Birbal: An efficient 7B instruct-model fine-tuned with curated datasets
This work addresses accessibility and reproducibility issues in LLM deployment for researchers and practitioners, though it is incremental as it builds on existing fine-tuning methods.
The paper tackles the high costs and lack of transparency in LLMOps by introducing Birbal, a Mistral-7B model fine-tuned on a single GPU, which achieved a 35% performance improvement over the second-best submission in the LLM Efficiency Challenge.
LLMOps incur significant costs due to hardware requirements, hindering their widespread accessibility. Additionally, a lack of transparency in model training methods and data contributes to the majority of models being non-reproducible. To tackle these challenges, the LLM Efficiency Challenge was introduced at NeurIPS Workshop, aiming to adapt foundation models on a diverse set of tasks via fine-tuning on a single GPU (RTX 4090 or A100 with 40GB) within a 24-hour timeframe. In this system description paper, we introduce Birbal, our Mistral-7B based winning model, fine-tuned on a single RTX 4090 for 16 hours. Birbal's success lies in curating high-quality instructions covering diverse tasks, resulting in a 35% performance improvement over second-best Qwen-14B based submission.