Spaces:
Running
Running
Download config.yaml from aisquared/bolt-hackathon-proxy: direct link, hf CLI and curl.
- Browser
- Download file 698 Bytes
-
https://huggingface.co/spaces/aisquared/bolt-hackathon-proxy/resolve/main/config.yaml
- Command line
-
hf download hf://spaces/aisquared/bolt-hackathon-proxy/config.yaml
-
curl -L -o config.yaml https://huggingface.co/spaces/aisquared/bolt-hackathon-proxy/resolve/main/config.yaml
698 Bytes
| model_list: | |
| - model_name: bolt-instruct-v2 | |
| litellm_params: | |
| model: openai/aisquared/bolt-instruct-v2 | |
| api_base: https://6ad27aecd2d0ab8296e58e32.endpoints.huggingface.cloud/v1 | |
| api_key: os.environ/HF_TOKEN | |
| # Generous default for students who don't set max_tokens themselves. | |
| # This is a DEFAULT, not a cap — a student's own max_tokens always wins, | |
| # up to the model's full 131K context. | |
| max_tokens: 16384 | |
| litellm_settings: | |
| callbacks: trace_logger.trace_logger | |
| drop_params: true | |
| general_settings: | |
| # One shared key for the whole hackathon — no database needed. | |
| # Students use this value as their API key. | |
| master_key: os.environ/MASTER_KEY | |