# Smart Studio

> Smart Studio

Version: 2.0

Documentation version: 2.0. This feed contains only capabilities available in the current public Smart Studio Console. Archived and compatibility-hidden content are intentionally excluded. Current release notes: https://smartstudio.console.alibabacloud.com/guides/llms/pages/changelog/2.0.0.txt.

## Table of Contents

- [AI Dataset Preparation](https://smartstudio.console.alibabacloud.com/guides/docs/datasets/ai-dataset-preparation): Configure AI-assisted dataset preparation and review generated training and evaluation labels.
- [Create Dataset](https://smartstudio.console.alibabacloud.com/guides/docs/datasets/create-dataset): Create a training or evaluation dataset and validate its accepted file format.
- [Manage Datasets](https://smartstudio.console.alibabacloud.com/guides/docs/datasets/manage-datasets): Find, inspect, edit, download, and remove datasets from the Datasets workspace.
- [Create Deployment](https://smartstudio.console.alibabacloud.com/guides/docs/deployments/create-deployment): Select a model, allocate an activated GPU cluster, and verify a new deployment.
- [Manage Deployments](https://smartstudio.console.alibabacloud.com/guides/docs/deployments/manage-deployments): Monitor deployment status and perform supported stop, restart, edit, and delete operations.
- [AI Auto Evaluation](https://smartstudio.console.alibabacloud.com/guides/docs/evaluations/ai-auto-evaluation): Evaluate a model with custom data, configurable metrics, and an LLM judge.
- [Benchmark Evaluation](https://smartstudio.console.alibabacloud.com/guides/docs/evaluations/benchmark-evaluation): Run a model against a standard benchmark dataset and inspect the evaluation report.
- [Evaluations Overview](https://smartstudio.console.alibabacloud.com/guides/docs/evaluations/overview): Choose between benchmark and AI auto evaluation for a Smart Studio model.
- [Data Distillation](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/data-distill): Generate training data from a teacher model and monitor the distillation workflow.
- [Direct Preference Optimization for LLMs](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/dpo-llm): Configure preference data and monitor a DPO training job for a language model.
- [Direct Preference Optimization for VLMs](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/dpo-vlm): Configure preference data and monitor a DPO training job for a vision-language model.
- [Fine-Tuning Methods](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/fine-tuning-methods): Compare LoRA and full-parameter fine-tuning before configuring a training job.
- [Knowledge Distillation](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/knowledge-distill): Transfer model behavior from a teacher model to a smaller student model.
- [Fine-Tuning Overview](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/overview): Choose a Smart Studio training approach based on your model, data, and optimization goal.
- [Supervised Fine-Tuning for LLMs](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/sft-llm): Configure and monitor supervised fine-tuning for a text-based large language model.
- [Supervised Fine-Tuning for VLMs](https://smartstudio.console.alibabacloud.com/guides/docs/fine-tuning/sft-vlm): Configure and monitor supervised fine-tuning for a vision-language model.
- [API Router Platform Requirements](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/api-router-requirements): Check server, network, and storage requirements before activating the API Router Platform.
- [Authorization](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/authorization): Understand the Alibaba Cloud permissions Smart Studio requires during cluster activation.
- [BYO-GPU Requirements](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/byo-gpu-requirement): Check server, network, storage, and GPU sizing requirements before activating a Smart Studio cluster.
- [Deployment Quickstart](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/deployment-quickstart): Deploy a model from activated GPU capacity and verify that the service reaches Ready.
- [GPU Dashboard](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/gpu-dashboard): Monitor an activated Smart Studio cluster, its GPU nodes, and model-serving workloads.
- [Smart Studio Overview](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/overview): Activate your own GPU cluster, deploy models, and operate AI workloads with Smart Studio.
- [Activate Cluster](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/self-service-activation): Connect, verify, and activate your own GPU cluster for Smart Studio.
- [Use with Agent](https://smartstudio.console.alibabacloud.com/guides/docs/get-started/use-with-ai): Give an AI agent current, versioned Smart Studio documentation without granting account access.
- [Model Gallery](https://smartstudio.console.alibabacloud.com/guides/docs/model-gallery/overview): Discover Smart Studio models and choose one for deployment.
- [My Models](https://smartstudio.console.alibabacloud.com/guides/docs/my-models/overview): Upload, inspect, and deploy custom and fine-tuned model assets.
- [Billing](https://smartstudio.console.alibabacloud.com/guides/docs/usage-billing/billing): Review Smart Studio charges and trace them back to GPU and AI workloads.
- [Usage](https://smartstudio.console.alibabacloud.com/guides/docs/usage-billing/usage&costs): Review Smart Studio consumption for self-deployed models and training workloads.