Spaces:
Sleeping
Sleeping
Commit History
-
Fix score range and [END] format for Phase 2 validation
Two issues: 1. Grader could return exactly 0.0 or 1.0; Phase 2 requires strictly (0, 1). Clamped compute_score output to max(0.01, min(0.99, raw)). 2. [END] stdout line was missing required score= field. Sample format: [END] success=... steps=... score=... rewards=... Added score tracking in _agent_loop and updated emit_end signature.07ca292 -
Fix inference.py crash when env vars not set
inference.py was calling sys.exit(1) when avg score < 0.1. Validator treats any non-zero exit as unhandled exception. Script should always exit 0 on normal completion regardless of scores.9eba455 -
Switch base image to AWS ECR Public mirror to avoid Docker Hub rate limits
The hackathon validator's build environment cannot reliably pull from registry-1.docker.io (repeated httpReadSeeker failures). public.ecr.aws is a rate-limit-free public mirror of the same official Docker Hub image.7f42f8e -
Use python:3.11-slim-bookworm for explicit base image tag
Avoids potential digest-resolution ambiguity with the floating python:3.11-slim tag during Docker Hub pulls.945dfa4 -
Add defaults for API_BASE_URL and MODEL_NAME per submission checklist
Checklist requires defaults for API_BASE_URL and MODEL_NAME but NOT for HF_TOKEN. Also tightened the missing-var check in main() to only gate on HF_TOKEN since the other two now have sensible defaults.02f8b38 -
Fix inference.py crash when env vars not set
Moved API_BASE_URL, MODEL_NAME, HF_TOKEN reads from module level (os.environ["KEY"] which raises KeyError) to os.environ.get() at module level. llm_client is now initialized inside main() after explicit validation. Missing vars print a usage message and exit 0 instead of crashing with an unhandled exception.a92cbbf
-
submission fixes
c2a35d8 -
Add required stdout format to inference.py
Emit [START]/[STEP]/[END] lines per OpenEnv spec. All human-readable output moved to stderr so stdout contains only the three structured line types expected by the automated evaluator.3b7ad8b
-
Fix inference Task4 hint, stale docstring, Dockerfile port comment
- inference.py: replace broken CREATE-INSERT-DROP-RENAME hint for Task 4 with correct ALTER TABLE RENAME COLUMN instructions; DROP TABLE on inventory/categories is an instant -1.0 safety violation - server/environment.py: update reset() docstring from seed%3/tasks 1-3 to seed%5/tasks 1-5 - Dockerfile: fix comment example port from 8000 to 7860c979cbb -
Add tasks 4-5, fix metadata, fix port, expand tests to 94
5e2a88b
-
Add uv.lock for multi-mode deployment validation
cd45cc6 -
gitignore thing
5663b4e -
Initial commit
822ca51