Skip to content

feat: Add Forge as LLM provider#34

Open
Yiiii0 wants to merge 1 commit intoLeapLabTHU:masterfrom
Yiiii0:feature/forge-20260307-234827
Open

feat: Add Forge as LLM provider#34
Yiiii0 wants to merge 1 commit intoLeapLabTHU:masterfrom
Yiiii0:feature/forge-20260307-234827

Conversation

@Yiiii0
Copy link
Copy Markdown

@Yiiii0 Yiiii0 commented Mar 8, 2026

Summary

Adds Forge as an OpenAI-compatible provider to LiveCodeBench runner. When FORGE_API_KEY is set, OpenAIRunner routes requests through Forge's API instead of OpenAI directly.

Changes

  • evaluation/code_eval/coding/LiveCodeBench/lcb_runner/lm_styles.py: Added LanguageModel entry for Forge with LMStyle.OpenAIChat
  • evaluation/code_eval/coding/LiveCodeBench/lcb_runner/runner/oai_runner.py: Refactored client initialization from class-level to instance-level _build_client() method that selects Forge or OpenAI based on env vars

Usage

export FORGE_API_KEY=your-key
# Optional: export FORGE_API_BASE=https://api.forge.tensorblock.co/v1
python -m lcb_runner.runner.main --model OpenAI/gpt-4o-mini

Tests

1 commit, 2 files changed. Existing behavior preserved when FORGE_API_KEY is not set.

I work at TensorBlock and will help maintain this integration.


About Forge

Forge is an open-source middleware service for unified AI model provider management. It routes requests across 40+ AI providers with access to thousands of models through a single OpenAI-compatible API.

## Changes

- Automated integration updates from manager loop.

Files modified:
evaluation/code_eval/coding/LiveCodeBench/lcb_runner/lm_styles.py
evaluation/code_eval/coding/LiveCodeBench/lcb_runner/runner/oai_runner.py
Sign up for free to join this conversation on GitHub. Already have an account? Sign in to comment

Labels

None yet

Projects

None yet

Development

Successfully merging this pull request may close these issues.

1 participant