linear_Merged_gpt2-large_0.3/mergekit_config.yml

dtype: float32
merge_method: linear
models:
- model: /kaggle/working/gpt2-large
  parameters:
    weight: 0.7
- model: /kaggle/working/debias_gpt2-large
  parameters:
    weight: 0.3
parameters:
  normalize: true
tokenizer:
  source: /kaggle/working/gpt2-large
初始化项目，由ModelHub XC社区提供模型 Model: trinhkhng/linear_Merged_gpt2-large_0.3 Source: Original Platform 2026-08-26 15:08:56 +08:00			`dtype: float32`
			`merge_method: linear`
			`models:`
			`- model: /kaggle/working/gpt2-large`
			`parameters:`
			`weight: 0.7`
			`- model: /kaggle/working/debias_gpt2-large`
			`parameters:`
			`weight: 0.3`
			`parameters:`
			`normalize: true`
			`tokenizer:`
			`source: /kaggle/working/gpt2-large`