RichardErkhov/Nitral-AI_-_Eris_PrimeV4.20-Vision-32k-7B-gguf

Quantization made by Richard Erkhov.

Github

Discord

Request more models

Eris_PrimeV4.20-Vision-32k-7B - GGUF

• Model creator: https://huggingface.co/Nitral-AI/
• Original model: https://huggingface.co/Nitral-AI/Eris_PrimeV4.20-Vision-32k-7B/

Name	Quant method	Size
Eris_PrimeV4.20-Vision-32k-7B.Q2_K.gguf	Q2_K	2.53GB
Eris_PrimeV4.20-Vision-32k-7B.IQ3_XS.gguf	IQ3_XS	2.81GB
Eris_PrimeV4.20-Vision-32k-7B.IQ3_S.gguf	IQ3_S	2.96GB
Eris_PrimeV4.20-Vision-32k-7B.Q3_K_S.gguf	Q3_K_S	2.95GB
Eris_PrimeV4.20-Vision-32k-7B.IQ3_M.gguf	IQ3_M	3.06GB
Eris_PrimeV4.20-Vision-32k-7B.Q3_K.gguf	Q3_K	3.28GB
Eris_PrimeV4.20-Vision-32k-7B.Q3_K_M.gguf	Q3_K_M	3.28GB
Eris_PrimeV4.20-Vision-32k-7B.Q3_K_L.gguf	Q3_K_L	3.56GB
Eris_PrimeV4.20-Vision-32k-7B.IQ4_XS.gguf	IQ4_XS	3.67GB
Eris_PrimeV4.20-Vision-32k-7B.Q4_0.gguf	Q4_0	3.83GB
Eris_PrimeV4.20-Vision-32k-7B.IQ4_NL.gguf	IQ4_NL	3.87GB
Eris_PrimeV4.20-Vision-32k-7B.Q4_K_S.gguf	Q4_K_S	3.86GB
Eris_PrimeV4.20-Vision-32k-7B.Q4_K.gguf	Q4_K	4.07GB
Eris_PrimeV4.20-Vision-32k-7B.Q4_K_M.gguf	Q4_K_M	4.07GB
Eris_PrimeV4.20-Vision-32k-7B.Q4_1.gguf	Q4_1	4.24GB
Eris_PrimeV4.20-Vision-32k-7B.Q5_0.gguf	Q5_0	4.65GB
Eris_PrimeV4.20-Vision-32k-7B.Q5_K_S.gguf	Q5_K_S	4.65GB
Eris_PrimeV4.20-Vision-32k-7B.Q5_K.gguf	Q5_K	4.78GB
Eris_PrimeV4.20-Vision-32k-7B.Q5_K_M.gguf	Q5_K_M	4.78GB
Eris_PrimeV4.20-Vision-32k-7B.Q5_1.gguf	Q5_1	5.07GB
Eris_PrimeV4.20-Vision-32k-7B.Q6_K.gguf	Q6_K	5.53GB
Eris_PrimeV4.20-Vision-32k-7B.Q8_0.gguf	Q8_0	7.17GB

Original model description:

base_model: - l3utterfly/mistral-7b-v0.2-layla-v4 - Nitral-AI/Eris_PrimeV4-Vision-32k-7B library_name: transformers tags: - mergekit - merge license: other

Eris Prime: Version 4.20 "Blaze it" 32k Edition

"Eris decided to pick up some recreational hobbies that led her to become a bit unhinged i guess."

Quants Available Here Thanks to Lewdiculus: https://huggingface.co/Lewdiculous/Eris_PrimeV4.20-Vision-32k-7B-GGUF-IQ-Imatrix

Vision/multimodal capabilities:

If you want to use vision functionality:

• You must use the latest versions of Koboldcpp.

To use the multimodal capabilities of this model and use vision you need to load the specified mmproj file, this can be found inside this model repo.

• You can load the mmproj by using the corresponding section in the interface: