You can deploy this model and start earning money today!

RichardErkhov/
maldv_-_winter-garden-7b-beta-gguf

Name: maldv_-_winter-garden-7b-beta-gguf
Availability: InStock

Views

No views yet

Quantization made by Richard Erkhov.

Github

Discord

Request more models

winter-garden-7b-beta - GGUF

Model creator: https://huggingface.co/maldv/
Original model: https://huggingface.co/maldv/winter-garden-7b-beta/

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

Original model description:

license: cc-by-nc-4.0 tags:

merge
conversational
multi-task pipeline_tag: text-generation

Winter Garden 7B - β - "The Writer"

It was mentioned that we are in the open ai dark winter; so I thought I would make myself a nice winter garden.

An experiment

I've merged four partitions successfully in the past, so lets go for 9! I started with:

Mistral-7B-v0.1

and merged in

ZySec-7B-v1
LemonadeRP-4.5.3
dpo-binarized-NeutrixOmnibe-7B
Multi-Verse-RP-7B
AlphaMonarch-7B
opus-v1.2-7b
Kunoichi-DPO-v2-7B
Noromaid-7B-0.4-DPO
ogno-monarch-jaskier-merge-7b-OH-PREF-DPO-v2

9-partition merge

All of the layers were partitioned in to 9 random bins. Alternating models were slerped at [0...1], and [1...0] gradients; except attention, which was slerped at 0.03.

This means that the model is still predominantly ordered around base mistral - including half of the input and output layers, and 28% of attention.

Other

Includes fast tokenizer.

Chat Template

There is currently a chat template for conversational turns. It may or may not be good for this model without more tuning. Initial tests show this model is a real talker and writer.

If you prompt with [WP] <prompt>\n\n it will take right off.

Scores

Metric	Score

RichardErkhov/maldv_-_winter-garden-7b-beta-gguf