Skip to content

Make the runner image runnable: litellm proxy-deps + cost-tracking hint #50

New issue

Have a question about this project? Sign up for a free GitHub account to open an issue and contact its maintainers and the community.

Sign up for GitHub

By clicking “Sign up for GitHub”, you agree to our terms of service and privacy statement. We’ll occasionally send you account related emails.

Already on GitHub? Sign in to your account

Merged
GiggleLiu merged 1 commit into from
Jul 17, 2026
Merged

Make the runner image runnable: litellm proxy-deps + cost-tracking hint #50

Show file tree
Hide file tree
Changes from all commits
Commits
Show all changes
1 commit
Select commit
File filter

Filter by extension

Filter by extension
Conversations
Failed to load comments.
Loading
Jump to
Jump to file
Failed to load files.
Loading
Diff view
Diff view
6 changes: 6 additions & 0 deletions benchmark/requirements.txt
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -2,3 +2,9 @@ mini-swe-agent==2.4.5
pyyaml>=6.0.3
pytest>=9.1.1
pytest-timeout>=2.4.0
# litellm 1.92.0 (pulled by mini-swe-agent) unconditionally imports proxy-only
# deps from completion() (litellm.responses.mcp → litellm.proxy.*). Declare them
# until a litellm release makes that import lazy; litellm[proxy] is NOT an
# option (its granian pin is unresolvable).
fastapi>=0.115
orjson>=3.10
4 changes: 4 additions & 0 deletions submission.env.example
View file
Open in desktop
Original file line number Diff line number Diff line change
Expand Up @@ -26,6 +26,10 @@ MODEL_NAME=openai/gpt-5.4
# API_KEY=... # generic key when no PROVIDER_API_KEY name fits
# MODEL_KWARGS={"custom_llm_provider":"openai","extra_headers":{"X-Org":"acme"}}
# # ^ JSON of extra litellm.completion kwargs (api_version, temperature, …)
# MSWEA_COST_TRACKING=ignore_errors
# # ^ required when your model isn't in litellm's price map (typical for custom
# # endpoints) — mini-swe otherwise aborts the session on the cost-calculation
# # error. Token usage is still recorded from API responses.

# ── LIMITS ────────────────────────────────────────────────────────────────────
# MAX_TOKENS=8192 # per-call output ceiling
Expand Down