Summer news from Mistral | S03 E23 | Build On Generative AI
Let's discover the latest news and model releases from Mistral !
Tiffany Souterre
Amazon Employee
Published Aug 8, 2024
In this episode we have the chance to welcome Mistral for a reviews of their latests news. The amazing Harizo Rajaona goes into the details of the new models that have been released including:
- Codestral Mamba: Unlike transformer model, it offer the advantage of linear time inference and the theoretical ability to model sequences of infinite length.
- Mathstral: Offers advanced math reasoning.
- Mistral NeMo: Offers a large context window of up to 128k tokens.
- Large Enough (aka Mistral Large 2): Mistral Large 2 is designed for single-node inference with long-context applications in mind – its size of 123 billion parameters allows it to run at large throughput on a single node.
Check out the recording here:
Loading...
Feedback:
Did you like this episode ? What are the other topics you would like us to talk about ? Let us know HERE.
Did you like this episode ? What are the other topics you would like us to talk about ? Let us know HERE.
Shared links:
- Blog post of Mistral Large 2: https://aws.amazon.com/blogs/machine-learning/mistral-large-2-is-now-available-in-amazon-bedrock/
- Blog post of Titan image generator: https://aws.amazon.com/blogs/aws/amazon-titan-image-generator-v2-is-now-available-in-amazon-bedrock/
- Darko's gist code: https://gist.github.com/darko-mesaros/c6f7f34cd97f9c16c7e45482d3d4c26d
- Blog post for Mathstral: https://mistral.ai/news/mathstral/
- Blog post for Codestral Mamba: https://mistral.ai/news/codestral-mamba/
- Blog post for Mistral NeMo: https://mistral.ai/news/mistral-nemo/
- Blog post for Large Enough: https://mistral.ai/news/mistral-large-2407/
- Blog post for build, tweak, repeat: https://mistral.ai/news/build-tweak-repeat/
Reach out to the hosts:
- Tiffany Souterre: linkedin.com/in/tiffanysouterre
- Harizo Rajaona: linkedin.com/in/harizorajaona/
Any opinions in this post are those of the individual author and may not reflect the opinions of AWS.