[7/7][multi-lora]: driver and example - fully async driver, typer launcher, adapter configs, service smoke client#1748
Conversation
…ml config, shared helpers, CLI flags and validation
…d, control-plane HTTP API, named Ray actor
…schedules, slot lifecycle, batch routing in get_batch
…ert push to engines, reconcile/train/save hooks
… per-adapter rewards, prefill logprob grouping
…apter buffers, round-robin collection, batch metadata
…cher, adapter configs, service smoke client
![gemini-code-assist[bot]](https://avatars.githubusercontent.com/in/956858?s=60&v=4){kind=small}
There was a problem hiding this comment.
Code Review
This pull request introduces a comprehensive Multi-LoRA training system, enabling concurrent training of multiple LoRA adapters on a shared base model. Key additions include a controller Ray actor for adapter management, a fully-async rollout mechanism with per-adapter buffering, and decoupled per-slot Adam optimizers. My review identified a redundant configuration check in the bridge helpers and recommended using isinstance() for type validation in the backend registry to follow Python best practices.
Important
The consumer version of Gemini Code Assist on GitHub is being sunset. Starting June 18, 2026, new organization installations will be blocked, and all code review activity will officially cease on July 17, 2026.
For more details on the timeline and next steps, please review the Help Documentation.
yushengsu-thu
left a comment
yushengsu-thu
left a comment
There was a problem hiding this comment.
Reviewed extensively on the umbrella PR #1638 (2xH200 E2E validated); with this final layer the merged tree is byte-identical to the E2E-validated head.
2 participants
#1638