Fimi 2
This is a merge of pre-trained language models created using mergekit.
If you enjoy this model, don't forget to leave feedback and a like on this model page.
Merge Details
Merge Method
This model was merged using the Reinforced Agent Merging Plus (Tensor-Local) merge method using Qwen/Qwen3.8-27B as a base.
Models Merged
The following models were included in the merge:
Configuration
The following YAML configuration was used to produce this model:
base_model: Qwen/Qwen3.8-27B
dtype: bfloat16
merge_method: ramplus_tl
modules:
default:
slices:
- sources:
- layer_range: [0, 0]
model: DavidAU/Qwen3.8-27B-Cold-Fusion-GAIN-V1.1
- layer_range: [0, 0]
model: beyoru/Kiwen1.1-27B
- layer_range: [0, 0]
model: TeichAI/Qwen3.8-27B-Fable-Distill
- layer_range: [0, 0]
model: Qwen/Qwen3.8-27B
model.language_model.layers:
slices:
- sources:
- layer_range: [0, 64]
model: DavidAU/Qwen3.8-27B-Cold-Fusion-GAIN-V1.1
- layer_range: [0, 64]
model: beyoru/Kiwen1.1-27B
- layer_range: [0, 64]
model: TeichAI/Qwen3.8-27B-Fable-Distill
- layer_range: [0, 64]
model: Qwen/Qwen3.8-27B
model.visual.blocks:
slices:
- sources:
- layer_range: [0, 27]
model: DavidAU/Qwen3.8-27B-Cold-Fusion-GAIN-V1.1
- layer_range: [0, 27]
model: beyoru/Kiwen1.1-27B
- layer_range: [0, 27]
model: TeichAI/Qwen3.8-27B-Fable-Distill
- layer_range: [0, 27]
model: Qwen/Qwen3.8-27B
out_dtype: bfloat16
parameters:
r: 0.25