Output limits were baked into pi.mjs inside the runtime image, so every change (GLM 32000 -> 131072, Qwen 8192 -> 65536) needed a new image and a full 785-trial recalibration, and the Qwen baseline ran on an image the repo could not rebuild. - MODELS holds max_tokens and its wire field per model; pi_run passes --model and --max-tokens to the adapter, which has no defaults left. - Receipts record model_max_tokens and model_sampling=service-default; evidence requires every payload to carry exactly that limit, which would have caught the SDK's silent 32000 cap. Co-Authored-By: Claude Opus 5.5 <noreply@anthropic.com>
15 lines
871 B
Docker
15 lines
871 B
Docker
FROM python:3.10-slim
|
|
RUN apt-get update && apt-get install -y --no-install-recommends \
|
|
build-essential git ca-certificates xz-utils \
|
|
&& rm -rf /var/lib/apt/lists/*
|
|
RUN python3 -c "import urllib.request,hashlib; p='/tmp/node.tar.xz'; urllib.request.urlretrieve('https://nodejs.org/dist/v22.23.3/node-v22.23.3-linux-x64.tar.xz',p); assert hashlib.sha256(open(p,'rb').read()).hexdigest()=='df450af89261115ef9f9e3830c3eeb2cc9213b63c720b1af623cb5dcbe2e02de'" \
|
|
&& tar -xJf /tmp/node.tar.xz --strip-components=1 -C /usr/local \
|
|
&& rm /tmp/node.tar.xz
|
|
WORKDIR /opt/embedded-rollout/agent
|
|
COPY agent/package.json agent/package-lock.json ./
|
|
RUN npm ci --omit=dev --ignore-scripts --no-audit --no-fund && npm cache clean --force
|
|
COPY agent/pi.mjs ./
|
|
ENV HOME=/tmp/pi-home PI_CODING_AGENT_DIR=/tmp/pi-home/agent \
|
|
PYTHONDONTWRITEBYTECODE=1
|
|
WORKDIR /workspace
|