🍀 touch grass

Nina Beerbower

nbeerbower

142 21 997

yamatazen's profile picture

McG-221's profile picture

jaepil's profile picture

https://ko-fi.com/birubawa

nsbeerbower
nbeerbower
nbeerbower

AI & ML interests

QLoRA finetuning and merging LLMs for fun. Also learning about diffusion and VLMs! (Out of space, won't be posting here anymore)

Recent Activity

liked a model about 9 hours ago

MiniMaxAI/MiniMax-H3

updated a dataset about 11 hours ago

schneewolflabs/Vernunft-SFT

published a dataset about 11 hours ago

schneewolflabs/Vernunft-SFT

View all activity

Organizations

nbeerbower 's collections 12

Qwen3.6-27B LoRAs

nbeerbower/Qwen3.6-27B-delegation-LoRA

Updated 8 days ago • 23 • 1
nbeerbower/Qwen3.6-27B-egirl-LoRA

Image-Text-to-Text • Updated 6 days ago
schneewolflabs/Qwen3.6-27B-Stimme-LoRA

Image-Text-to-Text • Updated 4 days ago
nbeerbower/Schierling-Qwen3.6-27B-LoRA

Updated 6 days ago • 14

Hemlock 0.1

First pass at training Qwen-Coder on Hemlock

nbeerbower/hemlock-sft-v0.1

Viewer • Updated Dec 17, 2025 • 867 • 28
nbeerbower/Hemlock-Qwen3-Coder-REAP-25B-A3B

Text Generation • 25B • Updated Dec 18, 2025 • 6 • 1
nbeerbower/Hemlock-Qwen2.5-Coder-32B

Text Generation • 33B • Updated Jan 3 • 10 • 1
nbeerbower/Hemlock-Qwen2.5-Coder-14B

Text Generation • 15B • Updated Dec 30, 2025 • 7 • 1

DPO

Various useful datasets with preference optimization

jondurbin/gutenberg-dpo-v0.1

Viewer • Updated Jan 12, 2024 • 918 • 968 • 167
nbeerbower/gutenberg2-dpo

Viewer • Updated Nov 16, 2024 • 293 • 73 • 24
jondurbin/truthy-dpo-v0.1

Viewer • Updated Jan 11, 2024 • 1.02k • 736 • 140
kyujinpy/orca_math_dpo

Viewer • Updated Apr 12, 2024 • 15.3k • 109 • 20

Dumplings

Qwen2.5 finetunes aiming for decensorship and improved english prose

nbeerbower/Dumpling-Qwen2.5-32B

Text Generation • 33B • Updated Jan 24, 2025 • 15 • • 11
nbeerbower/Dumpling-Qwen2.5-14B

Text Generation • 15B • Updated Feb 14, 2025 • 13 • • 4
nbeerbower/Dumpling-Qwen2.5-1.5B-v2

Text Generation • 2B • Updated Feb 4, 2025 • 10 • 2
nbeerbower/Dumpling-Qwen2.5-1.5B

Text Generation • 2B • Updated Feb 1, 2025 • 14 • 1

bruphin

Series of merge experiments attempting to make a small uncensored ChatML model based initially on ehartford/dolphin and rwitz/go-bruins-v2 mistral-7B

nbeerbower/bruphin-lambda

Text Generation • 7B • Updated Apr 3, 2024 • 54
nbeerbower/bruphin-kappa

Text Generation • 7B • Updated Mar 24, 2024 • 47 • 1
nbeerbower/bruphin-iota

Text Generation • 7B • Updated Mar 24, 2024 • 49 • 1
nbeerbower/bruphin-theta

Text Generation • 7B • Updated Mar 10, 2024 • 50

llama 3 experiments

Merge and finetune experiments with llama 3, aiming for decensorship and benchmark performance.

nbeerbower/llama-3-stella-truthy-dpo-8B

Text Generation • 8B • Updated May 6, 2024 • 14 • • 1
nbeerbower/llama-3-stella-8B

Text Generation • 8B • Updated May 6, 2024 • 18 • • 2
nbeerbower/llama-3-bible-dpo-8B

Text Generation • 8B • Updated May 6, 2024 • 7 • 4
nbeerbower/llama-3-wissenschaft-8B-v2

Text Generation • 8B • Updated May 5, 2024 • 10 • • 1

Grimoire Test

Testing various training methods in grimoire, a new training library

nbeerbower/Huihui-Qwen3.5-9B-abliterated-Grimoire-ORPO

Text Generation • 9B • Updated Mar 16 • 17 • 1
nbeerbower/Huihui-Qwen3.5-9B-abliterated-Grimoire-SFT

Text Generation • 9B • Updated Mar 18 • 12 • 1
nbeerbower/Huihui-Qwen3.5-9B-abliterated-Grimoire-DPO

Text Generation • 9B • Updated Mar 24 • 14
nbeerbower/Huihui-Qwen3.5-9B-abliterated-Grimoire-SimPO

Text Generation • 9B • Updated Mar 24 • 8

Qwen Image Edit LoRAs

QIE experiments!

nbeerbower/FIXBODY-r128

Updated Oct 2, 2025 • 1
nbeerbower/FIXBODYr128-QwenImageEdit2509

Updated Oct 9, 2025 • 2
nbeerbower/FIXBODY-r64-Qwen-Image-Edit

Updated Sep 26, 2025

Qwen3

my qwen3 tunes

nbeerbower/Yanfei-v2-Qwen3-32B

Text Generation • 33B • Updated Jun 11, 2025 • 11 • 1
nbeerbower/Vitus-Qwen3-14B

15B • Updated Jun 9, 2025 • 10 • 8
nbeerbower/Xiaolong-Qwen3-14B

Text Generation • 15B • Updated May 9, 2025 • 10 • 9
nbeerbower/Xiaolong-Qwen3-8B

Text Generation • 8B • Updated May 8, 2025 • 14 • • 4

abliteration loras

Extracted adapters for removing censorship in models

nbeerbower/Qwen2.5-32B-abliterated-LORA

Updated Apr 28, 2025 • 2
nbeerbower/Mistral-Nemo-12B-abliterated-LORA

Updated Oct 19, 2024 • 4
mlabonne/Llama-3-70B-Instruct-abliterated-LORA

Updated Aug 4, 2024 • 7

flammen

Merge & finetune experiments aiming for strong character roleplay, creative writing, and overall good performance.

flammenai/flammen32-mistral-7B

Text Generation • 7B • Updated Jun 1, 2024 • 5 • 1
flammenai/flammen31-mistral-7B

Text Generation • 7B • Updated Jun 1, 2024 • 6 • 2
flammenai/flammen30-mistral-7B

Text Generation • 7B • Updated May 29, 2024 • 8 • 1
flammenai/flammen29-mistral-7B

Text Generation • 7B • Updated May 27, 2024 • 11 • 1

Nemo

Mistral Nemo 12B finetunes and merges

nbeerbower/mistral-nemo-bophades-12B

Text Generation • 12B • Updated Sep 6, 2024 • 16 • • 9
nbeerbower/Lyra-Gutenberg-mistral-nemo-12B

Text Generation • 12B • Updated Sep 13, 2024 • 15 • • 20
nbeerbower/mistral-nemo-gutenberg-12B-v4

Text Generation • 12B • Updated Sep 6, 2024 • 17 • • 22
nbeerbower/mistral-nemo-gutenberg-12B-v3

Text Generation • 12B • Updated Sep 4, 2024 • 11 • 11