Summary

• Added AnyMoE
• Added the Starcoder 2 model
• Vision interactive mode
• X-LoRA and LoRA support for Gemma 2
• Optimized local loading speed
• Release using cargo dist

MSRV

cargo msrv

MSRV of mistral.rs is 1.75.0.

What's Changed

• Add a vision interactive mode by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/468
• Add support for X-LoRA/LoRA to Gemma 2 by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/499
• Add adapter activation support for Gemma 2 LoRA by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/500
• Decrease MSRV to 1.75.0 by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/501
• Fix docs with preprocessor config by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/502
• Allow for optional model field in OpenAI requests by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/504
• Support streaming for completion by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/508
• Select the best device in examples by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/507
• Handle Metal error from bf16 autodtype selection by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/509
• Use chat messages for the rust examples and show T/s by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/510
• Avoid f64 to f32 cast in phirope for dtype on metal by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/512
• Handle errors in rust examples by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/513
• AnyMoE: Build an MoE model from anything, quickly by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/476
• chore: update paths.rs by @eltociear in https://github.com/EricLBuehler/mistral.rs/pull/516
• Add AnyMoE support for vision models by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/515
• Support images in AnyMoE dataset by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/517
• Handle metal case for nonzero and bitwise ops by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/518
• Support saving and loading AnyMoE gating layer by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/519
• Implement the Starcoder 2 model architecture by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/522
• Docs for Starcoder 2 by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/524
• Starcoder2 docs by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/525
• Export loader for starcoder2 by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/526
• Support the new phi3 models by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/530
• Source the AnyMoE dataset from a structured JSON file by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/531
• Create an AnyMoE loss graph by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/532
• Add AnyMoE demo video by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/533
• Add judgment for whether model_id is a local path to speed up local model loading by @chenwanqq in https://github.com/EricLBuehler/mistral.rs/pull/523
• Update docs by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/534
• release using cargo-dist by @kranurag7 in https://github.com/EricLBuehler/mistral.rs/pull/480
• Bump version to 0.1.25 by @EricLBuehler in https://github.com/EricLBuehler/mistral.rs/pull/535

New Contributors

• @kranurag7 made their first contribution in https://github.com/EricLBuehler/mistral.rs/pull/480

Full Changelog: https://github.com/EricLBuehler/mistral.rs/compare/v0.1.24...v0.1.25

Download mistralrs-server 0.1.25