Mistral launches new services, SDK to allow customers to fine-tune their models

French AI startup Mistral is introducing new AI model customization options, including paid plans, to allow developers – and businesses – to fine-tune their generative models for specific use cases.

The first is self-service. Mistral has released a software development kit (SDK), Mistral-Finetuneto fit your models on workstations, servers, and small data center nodes.

In the readme of the SDK’s GitHub repository, Mistral notes that the SDK is optimized for multi-GPU configurations, but can be scaled down to a single Nvidia A100 or H100 GPU to fit smaller models like Mistral 7B. Fine-tuning a dataset like UltraChat, a collection of 1.4 million dialogues with OpenAI’s ChatGPT, takes about half an hour using Mistral-Finetune on eight H100s, says Mistral.

For developers and enterprises that prefer a more managed solution, there are Mistral’s recently launched fine-tuning services, available through the company’s API. Compatible with two Mistral models for now, the Mistral Small and the aforementioned Mistral 7B, Mistral states that the fine-tuning services will gain support for more of its models in the coming weeks.

Lastly, Mistral is launching custom training services – currently only available to select customers – to tune any Mistral model for an organization’s applications using its data. “This approach allows the creation of highly specialized models optimized for your specific domain”, explains the company in an official post blog.

Mistral, which my colleague Ingrid Lunden recently reported is seeking to raise around $600 million at a $6 billion valuation from investors including DST, General Catalyst and Lightspeed Venture Partners, is undoubtedly looking to grow revenue as it faces considerable – and growing – competition in the space of generative AI.

Since Mistral revealed its first generative model in September 2023, several others have been released, including a code generation modeland launched paid APIs. But it did not reveal how many users it has – nor what its revenues are like.