
Matt Sharp & Chris Brousseau - Writing "LLMs in Production" (the midway edition)
The Joe Reis Show
00:00
Navigating Data Contamination Risks and Model Selection
Exploring methods such as retrieval augmented generation to address contamination risks in data, highlighting the significance of selecting appropriate base models, comparing varied datasets, and stressing the importance of data cleaning in the fine-tuning process.
Transcript
Play full episode