Fine-tuning:
Streamline from fine-tuning to serving

Optimize generative AI performance by customizing models 

Achieve your business goals more effectively by fine-tuning pre-trained models with your enterprise's data, optimizing performance and saving both time and resources.

Parameter Efficient

Parameter-Efficient
Fine-Tuning

Fine-tune models efficiently by updating only relevant parameters, preserving performance while saving computational resources, making it a top choice for model refinement.

Faster Training

Instead of updating all model parameters, this method focuses on updating only a subset of the pre-trained model's parameters. Fewer updates means less time and resources, saving costs.

Maintains Accuracy

Preserves pre-trained model's valuable knowledge for seamless adaptation to new tasks. Despite using fewer resources, parameter-efficient fine-tuning maintains accuracy levels.

Effortlessly fine-tune and deploy LLMs

Fine-tune and deploy

Friendli Suite not only enables you to easily fine-tune your models but also streamlines the deployment process. You can run your fine-tuned models in your GPU environment with Friendli Container or on Friendli dedicated endpoints with just a few clicks. This seamless process ensures high performance and cost-efficiency for your operations.


SUPPORTED MODELS

Fine-tune open-source LLMs
on Friendli Dedicated Endpoints

Try now

LLAMA 2 7B HF

LLAMA 2 13B HF

META LLAMA 3 8B

META LLAMA 3 8B INSTRUCT

HOW TO USE

Serve a fine-tuned model
in one-click with Friendli

01

Friendli Container

Serve LLMs in your private environment

Learn more

02

Friendli Dedicated Endpoints

Build and serve custom LLMs on autopilot with Friendli Dedicated Endpoints

Learn more