The +package supports other free-response dynamics and assembled multi-specimen +workflows; a repair must preserve valid public interfaces and physically +meaningful behavior outside the three displayed profiles. + +Use the supplied source, paper, and reproducible workbench to repair the +implementation so that valid tapered-filament simulations retain physically +consistent free-response behavior across supported workflows. Preserve valid +public interfaces and scientific behavior outside the supplied example. Do not +hard-code the supplied profiles or output values, special-case the workbench, +or replace the simulation with fixed results. If an apparent local change does +not remain consistent when you vary valid geometry or motion, trace the model +and revise the repair rather than hiding the discrepancy. + +Work only under this task directory. The evaluation environment is Linux and +has no network access. You may modify files under `source/`; do not modify the +task statement, paper material, reproduction workflow, or evaluation +infrastructure. + +Run the public workbench with: + +```bash +python reproduce.py +``` + +The command writes a diagnostic observation report. A completed public +workflow shows that the experiment ran; it is not the final correctness +decision. diff --git a/tasks/task_111/metadata.json b/tasks/task_111/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..ad1e301d7d745094f0fe559c502770524d4659e1 --- /dev/null +++ b/tasks/task_111/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "111", + "title": "Restore consistent tapered-filament free response", + "domain": "computational-mechanics-and-soft-filament-dynamics", + "language": "python", + "source_repository": "https://github.com/GazzolaLab/PyElastica", + "source_commit": "46ad665fb84bda0c2d9b3392b6f2ef2388af1358", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "bc051a4f26a75f73535918cd1a8bc8754a4b1da712860acdff481ba9ff0efe8a" +} diff --git a/tasks/task_111/task.toml b/tasks/task_111/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..28928392bcf85dccc4569280dda2a69178beb1b1 --- /dev/null +++ b/tasks/task_111/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-111" +description = "Restore consistent tapered-filament free response" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "111" +language = "python" +base_commit_hash = "46ad665fb84bda0c2d9b3392b6f2ef2388af1358" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-111:v0.1.0@sha256:3ffa5e7326116f513a3a357db186b3c1d115e7b95328eec5a0c7646530eabf72" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_111" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-111:v0.1.0@sha256:3babb2ab453edfca93f09d4a4c2412a8eed02c686db64d4f122b656cb04be6cb" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_111" + +[[verifier.collect]] +command = "cd /app/task_111 && git config --global --add safe.directory /app/task_111 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_111/tests/Dockerfile b/tasks/task_111/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..02bd43646fe3eaadbfb4daa4cebbe7870ffc0ce1 --- /dev/null +++ b/tasks/task_111/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-111:v0.1.0@sha256:3babb2ab453edfca93f09d4a4c2412a8eed02c686db64d4f122b656cb04be6cb +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_111/tests/test.sh b/tasks/task_111/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_111/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_112/environment/Dockerfile b/tasks/task_112/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..1ee439e0af549d02dfcb655921492bda54b765b3 --- /dev/null +++ b/tasks/task_112/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_112 +COPY repo/ /app/task_112/source/ +COPY public/ /app/task_112/ +COPY requirements.lock /app/task_112/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_112:/app/task_112/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_112/instruction.md b/tasks/task_112/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..291c45e23ed2a0129e11abc43a375f14fe3b6329 --- /dev/null +++ b/tasks/task_112/instruction.md @@ -0,0 +1,14 @@ +I am evaluating a compact hybrid structural cross-section with constituents that +have different elastic and yield properties. The supplied run completes and reports +finite section results, but the plastic-capacity quantities for this valid +multi-material configuration are inconsistent with the behavior described in the +independent scientific material. + +Please inspect the supplied references and source snapshot, run the public +reproduction workflow, and repair the scientific implementation so that the +calculation is physically meaningful for supported valid inputs. Preserve the public +API and make the repair general rather than fitting the supplied example or its +material values. + +Keep the change focused, add or update regression tests where useful, and run the +relevant test suite plus the reproduction command before reporting the result. diff --git a/tasks/task_112/metadata.json b/tasks/task_112/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..056f24032857190eeed5e58903c737b36afc2b5c --- /dev/null +++ b/tasks/task_112/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "112", + "title": "Repair an inconsistent hybrid-section plastic capacity workflow", + "domain": "high-performance-fibers-and-composites", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "bb313b40e78959c6b6aa7c2e1d88bc13135cabb91be96ead6c7b02b595e09c91" +} diff --git a/tasks/task_112/task.toml b/tasks/task_112/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..21d44abd9b5e73239578ead0e54ad4ab31808104 --- /dev/null +++ b/tasks/task_112/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-112" +description = "Repair an inconsistent hybrid-section plastic capacity workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "112" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-112:v0.1.0@sha256:cf1d158d5fff8dc097eb9656e900a1db377e2d117cbcdd1f974abd1fd19c97d9" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_112" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-112:v0.1.0@sha256:30342a49f647f500a584e94e70157d8fde2a5f444c2d2c6f85eaf5cbf40ee421" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_112" + +[[verifier.collect]] +command = "cd /app/task_112 && git config --global --add safe.directory /app/task_112 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_112/tests/Dockerfile b/tasks/task_112/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..35d154649d925cb9f54d7534fb7276fb4e6adf1c --- /dev/null +++ b/tasks/task_112/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-112:v0.1.0@sha256:30342a49f647f500a584e94e70157d8fde2a5f444c2d2c6f85eaf5cbf40ee421 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_112/tests/test.sh b/tasks/task_112/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_112/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_113/environment/Dockerfile b/tasks/task_113/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..4835086c5739a496c242ebbf08e3bb0e900510a5 --- /dev/null +++ b/tasks/task_113/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_113 +COPY repo/ /app/task_113/source/ +COPY public/ /app/task_113/ +COPY requirements.lock /app/task_113/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_113:/app/task_113/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_113/instruction.md b/tasks/task_113/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..9e48cf6337ed82066580f389cc65444879789232 --- /dev/null +++ b/tasks/task_113/instruction.md @@ -0,0 +1,26 @@ +# Repair an inconsistent solar-resource workflow + +Inspect the supplied historical `pvlib` source snapshot and run the public +diagnostic. A fixed-site solar-resource calculation receives valid +timezone-aware timestamp series whose labels differ even though they describe +the same sampling instants. The calculation returns different physical +estimates for the two representations. + +Repair the implementation so equivalent valid aware representations produce +consistent observable solar-resource outputs and so the change generalizes to +supported scalar/series inputs and fixed-offset or named time zones. Preserve +public API shapes, output labels, and documented behavior for valid inputs +outside the observed discrepancy. + +Supported aware inputs in this task include pandas `Timestamp` and +`DatetimeIndex` values with fixed-offset or named time zones, in both scalar and +series forms. Parsing ambiguous or nonexistent naive local times is outside +the task scope. + +The supplied diagnostic is one small fixed-site clear-sky case. Your change +must generalize beyond those timestamps and values; do not special-case the +sample or hard-code any reported result. + +Keep the task offline and deterministic. Modify implementation code under +`source/` only. Do not edit `reproduce.py`, the paper material, generated +reports, or supplied data assets. diff --git a/tasks/task_113/metadata.json b/tasks/task_113/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..185ef26314b7d64c67d4b1ca1221ed5503a2f761 --- /dev/null +++ b/tasks/task_113/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "113", + "title": "Repair an inconsistent solar-resource workflow", + "domain": "renewable-energy-modelling", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "abff131018164171256d840b24b28f75874d438199c298730bb9e67a418069d1" +} diff --git a/tasks/task_113/task.toml b/tasks/task_113/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..2a1b36fdbb9c43c4d0b74e14e696efc7366b8f83 --- /dev/null +++ b/tasks/task_113/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-113" +description = "Repair an inconsistent solar-resource workflow" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "113" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-113:v0.1.0@sha256:ed2da0d993479cb7e2468edd003d7e0995b64c12f531af018f1e6342bd7b6298" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_113" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-113:v0.1.0@sha256:9cc086c8edab07c617306a43a0e2bf7e2e2f70003734287d40e84138fdd73c11" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_113" + +[[verifier.collect]] +command = "cd /app/task_113 && git config --global --add safe.directory /app/task_113 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_113/tests/Dockerfile b/tasks/task_113/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..a7326a76fb3d36b0c74862ea97a776b82a1d0d23 --- /dev/null +++ b/tasks/task_113/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-113:v0.1.0@sha256:9cc086c8edab07c617306a43a0e2bf7e2e2f70003734287d40e84138fdd73c11 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_113/tests/test.sh b/tasks/task_113/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_113/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_114/environment/Dockerfile b/tasks/task_114/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..a37e2e26bd69a26de7a18dc307fe77164e116eb8 --- /dev/null +++ b/tasks/task_114/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_114 +COPY repo/ /app/task_114/source/ +COPY public/ /app/task_114/ +COPY requirements.lock /app/task_114/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_114:/app/task_114/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_114/instruction.md b/tasks/task_114/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..87bc716877da4b7036f6023f8554d3bda2244916 --- /dev/null +++ b/tasks/task_114/instruction.md @@ -0,0 +1,19 @@ +# Repair an implausible storage history at a planning-epoch boundary + +An energy-system planning study uses the supplied historical PyPSA snapshot to +model one ordered storage experiment. The study's offline workbench completes, +but it reports an implausible availability transition around a labelled +chronology boundary. This creates an ambiguity in how much energy is actually +available for later demand. + +Use the supplied reproduction, scientific reference material, and project +source to investigate and repair the implementation. The resulting behavior +must be consistent with the documented storage models and must generalize to +valid storage configurations beyond the supplied experiment. + +With both per-period cyclic and initial-state conditions, cyclic governs the boundary; the initial state is not its predecessor. + +Run `python reproduce.py` from this task directory to inspect the experiment. +Do not hard-code the supplied fixture, report values, labels, dimensions, or +filenames. Do not modify the fixture, workflow, reproduction command, or +generated-report contract. diff --git a/tasks/task_114/metadata.json b/tasks/task_114/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..dee9c9eede25326926dbeb40d032011f60353a94 --- /dev/null +++ b/tasks/task_114/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "114", + "title": "Repair an implausible storage history at a planning-epoch boundary", + "domain": "power-system-energy-storage", + "language": "python", + "source_repository": "https://github.com/PyPSA/PyPSA", + "source_commit": "895bc26c721abbb826d7f8caf51587a66c69f15d", + "source_license": "MIT", + "license_source": "LICENSE.txt", + "gpl_family": false, + "public_payload_sha256": "11ed0f893a747a442a803de74162c64d83f5d2faaa768350901af033f8f20baf" +} diff --git a/tasks/task_114/task.toml b/tasks/task_114/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..ae2ebf0e8dda465d4fe93c848c9851949e840417 --- /dev/null +++ b/tasks/task_114/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-114" +description = "Repair an implausible storage history at a planning-epoch boundary" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "114" +language = "python" +base_commit_hash = "895bc26c721abbb826d7f8caf51587a66c69f15d" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-114:v0.1.0@sha256:d00e869c664e260a5f304cefea76ebc24b7771534511614557b49dabff377790" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_114" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-114:v0.1.0@sha256:413bf5a0fb0be52e587cf96314cfa625a9a55c8214506562b79d74aa5221074e" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_114" + +[[verifier.collect]] +command = "cd /app/task_114 && git config --global --add safe.directory /app/task_114 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_114/tests/Dockerfile b/tasks/task_114/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..9d2b266550f0db59819b8d689f1810578b6196de --- /dev/null +++ b/tasks/task_114/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-114:v0.1.0@sha256:413bf5a0fb0be52e587cf96314cfa625a9a55c8214506562b79d74aa5221074e +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_114/tests/test.sh b/tasks/task_114/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_114/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_115/environment/Dockerfile b/tasks/task_115/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..d0dedb3b0d4de3ba659b5d321ce5ea4d3ee74c87 --- /dev/null +++ b/tasks/task_115/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_115 +COPY repo/ /app/task_115/source/ +COPY public/ /app/task_115/ +COPY requirements.lock /app/task_115/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_115:/app/task_115/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_115/instruction.md b/tasks/task_115/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..2e5261c45cfe84a936f7832757d83ddfe5b8dd0d --- /dev/null +++ b/tasks/task_115/instruction.md @@ -0,0 +1,24 @@ +# Repair a static structural-analysis workflow + +The supplied Pynite source is an historical snapshot of a small structural +finite-element library. The public workbench builds two legal frame +configurations and records their static-analysis observations. + +One configuration is an ordinary restrained frame. For the other legal +configuration, the historical snapshot can +nevertheless finish and report non-finite, implausibly large, or +solver-dependent values. The official manual in `paper.md` supplies the +stability and solver definitions needed to interpret those observations. + +Repair the historical workflow so it no longer silently presents a misleading +static result for configurations that the documented model semantics cannot +support, while retaining finite, consistent results for supported structures. +Keep the public analysis interfaces and their documented options usable. Do +not replace the finite-element model with a fixture-specific answer or change +the public workbench to hide an observation. + +Start by running `python reproduce.py`; the JSON output is an observation +workbench rather than a reference answer. You may add your own small models +and tests while reasoning about the source. The final implementation should +remain valid for additional supported structural models while retaining the +documented scientific contract. diff --git a/tasks/task_115/metadata.json b/tasks/task_115/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..d73cf4ea379e28f27883c6a88e6c00ad0dde8b9a --- /dev/null +++ b/tasks/task_115/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "115", + "title": "Restore physically meaningful structural stability checks", + "domain": "structural-engineering-and-disaster-resilience", + "language": "python", + "source_repository": "https://github.com/JWock82/Pynite", + "source_commit": "147c7aa24ec30b1c6a4d4820446915380752877b", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "9fd0984b691e2f02e10254a1bd81bce154c5487a07bdd0b418feb9a291468abc" +} diff --git a/tasks/task_115/task.toml b/tasks/task_115/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..82d98ae6d91e2fe2f35d2d09ca115074ab651040 --- /dev/null +++ b/tasks/task_115/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-115" +description = "Restore physically meaningful structural stability checks" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "115" +language = "python" +base_commit_hash = "147c7aa24ec30b1c6a4d4820446915380752877b" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-115:v0.1.0@sha256:b1349937225fba8160651c9b9b218d1c8eef9c8c0f38e2a97620ac6f69d285e8" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_115" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-115:v0.1.0@sha256:a1b79491c7ee62a7541fe24e1053d2094f439d23b7b1fadff3f2086f9044b3ed" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_115" + +[[verifier.collect]] +command = "cd /app/task_115 && git config --global --add safe.directory /app/task_115 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_115/tests/Dockerfile b/tasks/task_115/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..ca347495d3665a89b182b95530e2fc5e5bc734ac --- /dev/null +++ b/tasks/task_115/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-115:v0.1.0@sha256:a1b79491c7ee62a7541fe24e1053d2094f439d23b7b1fadff3f2086f9044b3ed +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_115/tests/test.sh b/tasks/task_115/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_115/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_116/environment/Dockerfile b/tasks/task_116/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..4f686d12c436437a18270986d5ee2cea03d607ca --- /dev/null +++ b/tasks/task_116/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_116 +COPY repo/ /app/task_116/source/ +COPY public/ /app/task_116/ +COPY requirements.lock /app/task_116/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_116:/app/task_116/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_116/instruction.md b/tasks/task_116/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..7a78eccd6c0fd59f65a2a3f620b870ab566006a2 --- /dev/null +++ b/tasks/task_116/instruction.md @@ -0,0 +1,12 @@ +# Structural analysis repair + +You are given a small, offline 3-D frame-analysis workbench in `reproduce.py` and the task-local snapshot under `source/`. The workbench contains a framed-structure case and a control case. It runs each case through both matrix-storage modes and prints bounded observations; use those observations, the entered topology, and the method references to determine what behavior is physically meaningful. + +Restore the library's documented static-analysis contract for the entered finite-element models: + +- the control case must continue to return finite, equilibrium-consistent displacements and reactions; +- each entered model must produce a physically interpretable outcome, and a valid stable response must not be changed into a misleading result; +- when a response is reported, it must remain consistent when the same model is solved through either matrix-storage mode and when supported model data or numerical options are varied; +- callers using the library's documented checking options must retain their documented behavior. + +Keep the API usable by the existing workbench and avoid changing the meaning of valid stable analyses. Preserve the documented API semantics. Do not add network access, new external data, or fixture-specific name checks. Use the method references in `paper.md` to justify the behavior. Run the public reproduction repeatedly while developing; its output is an observation aid, not the complete acceptance test. diff --git a/tasks/task_116/metadata.json b/tasks/task_116/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..5e8e521b9ecc8ae8b5c4af63049cafed48cb8736 --- /dev/null +++ b/tasks/task_116/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "116", + "title": "Repair mode-dependent structural analysis observations", + "domain": "structural-engineering-finite-element-resilience", + "language": "python", + "source_repository": "https://github.com/JWock82/Pynite", + "source_commit": "", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "dbc513a429a96f2f932c6361a67d673a4473c834f540432b3c60d51c91cfcc9d" +} diff --git a/tasks/task_116/task.toml b/tasks/task_116/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..1cafea93dee7ba09d0049f88bc37e45bb916cc28 --- /dev/null +++ b/tasks/task_116/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-116" +description = "Repair mode-dependent structural analysis observations" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "116" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-116:v0.1.0@sha256:915ef0a612a06d17d0dae56aa1481524f7ff3a522e8373acf846faf6e5bd9423" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_116" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-116:v0.1.0@sha256:0f6df9acdb29e887cbcce79b2fe207237fee4edfbf825459002da778c7ccf90f" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_116" + +[[verifier.collect]] +command = "cd /app/task_116 && git config --global --add safe.directory /app/task_116 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_116/tests/Dockerfile b/tasks/task_116/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..91a0588234d325f8ef703450aa63a9b9030d38db --- /dev/null +++ b/tasks/task_116/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-116:v0.1.0@sha256:0f6df9acdb29e887cbcce79b2fe207237fee4edfbf825459002da778c7ccf90f +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_116/tests/test.sh b/tasks/task_116/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_116/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_117/environment/Dockerfile b/tasks/task_117/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..689710db4b89ce3f2aa988321f2e50c3aeac6710 --- /dev/null +++ b/tasks/task_117/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_117 +COPY repo/ /app/task_117/source/ +COPY public/ /app/task_117/ +COPY requirements.lock /app/task_117/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_117:/app/task_117/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_117/instruction.md b/tasks/task_117/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..92e688f468c20ce63879a331cdee7542a317a1b0 --- /dev/null +++ b/tasks/task_117/instruction.md @@ -0,0 +1,31 @@ +# Repair inconsistent lateral-frame member observations + +A structural-engineering group is replaying a small lateral-load frame used to +compare alternative bracing layouts for earthquake and wind resilience. The +model is valid, the nonlinear analysis completes, and the node and member +quantities are finite. However, one supported member report is not consistent +with the geometry implied by the solved frame, and the scalar and sampled forms +of that report do not agree. + +Run the offline reproduction, inspect the complete source snapshot and the +accompanying official project material, and repair the implementation so that +supported frame analyses report physically consistent member observations. The +repair must generalize to other valid geometries, member orientations, load +signs, unilateral-element states, physical-member subdivisions, and public +result-query paths. Preserve ordinary active-member force, bending, and +second-order behavior. + +Do not hard-code the displayed frame, member names, node displacements, or +reported values. Do not modify the task statement, public reproduction, +scientific material, or evaluation infrastructure. Work only under this task +directory and keep the workflow offline. + +Use: + +```bash +python reproduce.py +``` + +The command writes a finite JSON observation report under `outputs/`. Its +`workflow_completed` status means that the scientific workflow ran and exposed +surface observations; it is not the final correctness decision. diff --git a/tasks/task_117/metadata.json b/tasks/task_117/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..09e5c299982fd94f2a943b93b62f6267b6386840 --- /dev/null +++ b/tasks/task_117/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "117", + "title": "Repair inconsistent lateral-frame member observations", + "domain": "structural-engineering-disaster-resilience", + "language": "python", + "source_repository": "https://github.com/JWock82/Pynite", + "source_commit": "573e51689fe5fe54f279a7a875e431a35355fe73", + "source_license": "MIT", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "576548fdac2012b5964ab55f374077468fdce4afddd9599cfbf6301b1f3ea490" +} diff --git a/tasks/task_117/task.toml b/tasks/task_117/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..5a2024949ff3ce8b28dc46837205bc8f45fb4f9a --- /dev/null +++ b/tasks/task_117/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-117" +description = "Repair inconsistent lateral-frame member observations" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "117" +language = "python" +base_commit_hash = "573e51689fe5fe54f279a7a875e431a35355fe73" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-117:v0.1.0@sha256:194616d6aabade2a36d9d1076a2628f50790c9a88beccd353de1a508b3a5f417" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_117" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-117:v0.1.0@sha256:4bff5cf1692b605c6de4c67fcaa834de3c63bcece3857dfb0074aee990b8a7ca" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_117" + +[[verifier.collect]] +command = "cd /app/task_117 && git config --global --add safe.directory /app/task_117 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_117/tests/Dockerfile b/tasks/task_117/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..48efb7fc4a0c1b5b688e7879dbeadc2f7b052360 --- /dev/null +++ b/tasks/task_117/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-117:v0.1.0@sha256:4bff5cf1692b605c6de4c67fcaa834de3c63bcece3857dfb0074aee990b8a7ca +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_117/tests/test.sh b/tasks/task_117/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_117/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_118/environment/Dockerfile b/tasks/task_118/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..c5971a96cd22e7c5d3f137faa71473e43455b16e --- /dev/null +++ b/tasks/task_118/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_118 +COPY repo/ /app/task_118/source/ +COPY public/ /app/task_118/ +COPY requirements.lock /app/task_118/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_118:/app/task_118/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_118/instruction.md b/tasks/task_118/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..b121f2850b03d8fa9efbb610ac557075e8ebdca2 --- /dev/null +++ b/tasks/task_118/instruction.md @@ -0,0 +1,17 @@ +# Repair the periodic molecular calculation + +A small research group uses the supplied high-level Python interface to compare +single-point energies for a heteroatom-rich molecular model in several valid +periodic-cell representations. The calculations finish and return finite +values, but the results are not reliable enough for a scientific workflow. + +Work only with the offline source snapshot in `source/`. Use the public +reproduction command and the included method paper and API documentation to +understand the observations. Repair the implementation so that equivalent +valid periodic representations give consistent scientific results while +preserving the public interfaces and units. + +Do not change `reproduce.py` or hard-code its numbers. Keep the fix general: +the same contract should hold beyond the displayed observation. Stop and +report a runner failure if the offline build or calculation cannot be completed +rather than hiding it with fabricated output. diff --git a/tasks/task_118/metadata.json b/tasks/task_118/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..6593e9a5d0bca3bf23e106bcd715ac37cfb590bc --- /dev/null +++ b/tasks/task_118/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "118", + "title": "Repair inconsistent periodic molecular energy representations", + "domain": "computational-high-energy-molecular-chemistry-periodic-models", + "language": "python", + "source_repository": "", + "source_commit": "", + "source_license": "AGPL-family", + "license_source": "COPYING", + "gpl_family": true, + "public_payload_sha256": "1d2f2929496088cc56e5ace00807c26de51b2ca99b758ced06ebe7b6f8a057e1" +} diff --git a/tasks/task_118/task.toml b/tasks/task_118/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..494a238fe160dac3bca3aa49c56a6d1cc7fc7ecc --- /dev/null +++ b/tasks/task_118/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-118" +description = "Repair inconsistent periodic molecular energy representations" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "118" +language = "python" +base_commit_hash = "" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-118:v0.1.0@sha256:d368971b40f90dbf4197b0a285dd19086c66fb8287bc96f8c4601be5b2c0f95b" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_118" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-118:v0.1.0@sha256:1e5311b58eae5b34988b4a7b193b50e3dae7e6037605b1deaeb2c854a3de4cd5" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_118" + +[[verifier.collect]] +command = "cd /app/task_118 && git config --global --add safe.directory /app/task_118 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_118/tests/Dockerfile b/tasks/task_118/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..4137b3f2da9bb0d1000128a7ddc1c8fc4767ae67 --- /dev/null +++ b/tasks/task_118/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-118:v0.1.0@sha256:1e5311b58eae5b34988b4a7b193b50e3dae7e6037605b1deaeb2c854a3de4cd5 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_118/tests/test.sh b/tasks/task_118/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_118/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tasks/task_119/environment/Dockerfile b/tasks/task_119/environment/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..92bba58fff864dc7913f8fd9a9697f65e675a30c --- /dev/null +++ b/tasks/task_119/environment/Dockerfile @@ -0,0 +1,60 @@ +ARG BASE_IMAGE=python:3.11-slim +FROM ${BASE_IMAGE} + +ARG DEBIAN_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian +ARG DEBIAN_SECURITY_MIRROR=https://mirrors.tuna.tsinghua.edu.cn/debian-security +ARG PIP_INDEX_URL=https://pypi.tuna.tsinghua.edu.cn/simple +ARG INSTALL_FORTRAN=0 +ARG INSTALL_OCTAVE=0 +ENV DEBIAN_FRONTEND=noninteractive \ + PIP_DISABLE_PIP_VERSION_CHECK=1 \ + PIP_INDEX_URL=${PIP_INDEX_URL} \ + PIP_NO_CACHE_DIR=1 \ + PYTHONDONTWRITEBYTECODE=1 + +RUN sed -i \ + -e "s|http://deb.debian.org/debian-security|${DEBIAN_SECURITY_MIRROR}|g" \ + -e "s|http://deb.debian.org/debian|${DEBIAN_MIRROR}|g" \ + /etc/apt/sources.list.d/debian.sources + +RUN apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends \ + build-essential \ + ca-certificates \ + curl \ + git \ + ripgrep \ + && rm -rf /var/lib/apt/lists/* + +RUN if [ "${INSTALL_FORTRAN}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends gfortran \ + && rm -rf /var/lib/apt/lists/*; \ + fi \ + && if [ "${INSTALL_OCTAVE}" = "1" ]; then \ + apt-get -o Acquire::Retries=10 update \ + && apt-get -o Acquire::Retries=10 install -y --no-install-recommends octave \ + && rm -rf /var/lib/apt/lists/*; \ + fi + +WORKDIR /app/task_119 +COPY repo/ /app/task_119/source/ +COPY public/ /app/task_119/ +COPY requirements.lock /app/task_119/requirements.lock + +RUN python -m pip install --no-cache-dir --upgrade pip \ + && python -m pip install --no-cache-dir -r requirements.lock +ENV PYTHONPATH=/app/task_119:/app/task_119/source + +RUN git init \ + && git checkout -b main \ + && git config user.email science-bench@example.invalid \ + && git config user.name "SWE-bench Science baseline" \ + && git add -A \ + && GIT_AUTHOR_DATE=2025-01-01T00:00:00Z \ + GIT_COMMITTER_DATE=2025-01-01T00:00:00Z \ + git commit -m baseline \ + && git reflog expire --expire=now --all \ + && git gc --prune=now + +CMD ["bash"] diff --git a/tasks/task_119/instruction.md b/tasks/task_119/instruction.md new file mode 100644 index 0000000000000000000000000000000000000000..d928284f1a836dbd08625858fbfd651e7cb497a1 --- /dev/null +++ b/tasks/task_119/instruction.md @@ -0,0 +1,25 @@ +# Repair a molecular curvature workflow for reaction-path analysis + +A chemistry group is replaying a small, archived curvature calculation while +checking a high-energy molecular reaction path. The supplied molecular +geometry and electronic-structure output are valid, and the archived record +contains finite numeric values, but the local curvature used +for vibrational and transition-state analysis is not recovered as a complete +scientific object. + +Run the offline reproduction, read the accompanying method material, and study +the complete source snapshot. Repair the implementation so that valid +curvature outputs are represented consistently for supported molecular sizes +and valid calculations. The repair must generalize to other molecules, matrix +dimensions, and valid calculations; do not hard-code the supplied geometry, +matrix, eigenvalues, or report. + +Work only in this task directory. The evaluation environment has no network +access. You may modify files under `source/`; do not modify the task +statement, public reproduction, fixtures, or evaluation infrastructure. + +Run the diagnostic with: + +```bash +python reproduce.py +``` diff --git a/tasks/task_119/metadata.json b/tasks/task_119/metadata.json new file mode 100644 index 0000000000000000000000000000000000000000..679db50693d8f6f3996f22ba69d783b97bb62353 --- /dev/null +++ b/tasks/task_119/metadata.json @@ -0,0 +1,12 @@ +{ + "task_id": "119", + "title": "Repair a molecular curvature workflow for reaction-path analysis", + "domain": "quantum-chemistry-reaction-paths", + "language": "python", + "source_repository": "https://github.com/MolSSI/QCEngine", + "source_commit": "dd5eec6bfefa11ffe0ec81e2c4018003e8e39f2b", + "source_license": "BSD-3-Clause", + "license_source": "LICENSE", + "gpl_family": false, + "public_payload_sha256": "3871acb535cf909a57445869a10486318e8941c87fba318374acd51d562b1ee8" +} diff --git a/tasks/task_119/task.toml b/tasks/task_119/task.toml new file mode 100644 index 0000000000000000000000000000000000000000..cf3114cce2a0d9856b0a1a54082dcac5f8ff2adf --- /dev/null +++ b/tasks/task_119/task.toml @@ -0,0 +1,44 @@ +schema_version = "1.3" +artifacts = ["/logs/artifacts/model.patch"] + +[task] +name = "openmoss/swe-bench-science-119" +description = "Repair a molecular curvature workflow for reaction-path analysis" +authors = [] +keywords = ["scientific-computing", "software-engineering"] + +[metadata] +task_id = "119" +language = "python" +base_commit_hash = "dd5eec6bfefa11ffe0ec81e2c4018003e8e39f2b" + +[agent] +network_mode = "no-network" +timeout_sec = 5400.0 + +[environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-environment-python-task-119:v0.1.0@sha256:bd74671fe69b403be15607d8f221b5e2528cb8fac5ea76561542af3ac3421bdc" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_119" + +[verifier] +network_mode = "no-network" +environment_mode = "separate" +timeout_sec = 1800.0 + +[verifier.environment] +docker_image = "docker.io/kevinxulearning/swe-bench-science-verifier-python-task-119:v0.1.0@sha256:17807f4660a135d0d24a77e806d34f7fb1708117ab490bcdd0c4711ef5c5b560" +os = "linux" +allow_internet = false +cpus = 2 +memory_mb = 8192 +storage_mb = 20480 +workdir = "/app/task_119" + +[[verifier.collect]] +command = "cd /app/task_119 && git config --global --add safe.directory /app/task_119 && baseline=$(git rev-list --max-parents=0 --reverse HEAD | head -n 1) && git add -A && git diff --cached --binary \"$baseline\" > /logs/artifacts/model.patch" +timeout_sec = 300.0 diff --git a/tasks/task_119/tests/Dockerfile b/tasks/task_119/tests/Dockerfile new file mode 100644 index 0000000000000000000000000000000000000000..47dcc388d4095823f32526d82827a5adcc472fc8 --- /dev/null +++ b/tasks/task_119/tests/Dockerfile @@ -0,0 +1,2 @@ +ARG VERIFIER_IMAGE=docker.io/kevinxulearning/swe-bench-science-verifier-python-task-119:v0.1.0@sha256:17807f4660a135d0d24a77e806d34f7fb1708117ab490bcdd0c4711ef5c5b560 +FROM ${VERIFIER_IMAGE} diff --git a/tasks/task_119/tests/test.sh b/tasks/task_119/tests/test.sh new file mode 100644 index 0000000000000000000000000000000000000000..714959da692677fd78953da0372dfeb595111b19 --- /dev/null +++ b/tasks/task_119/tests/test.sh @@ -0,0 +1,6 @@ +#!/bin/sh +set -eu + +# Pier requires a public tests entrypoint when loading a local task. In +# separate-verifier mode the executable grader remains inside the verifier image. +exec python /tests/grader.py diff --git a/tools/materialize.py b/tools/materialize.py new file mode 100644 index 0000000000000000000000000000000000000000..8277960c7e9b4b8f92543dcb2858e5a3578c7eb4 --- /dev/null +++ b/tools/materialize.py @@ -0,0 +1,88 @@ +#!/usr/bin/env python3 +"""Materialize an explicit local Pier task selection from the release table.""" + +from __future__ import annotations + +import argparse +import json +import shutil +from pathlib import Path + + +ROOT = Path(__file__).resolve().parents[1] +GPL_IDS = {"003", "021", "023", "057", "066", "074", "075", "083", "084", "085", "100", "118"} + + +def normalize_task_id(value: str) -> str: + raw = value.strip().removeprefix("task_") + if not raw.isdigit() or not 1 <= int(raw) <= 119: + raise ValueError(f"task id must be in 001..119: {value!r}") + return f"{int(raw):03d}" + + +def load_rows() -> list[dict[str, object]]: + return [json.loads(line) for line in (ROOT / "manifests" / "tasks.jsonl").read_text(encoding="utf-8").splitlines() if line.strip()] + + +def main() -> int: + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--output", type=Path, default=ROOT / "tasks-selected") + parser.add_argument("--task-id", action="append", dest="task_ids") + parser.add_argument("--allow-GPL", action="store_true", dest="allow_gpl") + parser.add_argument( + "--include-build-context", + action="store_true", + help="Copy full environment source/build inputs instead of the thin HF task bundle", + ) + parser.add_argument("--force", action="store_true") + args = parser.parse_args() + selected = {normalize_task_id(value) for value in args.task_ids} if args.task_ids else None + rows = [] + for row in load_rows(): + task_id = str(row["release_id"]) + if selected is not None and task_id not in selected: + continue + if bool(row.get("gpl_family")) and not args.allow_gpl: + continue + rows.append(row) + if args.output.exists(): + if not args.force: + raise FileExistsError(f"output exists; use --force: {args.output}") + shutil.rmtree(args.output) + args.output.mkdir(parents=True) + for row in rows: + source = ROOT / str(row["task_path"]) + destination = args.output / source.name + if args.include_build_context: + if not (source / "environment" / "repo").is_dir(): + raise ValueError( + "full build context is absent from this snapshot; " + "use the thin prebuilt-image bundle" + ) + shutil.copytree(source, destination) + continue + destination.mkdir(parents=True) + for relative in ( + "task.toml", + "instruction.md", + "metadata.json", + "environment/Dockerfile", + "tests/Dockerfile", + "tests/test.sh", + ): + path = source / relative + if path.is_file(): + target = destination / relative + target.parent.mkdir(parents=True, exist_ok=True) + shutil.copy2(path, target) + selection = { + "allow_gpl": args.allow_gpl, + "task_ids": [str(row["release_id"]) for row in rows], + } + (args.output / "selection.json").write_text(json.dumps(selection, indent=2) + "\n", encoding="utf-8") + print(json.dumps(selection, indent=2)) + return 0 + + +if __name__ == "__main__": + raise SystemExit(main()) diff --git a/tools/pier_adapters.py b/tools/pier_adapters.py new file mode 100644 index 0000000000000000000000000000000000000000..a978e05a65191332a24501810670e083de7cd010 --- /dev/null +++ b/tools/pier_adapters.py @@ -0,0 +1,19 @@ +#!/usr/bin/env python3 +"""Small runtime-only Pier adapters shared by all task images.""" + +from __future__ import annotations + +from pier.agents.installed.codex import Codex + + +class ScienceBenchCodex(Codex): + """Use npm's optional platform package when Pier installs Codex.""" + + def install_spec(self): + spec = super().install_spec() + for step in spec.steps: + step.run = step.run.replace( + "npm install -g @openai/codex", + "npm install -g --include=optional @openai/codex", + ) + return spec diff --git a/tools/provider_config.py b/tools/provider_config.py new file mode 100644 index 0000000000000000000000000000000000000000..4000ca6ad71971936b44201a2f435649a906ed37 --- /dev/null +++ b/tools/provider_config.py @@ -0,0 +1,101 @@ +#!/usr/bin/env python3 +"""Resolve a Codex gateway profile without persisting credential values.""" + +from __future__ import annotations + +import json +import shlex +from collections.abc import Mapping +from dataclasses import dataclass +from pathlib import Path +from urllib.parse import urlsplit, urlunsplit + + +@dataclass(frozen=True) +class CodexProfile: + model: str + base_url: str + wire_api: str + version: str | None + reasoning_effort: str | None + + @property + def safe_base_url(self) -> str: + parts = urlsplit(self.base_url) + host = parts.hostname or "" + if parts.port: + host = f"{host}:{parts.port}" + return urlunsplit((parts.scheme, host, parts.path, "", "")) + + +def parse_dotenv(path: Path) -> dict[str, str]: + if not path.is_file(): + raise FileNotFoundError(path) + values: dict[str, str] = {} + for line in path.read_text(encoding="utf-8").splitlines(): + text = line.strip() + if not text or text.startswith("#"): + continue + if text.startswith("export "): + text = text[7:].lstrip() + if "=" not in text: + continue + key, value = text.split("=", 1) + key = key.strip() + if not key or not key.replace("_", "").isalnum() or key[0].isdigit(): + continue + value = value.strip() + if value and value[0] in "\"'" and value[-1:] == value[0]: + parts = shlex.split(value, comments=False, posix=True) + value = parts[0] if parts else "" + elif " #" in value: + value = value.split(" #", 1)[0].rstrip() + values[key] = value + return values + + +def resolve_codex_profile(environ: Mapping[str, str]) -> CodexProfile: + wire = environ.get("CODEX_WIRE_API", "responses").strip().lower().replace("-", "_") + aliases = { + "responses": "responses", + "openai_responses": "responses", + "chat": "chat", + "openai_chat": "chat", + } + if wire not in aliases: + raise ValueError("CODEX_WIRE_API must be responses or chat") + base_url = ( + environ.get("CODEX_BASE_URL") + or environ.get("OPENAI_BASE_URL") + or "https://api.openai.com/v1" + ).strip() + parts = urlsplit(base_url) + if parts.scheme not in {"http", "https"} or not parts.hostname: + raise ValueError("CODEX_BASE_URL must be an http(s) URL") + return CodexProfile( + model=(environ.get("MODEL") or "gpt-5").strip(), + base_url=base_url, + wire_api=aliases[wire], + version=(environ.get("CODEX_VERSION") or "").strip() or None, + reasoning_effort=( + environ.get("CODEX_REASONING_EFFORT") + or environ.get("REASONING_EFFORT") + or "" + ).strip() + or None, + ) + + +def render_codex_config(profile: CodexProfile) -> str: + return "\n".join( + [ + 'model_provider = "science_bench_gateway"', + "", + "[model_providers.science_bench_gateway]", + 'name = "SWE-bench Science Gateway"', + f"base_url = {json.dumps(profile.base_url)}", + f"wire_api = {json.dumps(profile.wire_api)}", + 'env_key = "OPENAI_API_KEY"', + "", + ] + ) diff --git a/tools/run_batch.py b/tools/run_batch.py new file mode 100644 index 0000000000000000000000000000000000000000..d4ecc5cd299d6ce3f4afa81d08ad70acb03799a6 --- /dev/null +++ b/tools/run_batch.py @@ -0,0 +1,234 @@ +#!/usr/bin/env python3 +"""Pull prebuilt task images and run an explicit local selection with Pier.""" + +from __future__ import annotations + +import argparse +import hashlib +import json +import os +import shlex +import subprocess +import sys +from pathlib import Path + +try: + from .provider_config import parse_dotenv, render_codex_config, resolve_codex_profile +except ImportError: # Direct execution: python3 scripts/run_batch.py + from provider_config import parse_dotenv, render_codex_config, resolve_codex_profile + +try: + import tomllib +except ModuleNotFoundError: # Python 3.10 and earlier + try: + import tomli as tomllib + except ModuleNotFoundError as exc: # pragma: no cover - depends on host Python + raise SystemExit( + "run_batch.py requires Python 3.11+ or the backport: " + "python3 -m pip install tomli" + ) from exc + + +def task_dirs(root: Path) -> list[Path]: + if root.name.startswith("task_") and (root / "task.toml").is_file(): + return [root] + return sorted( + (path for path in root.glob("task_*") if path.is_dir()), + key=lambda path: path.name, + ) + + +def load_image_refs(task_dir: Path) -> list[str]: + config = tomllib.loads((task_dir / "task.toml").read_text(encoding="utf-8")) + refs = [ + config.get("environment", {}).get("docker_image", ""), + config.get("verifier", {}).get("environment", {}).get("docker_image", ""), + ] + missing = [ref for ref in refs if not ref or "pending" in ref] + if missing: + raise ValueError(f"{task_dir.name} has unpublished image references") + return list(dict.fromkeys(refs)) + + +def pull_images(task_dirs_: list[Path], *, platform: str) -> list[str]: + refs: list[str] = [] + for task_dir in task_dirs_: + refs.extend(load_image_refs(task_dir)) + refs = list(dict.fromkeys(refs)) + for ref in refs: + command = ["docker", "pull", "--platform", platform, ref] + print("+ " + shlex.join(command), flush=True) + subprocess.run(command, check=True) + return refs + + +def selection_payload(root: Path, dirs: list[Path]) -> dict[str, object]: + selection_file = root / "selection.json" + if selection_file.is_file(): + payload = json.loads(selection_file.read_text(encoding="utf-8")) + task_ids = [str(value) for value in payload.get("task_ids", [])] + else: + task_ids = [path.name.removeprefix("task_") for path in dirs] + payload = {"allow_gpl": None, "task_ids": task_ids} + canonical = json.dumps({"task_ids": task_ids}, sort_keys=True).encode("utf-8") + payload["task_ids"] = task_ids + payload["selection_sha256"] = hashlib.sha256(canonical).hexdigest() + return payload + + +def redacted_command(command: list[str]) -> str: + redacted: list[str] = [] + index = 0 + while index < len(command): + value = command[index] + if value == "--agent-env" and index + 1 < len(command): + redacted.extend([value, ""]) + index += 2 + continue + if ( + value == "--agent-kwarg" + and index + 1 < len(command) + and command[index + 1].startswith("config_toml=") + ): + redacted.extend([value, "config_toml="]) + index += 2 + continue + redacted.append(value) + index += 1 + return shlex.join(redacted) + + +def main() -> int: + parser = argparse.ArgumentParser(description=__doc__) + parser.add_argument("--path", type=Path, required=True, help="Materialized task directory") + parser.add_argument("--agent", default="nop") + parser.add_argument("--env", default="docker") + parser.add_argument("--env-file", type=Path) + parser.add_argument("--model", action="append", default=[]) + parser.add_argument("--agent-env", action="append", default=[]) + parser.add_argument("--agent-kwarg", action="append", default=[]) + parser.add_argument("--n-concurrent", type=int, default=1) + parser.add_argument("--n-attempts", type=int, default=1) + parser.add_argument("--max-retries", type=int, default=0) + parser.add_argument("--agent-timeout-multiplier", type=float) + parser.add_argument("--verifier-timeout-multiplier", type=float) + parser.add_argument("--jobs-dir", type=Path, default=Path("jobs")) + parser.add_argument("--job-name") + parser.add_argument("--platform", default="linux/amd64") + parser.add_argument("--pier-bin", default="pier", help="Pier executable or absolute path") + parser.add_argument("--agent-import-path") + parser.add_argument("--skip-pull", action="store_true", help="Use images already present locally") + parser.add_argument( + "--no-auto-provider", + action="store_true", + help="Do not translate CODEX_* values from --env-file into native Pier kwargs", + ) + parser.add_argument( + "--no-auto-agent-adapter", + action="store_true", + help="Use Pier's built-in agent class instead of the runtime-only Codex adapter", + ) + parser.add_argument("--dry-run", action="store_true") + args = parser.parse_args() + + root = args.path.resolve() + dirs = task_dirs(root) + if not dirs: + raise ValueError(f"no task_NNN directories found under {root}") + selection = selection_payload(root, dirs) + image_refs: list[str] = [] + for task_dir in dirs: + image_refs.extend(load_image_refs(task_dir)) + image_refs = list(dict.fromkeys(image_refs)) + if not args.skip_pull: + pull_images(dirs, platform=args.platform) + models = list(args.model) + agent_kwargs = list(args.agent_kwarg) + agent_import_path = args.agent_import_path + if args.agent == "codex" and not args.no_auto_agent_adapter and not agent_import_path: + package = Path(__file__).resolve().parent.name + agent_import_path = f"{package}.pier_adapters:ScienceBenchCodex" + provider_metadata: dict[str, object] | None = None + if args.agent == "codex" and not args.no_auto_provider: + profile_env = dict(os.environ) + if args.env_file: + profile_env.update(parse_dotenv(args.env_file)) + profile = resolve_codex_profile(profile_env) + if not models: + models.append(profile.model) + if not any(value.startswith(("config_toml=", "config_toml_file=")) for value in agent_kwargs): + agent_kwargs.append("config_toml=" + render_codex_config(profile)) + # Pier normally strips a provider prefix before invoking Codex. Gateways + # may use that prefix for routing, so preserve the exact model identifier + # unless the caller supplied an explicit command override. + if not any(value.startswith("command_model_name=") for value in agent_kwargs): + agent_kwargs.append("command_model_name=" + profile.model) + if profile.version and not any(value.startswith("version=") for value in agent_kwargs): + agent_kwargs.append("version=" + profile.version) + if profile.reasoning_effort and not any( + value.startswith("reasoning_effort=") for value in agent_kwargs + ): + agent_kwargs.append("reasoning_effort=" + profile.reasoning_effort) + provider_metadata = { + "protocol": profile.wire_api, + "base_url": profile.safe_base_url, + "credential_env": "OPENAI_API_KEY", + } + + command = [ + args.pier_bin, "run", "--path", str(root), "--agent", args.agent, "--env", args.env, + "--n-concurrent", str(args.n_concurrent), "--n-attempts", str(args.n_attempts), + "--max-retries", str(args.max_retries), "--no-force-build", "--no-delete", "--yes", + ] + if args.agent_timeout_multiplier is not None: + command.extend(["--agent-timeout-multiplier", str(args.agent_timeout_multiplier)]) + if args.verifier_timeout_multiplier is not None: + command.extend(["--verifier-timeout-multiplier", str(args.verifier_timeout_multiplier)]) + if args.env_file: + command.extend(["--env-file", str(args.env_file)]) + if agent_import_path: + command.extend(["--agent-import-path", agent_import_path]) + for model in models: + command.extend(["--model", model]) + for value in args.agent_env: + command.extend(["--agent-env", value]) + for value in agent_kwargs: + command.extend(["--agent-kwarg", value]) + if args.jobs_dir: + command.extend(["--jobs-dir", str(args.jobs_dir)]) + if args.job_name: + command.extend(["--job-name", args.job_name]) + + metadata = { + "task_ids": selection["task_ids"], + "allow_gpl": selection.get("allow_gpl"), + "selection_sha256": selection["selection_sha256"], + "image_refs": image_refs, + "platform": args.platform, + "pier_command": redacted_command(command), + "agent_import_path": agent_import_path, + "provider": provider_metadata, + } + metadata_path = root / "batch-run.json" + metadata_path.write_text(json.dumps(metadata, indent=2) + "\n", encoding="utf-8") + print(json.dumps(metadata, indent=2, sort_keys=True)) + if args.dry_run: + return 0 + pier_environment = os.environ.copy() + # Pier may build an ephemeral environment+agent image. Keep that derived + # build on the same architecture as the prebuilt task images. + pier_environment["DOCKER_DEFAULT_PLATFORM"] = args.platform + tool_root = str(Path(__file__).resolve().parent.parent) + existing_pythonpath = pier_environment.get("PYTHONPATH", "") + pier_environment["PYTHONPATH"] = os.pathsep.join( + value for value in (tool_root, existing_pythonpath) if value + ) + return subprocess.run(command, check=False, env=pier_environment).returncode + + +if __name__ == "__main__": + try: + raise SystemExit(main()) + except (FileNotFoundError, ValueError, tomllib.TOMLDecodeError) as exc: + print(f"error: {exc}", file=sys.stderr) + raise SystemExit(2)