Quantization made by Richard Erkhov.
DARE_TIES_13B - GGUF
- Model creator: https://huggingface.co/yunconglong/
- Original model: https://huggingface.co/yunconglong/DARE_TIES_13B/
Name | Quant method | Size |
---|---|---|
DARE_TIES_13B.Q2_K.gguf | Q2_K | 4.43GB |
DARE_TIES_13B.IQ3_XS.gguf | IQ3_XS | 4.94GB |
DARE_TIES_13B.IQ3_S.gguf | IQ3_S | 5.22GB |
DARE_TIES_13B.Q3_K_S.gguf | Q3_K_S | 5.2GB |
DARE_TIES_13B.IQ3_M.gguf | IQ3_M | 5.34GB |
DARE_TIES_13B.Q3_K.gguf | Q3_K | 5.78GB |
DARE_TIES_13B.Q3_K_M.gguf | Q3_K_M | 5.78GB |
DARE_TIES_13B.Q3_K_L.gguf | Q3_K_L | 6.27GB |
DARE_TIES_13B.IQ4_XS.gguf | IQ4_XS | 6.5GB |
DARE_TIES_13B.Q4_0.gguf | Q4_0 | 6.78GB |
DARE_TIES_13B.IQ4_NL.gguf | IQ4_NL | 6.85GB |
DARE_TIES_13B.Q4_K_S.gguf | Q4_K_S | 6.84GB |
DARE_TIES_13B.Q4_K.gguf | Q4_K | 7.25GB |
DARE_TIES_13B.Q4_K_M.gguf | Q4_K_M | 7.25GB |
DARE_TIES_13B.Q4_1.gguf | Q4_1 | 7.52GB |
DARE_TIES_13B.Q5_0.gguf | Q5_0 | 8.26GB |
DARE_TIES_13B.Q5_K_S.gguf | Q5_K_S | 8.26GB |
DARE_TIES_13B.Q5_K.gguf | Q5_K | 8.51GB |
DARE_TIES_13B.Q5_K_M.gguf | Q5_K_M | 8.51GB |
DARE_TIES_13B.Q5_1.gguf | Q5_1 | 9.01GB |
DARE_TIES_13B.Q6_K.gguf | Q6_K | 9.84GB |
DARE_TIES_13B.Q8_0.gguf | Q8_0 | 12.75GB |
Original model description:
license:
- other
tags:
- dare ties
DARE_TIES_13B
This is a merge of pre-trained language models created using mergekit.
Merge Details
Merge Method
This model was merged using the DARE TIES merge method using yunconglong/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B as a base.
Models Merged
The following models were included in the merge:
- ./13B_DPO
- ./13B_MATH_DPO
Configuration
The following YAML configuration was used to produce this model:
models:
- model: yunconglong/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B
parameters:
density: 1.0
weight: 1.0
- model: ./13B_MATH_DPO
parameters:
density: 0.5
weight: [0.33, 0.4, 0.33]
- model: ./13B_DPO
parameters:
density: [0.33, 0.45, 0.66]
weight: 0.66
merge_method: dare_ties
base_model: yunconglong/Truthful_DPO_TomGrc_FusionNet_7Bx2_MoE_13B
parameters:
normalize: true
int8_mask: true
dtype: bfloat16
tokenizer_source : union