Quantization made by Richard Erkhov.

MetaMath-Tulpar-7b-v2-Slerp - GGUF

Model creator: https://huggingface.co/Weyaxi/
Original model: https://huggingface.co/Weyaxi/MetaMath-Tulpar-7b-v2-Slerp/

Name	Quant method	Size
MetaMath-Tulpar-7b-v2-Slerp.Q2_K.gguf	Q2_K	2.53GB
MetaMath-Tulpar-7b-v2-Slerp.IQ3_XS.gguf	IQ3_XS	2.81GB
MetaMath-Tulpar-7b-v2-Slerp.IQ3_S.gguf	IQ3_S	2.96GB
MetaMath-Tulpar-7b-v2-Slerp.Q3_K_S.gguf	Q3_K_S	2.95GB
MetaMath-Tulpar-7b-v2-Slerp.IQ3_M.gguf	IQ3_M	3.06GB
MetaMath-Tulpar-7b-v2-Slerp.Q3_K.gguf	Q3_K	3.28GB
MetaMath-Tulpar-7b-v2-Slerp.Q3_K_M.gguf	Q3_K_M	3.28GB
MetaMath-Tulpar-7b-v2-Slerp.Q3_K_L.gguf	Q3_K_L	3.56GB
MetaMath-Tulpar-7b-v2-Slerp.IQ4_XS.gguf	IQ4_XS	3.67GB
MetaMath-Tulpar-7b-v2-Slerp.Q4_0.gguf	Q4_0	3.83GB
MetaMath-Tulpar-7b-v2-Slerp.IQ4_NL.gguf	IQ4_NL	3.87GB
MetaMath-Tulpar-7b-v2-Slerp.Q4_K_S.gguf	Q4_K_S	3.86GB
MetaMath-Tulpar-7b-v2-Slerp.Q4_K.gguf	Q4_K	4.07GB
MetaMath-Tulpar-7b-v2-Slerp.Q4_K_M.gguf	Q4_K_M	4.07GB
MetaMath-Tulpar-7b-v2-Slerp.Q4_1.gguf	Q4_1	4.24GB
MetaMath-Tulpar-7b-v2-Slerp.Q5_0.gguf	Q5_0	4.65GB
MetaMath-Tulpar-7b-v2-Slerp.Q5_K_S.gguf	Q5_K_S	4.65GB
MetaMath-Tulpar-7b-v2-Slerp.Q5_K.gguf	Q5_K	4.78GB
MetaMath-Tulpar-7b-v2-Slerp.Q5_K_M.gguf	Q5_K_M	4.78GB
MetaMath-Tulpar-7b-v2-Slerp.Q5_1.gguf	Q5_1	5.07GB
MetaMath-Tulpar-7b-v2-Slerp.Q6_K.gguf	Q6_K	5.53GB

Original model description:

license: apache-2.0 tags: - merge

MetaMath-Tulpar-7b-v2-Slerp

This is the model for MetaMath-Tulpar-7b-v2-Slerp. I used mergekit to merge models.

Yaml Config to reproduce


slices:
  - sources:
      - model: meta-math/MetaMath-Mistral-7B
        layer_range: [0, 32]
      - model: HyperbeeAI/Tulpar-7b-v2
        layer_range: [0, 32]
merge_method: slerp
base_model: mistralai/Mistral-7B-v0.1
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 # fallback for rest of tensors
dtype: float16