Introduction to Tracking with FinetuneDB

Step 1: Sign Up and Create an Account

Step 2: Create a Workspace and Generate an API Key

Step 3: Start Tracking Data

Step 4: Create Datasets

Step 5: Finetune and Compare

FinetuneDB

Get started free

Sign in

Getting started with setting up and data tracking.

Quickstart

FinetuneDB streamlines AI fine-tuning by providing an end-to-end platform to create and manage datasets, fine-tune and serve LLMs that are significantly cheaper and better performing. Use FinetuneDB to track production data, leverage collaborative tools for dataset creation, test different prompt and model versions, and create fine-tuned models.

What is FinetuneDB

Our platform is designed around a cyclical process that simplifies LLM fine-tuning and enables continuous model improvements via monitoring and output evaluations.

Overview

Fine-tune, compare, and serve open-source, proprietary, and custom models.

Base Models

Flexible plans for individuals and teams.

Pricing

Switch from a proprietary model like GPT-4 Turbo to an open-source model using current production logs.

Switch to Open-Source LLM

Get together with domain experts to create and collaborate on fine-tuning datasets using the FinetuneDB Dataset Manager.

Collaborate on Datasets

Create and manage custom fine-tuning datasets all in one place. Upload datasets, use production data, or create datasets from scratch with domain experts.

How it works

Upload datasets in JSONL format using a simple drag-and-drop interface. Multiple datasets can be uploaded into the same project for easy management and editing.

Upload Datasets

Easily edit fine-tuning datasets with our user-friendly graphical interface. Keep track of changes and experiment confidently with version histories, making sure you can always revert or compare different dataset versions.

Version Controlled Editing

Simplify complex function applications with a visual interface, making it easier to apply transformations and data enrichment without code.

Visual Function Calling Editor

Our system automatically validates training data integrity.

Automatic Dataset Validation

Easily export your datasets in JSONL format.

Export Datasets

The Log Viewer enables tracking, monitoring, analysis of production logs, and building fine-tuning datasets from production data.

Logging Requests

View and filter logs based on various criteria.

View and Filter Requests

In FinetuneDB, you can search and filter past LLM requests to inspect responses and build a training dataset.

Training on Logs

Fine-tune open-source and proprietary LLMs straight from the FinetuneDB platform using production logs or uploaded datasets for a streamlined fine-tuning process.

Fine-Tuning

Choose your fine-tuned models, view model details, and access API endpoints to integrate your models into production.

Model Serving

Evaluations in FinetuneDB are designed to measure and improve the quality of model outputs through collaborative and automated processes.

Evaluations

The Studio is a dedicated space for developing and optimizing AI model prompts and comparing different model versions.

Prompt Studio

Workspaces are a centralized hub for managing and collaborating on multiple projects. Each project can be integrated with various model providers and supports shared resources for seamless team collaboration.

Welcome

Getting Started

Use Cases

Features

Logging data

Quickstart

Introduction to Tracking with FinetuneDB

Step 2: Create a Workspace and Generate an API Key

Step 3: Start Tracking Data

Integration

Step 4: Create Datasets

Step 5: Finetune and Compare

Welcome

Getting Started

Use Cases

Features

Logging data

​Introduction to Tracking with FinetuneDB

​Step 1: Sign Up and Create an Account

​Step 2: Create a Workspace and Generate an API Key

​Step 3: Start Tracking Data

Integration

​Step 4: Create Datasets

​Step 5: Finetune and Compare

Introduction to Tracking with FinetuneDB

Step 1: Sign Up and Create an Account

Step 2: Create a Workspace and Generate an API Key

Step 3: Start Tracking Data

Step 4: Create Datasets

Step 5: Finetune and Compare