The fine-tuning benchmarking blueprint streamlines infrastructure benchmarking for fine-tuning using the MLCommons methodology. It fine-tunes a quantized Llama-2-70B model and a standard dataset.
Once complete, benchmarking results, such as training time and resource utilization, are available in MLFlow and Grafana for easy tracking. This blueprint enables data-driven infrastructure decisions for your fine-tuning jobs.
| Feature Showcase | Title | Description | Blueprint File |
|---|---|---|---|
| Benchmark LoRA fine-tuning performance using MLCommons methodology with quantized large language models | LoRA fine-tuning of quantitized Llama-2-70B model on A100 node using MLCommons methodology | Deploys LoRA fine-tuning of quantitized Llama-2-70B model on A100 node using MLCommons methodology on BM.GPU.A100.8 with 8 GPU(s). | mlcommons_lora_finetune_nvidia_sample_recipe.json |