--- license: apache-2.0 tags: - merge - mergekit - lazymergekit - google/gemma-3-4b-it - Qwen/Qwen2.5-Coder-7B --- # CodeGem CodeGem is a merge of the following models using [mergekit](https://github.com/cg123/mergekit): * [google/gemma-3-4b-it](https://huggingface.co/google/gemma-3-4b-it) * [Qwen/Qwen2.5-Coder-7B](https://huggingface.co/Qwen/Qwen2.5-Coder-7B) ## 🧩 Configuration ```yaml slices: - sources: - model: google/gemma-3-4b-it layer_range: [0, 32] - model: Qwen/Qwen2.5-Coder-7B layer_range: [0, 32] merge_method: slerp base_model: google/gemma-3-4b-it parameters: t: - filter: self_attn value: [0, 0.5, 0.3, 0.7, 1] - filter: mlp value: [1, 0.5, 0.7, 0.3, 0] - value: 0.5 dtype: bfloat16 ```