CheatsheetΒΆ

The following cheatsheet provides some example command lines for common cases.

Use Example Command
Run the basic example 
python -m celi_framework.main \
   --job-description=celi_framework.examples.human_eval.job_description.job_description
Pass parameters to the ToolImplmentations class 
python -m celi_framework.main \
   --job-description=celi_framework.examples.human_eval.job_description.job_description
   --tool-config='{"single_example":"HumanEval/3"}'
Use Claude models 
python -m celi_framework.main \
   --job-description=celi_framework.examples.human_eval.job_description.job_description
   --anthropic-api-key=
   --primary-model-name=claude-3-5-sonnet-20240620
Turn off parallelization and caching for debugging 
python -m celi_framework.main \
   --job-description=celi_framework.examples.human_eval.job_description.job_description
   --no-cache --serialize
Set the size of an individual LLM response to 4096 and a total limit of 1M input and output tokens 
python -m celi_framework.main \
   --job-description=celi_framework.examples.human_eval.job_description.job_description
   --max-tokens=4096 --token-budget=1000000