axeron-44
This is a forge of pre-trained language models created using forgelm.
Forge Details
Forge Method
This model was forged using the SLERP forge method.
Models Forged
The following models were included in the forge:
Configuration
The following YAML configuration was used to produce this model:
base_model: Qwen/Qwen2.5-3B
dtype: bfloat16
forge_method: slerp
modules:
default:
slices:
- sources:
- layer_range: [0, 36]
model: Qwen/Qwen2.5-3B
- layer_range: [0, 36]
model: Qwen/Qwen2.5-3B-Instruct
parameters:
t: 0.5
tokenizer:
source: base