Repository Issue Activity (beta)

#454 Base Model generation time increases when passed through the MergeKit

#6 How to merge models of two different sizes and architectures?

#697 Tiny bug in mergekit/moe/deepseek.py 1 line change.

#669 Qwen3.5 27B

#681 Pydantic error when merging models

#494 how to merge gguf quntizated models?

#687 Why I can't merge Deepseek R1 with Deepseek V3.1?

#524 NameError: name 'LLM' is not defined when using vllm in lm_eval

#680 Feature Request: Support strict parameter filtering (exclude non-matched parameters from merging)

#676 Merging multiple different architectures using MergeKit.

#675 📋 Documentation Enhancement Suggestion

#671 Does Mergekit support Qwen3 models?

#654 Gated MOE - 4 bit bug

#672 It would be greatly appreciated if you could support the Qwen3Next 80A3B model.

#659 When will the Qwen3-VL series models be supported?

#667 max_position_embeddings

#561 Multislerp is not documented

#627 Will GLM4MoE be supported?

#626 Qwen3 MoE Error

#660 TypeError in SLERP merging with ApertusForCausalLM

#395 Error at MoE Qwen 1.5B

#68 Has anyone tried modal.com for merging models ?

#653 KeyError: ModelReference Problem

#652 Clarification on gpt-oss support

#649 Collect all weights in a single task