Final_Assignment_Template / model_provider.py

Commit History

Fix Cerebras LiteLLM routing and include all providers in the chain.
74e9d85

ken2ki commited on

Use Cerebras first, then Google, then Groq on the Space.
ccf9702

ken2ki commited on

Rotate to Cerebras and Google when Groq limits are hit.
2abc385

ken2ki commited on

Add more free Groq models to the fallback rotation.
edf7a56

ken2ki commited on

Fix Groq fallback to use chat models that actually exist.
291a663

ken2ki commited on

Improve Space reliability when Groq limits are hit.
b09c490

ken2ki commited on

Stop sending unsupported num_ctx to Groq on the Space.
f7c43af

ken2ki commited on

Fix Groq model name for Space runs.
966fee7

ken2ki commited on

Add Groq as a free inference option on the Space.
583e697

ken2ki commited on

Ship the full GAIA agent for course submission.
3af908f

ken2ki commited on

Stabilize Ollama agent steps and handle missing benchmark files
4879dd8

ken2ki commited on

Run the GAIA agent locally with Ollama
4a95c80

ken2ki commited on