Modelpublic

lejelly/deepseek-7b-math-hendrycksmath-lambda00

sourceHugging Faceupdated 8mo agoView on Hugging Face

0likes14downloads

Model Card

deepseek-7b-math-hendrycksmath-lambda00

2モデルの線形補間マージモデル。

Parameter	Value
Model A	`deepseek-ai/deepseek-math-7b-instruct`
Model B	`jahyungu/deepseek-math-7b-instruct_hendrycks_math`
λ_a	0.00
λ_b	1.00
Formula	θ* = 0.00 × θa + 1.00 × θb
dtype	torch.float16

Union tokenizer (mergekit-style): vocabularies of both models are merged.

For tokens missing from a model, the other model's embedding is used as fallback before linear interpolation.

This model was created by linearly interpolating the parameters of two models:

—Model A (deepseek-ai/deepseek-math-7b-instruct): weight = 0.00
—Model B (jahyungu/deepseek-math-7b-instruct_hendrycks_math): weight = 1.00