Spaces:

Himasree
/

sre-bench

Sleeping

App Files Files Community

87.8 kB

Ctrl+K

Ctrl+K

3 contributors

History: 19 commits

Himasree's picture

fix: use score= field and clamp scores to (0.01, 0.99) in [END] output

8cdab66 7 months ago

server
fix: harden Phase 2 scoring guardrails; align tests; pin Docker; fix edge-case expectations 7 months ago
.gitattributes

87 Bytes
first commit 7 months ago
.gitignore

3.11 kB
first commit 7 months ago
Dockerfile

264 Bytes
fix: Complete Phase 2 grader validation fixes 7 months ago
README.md

2.82 kB
Fix strict score bounds and align task config 7 months ago
app.py

1.23 kB
Rename server.py to app.py to fix import namespace conflict 7 months ago
inference.py

9.7 kB
fix: use score= field and clamp scores to (0.01, 0.99) in [END] output 7 months ago
openenv.yaml

1.02 kB
fix: reward_range [0.01, 0.99] — satisfy strict (0,1) grader constraint 7 months ago
pyproject.toml

577 Bytes
first commit 7 months ago
requirements.txt

169 Bytes
first commit 7 months ago
test.py

240 Bytes
Compliance fixes: app.py rename, auto-reset logic, and validate-submission script updates 7 months ago
test_all_tasks.py

13.4 kB
fix: reward never 0.0 — use 0.05 for intermediate steps, cache terminal score 7 months ago
test_api.py

583 Bytes
Compliance fixes: app.py rename, auto-reset logic, and validate-submission script updates 7 months ago
test_cumulative_reward.py

2.14 kB
fix: Complete Phase 2 grader validation fixes 7 months ago
test_grader.py

6.32 kB
fix: harden Phase 2 scoring guardrails; align tests; pin Docker; fix edge-case expectations 7 months ago
uv.lock

0 Bytes
first commit 7 months ago
validate-submission.sh

7.62 kB
fix: Complete Phase 2 grader validation fixes 7 months ago
validation_output.txt

1.59 kB
Compliance fixes: app.py rename, auto-reset logic, and validate-submission script updates 7 months ago