Modelpublic

RichardErkhov/Nitral-Archive_-_Eris-Floramix-7b-gguf

sourceHugging Faceupdated 2y agoView on Hugging Face

0likes238downloads

Model Card

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Eris-Floramix-7b - GGUF

—Model creator: https://huggingface.co/Nitral-Archive/
—Original model: https://huggingface.co/Nitral-Archive/Eris-Floramix-7b/

Name	Quant method	Size
Eris-Floramix-7b.Q2_K.gguf	Q2_K	2.53GB
Eris-Floramix-7b.IQ3_XS.gguf	IQ3_XS	2.81GB
Eris-Floramix-7b.IQ3_S.gguf	IQ3_S	2.96GB
Eris-Floramix-7b.Q3_K_S.gguf	Q3KS	2.95GB
Eris-Floramix-7b.IQ3_M.gguf	IQ3_M	3.06GB
Eris-Floramix-7b.Q3_K.gguf	Q3_K	3.28GB
Eris-Floramix-7b.Q3_K_M.gguf	Q3KM	3.28GB
Eris-Floramix-7b.Q3_K_L.gguf	Q3KL	3.56GB
Eris-Floramix-7b.IQ4_XS.gguf	IQ4_XS	3.67GB
Eris-Floramix-7b.Q4_0.gguf	Q4_0	3.83GB
Eris-Floramix-7b.IQ4_NL.gguf	IQ4_NL	3.87GB
Eris-Floramix-7b.Q4_K_S.gguf	Q4KS	3.86GB
Eris-Floramix-7b.Q4_K.gguf	Q4_K	4.07GB
Eris-Floramix-7b.Q4_K_M.gguf	Q4KM	4.07GB
Eris-Floramix-7b.Q4_1.gguf	Q4_1	4.24GB
Eris-Floramix-7b.Q5_0.gguf	Q5_0	4.65GB
Eris-Floramix-7b.Q5_K_S.gguf	Q5KS	4.65GB
Eris-Floramix-7b.Q5_K.gguf	Q5_K	4.78GB
Eris-Floramix-7b.Q5_K_M.gguf	Q5KM	4.78GB
Eris-Floramix-7b.Q5_1.gguf	Q5_1	5.07GB
Eris-Floramix-7b.Q6_K.gguf	Q6_K	5.53GB
Eris-Floramix-7b.Q8_0.gguf	Q8_0	7.17GB

Original model description: --- base_model:

—ChaoticNeutrals/ErisRemixDPO_7B
—ResplendentAI/FloraDPO7B library_name: transformers tags:
—mergekit
—merge license: other ---

image/jpeg

This model was merged using the SLERP merge method.

Models Merged

The following models were included in the merge:

Configuration

The following YAML configuration was used to produce this model:

yaml

slices:
  - sources:
      - model: ChaoticNeutrals/Eris_Remix_DPO_7B
        layer_range: [0, 32]
      - model: ResplendentAI/Flora_DPO_7B
        layer_range: [0, 32]
merge_method: slerp
base_model: ChaoticNeutrals/Eris_Remix_DPO_7B
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
dtype: bfloat16