VITA Lab ← PRODUCTS
PRODUCT / AI.VITA START A PROJECT ↗
BETA · AI · PRIVATE LLM

ai.vita BETA

AI trained on your data, running where you choose

ai.vita fine-tunes open-source language models on your organisation's data, evaluates them and prepares them to run on a single GPU with 20 GB of memory. The finished model is deployed either on your own on-premise server or on a virtual server in your own account with the cloud of your choice. GPUs are rented only for as long as training needs them; your data and model stay with you. The product is in beta.

FIELDS
Fine-tuning (LoRA / full)EvaluationQuantizationOn-premise or cloud deployment
ADDRESS
ai.vitaactionlab.com ↗
STATUS
Beta. Sign-in and user management work in the panel; other screens are previews. GPU provider selection is in progress.
COMPANY
VİTA Teknoloji Eğitim ve Danışmanlık Ltd. Şti. (VITA Lab)
Five steps: prepare data, fine-tune, evaluate, quantize and deploy on-premise or in your own cloud account
01 / PREPARE DATA

A human-approved training dataset.

Training examples are written and approved by the organisation's experts, so the dataset carries its knowledge and language.

02 / FINE-TUNE

An open-source model, tuned on your data.

Qwen 27B is fine-tuned (LoRA or full) on an on-demand GPU server, with progress, loss and cost monitored live.

03 / EVALUATE

Before and after, on the same questions.

The fine-tuned model is compared with the base model on a held-out test set, question by question.

04 / QUANTIZE

A measured model that fits 20 GB.

The model is quantized to fit in 20 GB of memory and measured on the same questions before and after, so any quality loss is visible.

05 / DEPLOY

On-premise or in your cloud.

The finished model runs on a server with a single 20 GB GPU. You choose where: your own on-premise server, or a virtual server in your own account with the cloud of your choice. The training server is then shut down.

Two options: an on-premise server or a virtual server in your own cloud account; your data and model stay with you
COST CONTROL · BETA

Every cost shown upfront.

The hourly price is shown before a GPU server starts, idle servers shut down automatically, and every action that spends money is confirmed with its amount.

Beta status: sign-in, sessions, users and activity work in the panel; training, servers and the other screens are currently previews with sample data. GPU provider selection is in progress. Voice, search and image modules are on the roadmap.

TECHNOLOGY

Train in the cloud, infer where you choose.

MODEL
Qwen 27B (open source) · LoRA / full fine-tuning
TARGET HARDWARE
A single 20 GB GPU
DEPLOYMENT
On-premise server or cloud virtual server
INFRASTRUCTURE
On-demand GPUs · automatic shutdown
PANEL API
Cloudflare Worker · D1
INTERFACE
English and Turkish
NEXT STEP
Let's run your model where you choose.
Get in touch with VITA ↗