support kimi 2.5/6 full param and lora - #1057
Conversation
There was a problem hiding this comment.
Code Review
This pull request introduces support for Kimi VL and Kimi K2.5 models, including specialized weight conversion logic and multimodal token expansion for training. It significantly enhances the LoRA weight synchronization mechanism by implementing IPC staging buffers for faster transfers and a chunked streaming approach to handle large adapters with SGLang. Additionally, it adds support for shared-outer grouped-expert LoRA and updates various utility functions and quantization scripts to accommodate vision-tower and projector components. Feedback was provided to improve the robustness of rollout batch processing by using strict zipping.
|
@gongyisheng - weight sync in this PR. |
|
sglang-miles: sgl-project/sglang#25141 megatron-bridge: |
yushengsu-thu
left a comment
There was a problem hiding this comment.
unblock the ci test
thanks @GeLee-Q for kimi 2.5 full param support
sglang-miles: sgl-project/sglang#25141
megatron-bridge: radixark/Megatron-Bridge#3