Skip to content
View original post on X: RadixArk· 46/100AI score46/100

RadixArk releases Miles v0.1.1 with multi-LoRA and expanded model support

AISummary

RadixArk has released Miles v0.1.1, adding multi-LoRA with Tinker API compatibility so multiple training jobs can share one base model.

The update also supports agentic training with harnesses like Claude Code and runs Harbor tasks in sandboxes including AgentENV, Daytona, E2B, and Modal.

It further reduces memory needs for training larger models on validated NVIDIA and AMD GPUs and adds stable support for Qwen3.8-Flash-Next, GLM-5.3-Flash, and Kimi-K3.

Post on XView on X
RadixArk
@radixark

Miles v0.1.1 is out.

This release expands what teams can run with Miles.

Multi-LoRA with Tinker API compatibility lets multiple training jobs share one base model without interfering with one another. For agentic training, Miles works natively with harnesses like Claude Code and runs Harbor tasks directly in sandboxes like AgentENV, Daytona, E2B, and Modal.

Training larger models requires less memory, with validated support across NVIDIA and AMD GPUs.

Model support lands into the stable version: Qwen3.8-Flash-Next, GLM-5.3-Flash, Kimi-K3, and more.

Full release notes 👇

Source: RadixArk · x.comPublished · added here

Topics

AI agentsOpen-source ecosystemDeployment & engineeringProduct updates