AI & ML interests

I like breaking things and playing the mad scientist. Did you know we still haven't completely untapped the full potential of SD 1.5 yet?

Recent Activity

new activity about 6 hours ago

SL-AI/GRaPE-2.5-Helios-GGUF:This model is insane.

new activity 10 days ago

google/gemma-4-26B-A4B-it:Gemma 4.2 26B A4B?

new activity 14 days ago

OmniSVG/OmniSVG:Adding `safetensors` variant of this model

View all activity

Organizations

None yet

New activity in SL-AI/GRaPE-2.5-Helios-GGUF about 6 hours ago

This model is insane.

#1 opened about 6 hours ago by

Koitenshin

New activity in google/gemma-4-26B-A4B-it 10 days ago

Gemma 4.2 26B A4B?

#58 opened 29 days ago by

kozusenok

New activity in OmniSVG/OmniSVG 14 days ago

Adding `safetensors` variant of this model

🔥 4

#1 opened about 1 year ago by

SFconvertbot

New activity in Qwen/Qwen3.6-35B-A3B 15 days ago

Prescence penalty 1.5

👀👍 4

#5 opened 5 months ago by

Sliderpro93

New activity in SC117/Ornith-1.5-35B-A3B-Heretic-MTP-APEX-GGUF 17 days ago

The high-quality is actually pretty bad

#3 opened 25 days ago by

I-Am-Sam

New activity in aixonlab/Eurydice-24b-v3 18 days ago

Extremely slow, but amazing

#2 opened 18 days ago by

Koitenshin

New activity in Qwen/Qwen3.8-27B 26 days ago

Le Cirque du Raisonnement: 27B Parameters, 262K Tokens of Hot Air, and a Packaging Bug That Silently Truncates Every Prompt)

❤️➕ 5

#144 opened 28 days ago by

AdrienneNoctis

New activity in JetBrains/Mellum2-12B-A2.5B-Thinking 2 months ago

not particularly smart

👍 14

#7 opened 3 months ago by

toriset

New activity in empero-ai/Qwythos-9B-Claude-Mythos-5-1M-GGUF 3 months ago

Thanks - it works well in opencode

🤝❤️ 7

#1 opened 3 months ago by

Milor123

New activity in mistralai/Ministral-3-14B-Instruct-2512 4 months ago

The model suffers from ADHD

👍 1

#18 opened 9 months ago by

Candala

New activity in deucebucket/Gemma-4-26B-A4B-it-Cerebellum-v6-GGUF 4 months ago

What about uncensored/abliterated version?

❤️ 2

#2 opened 4 months ago by

tima2431

New activity in google/gemma-4-E4B-it 5 months ago

Image Dimension Test Results

#27 opened 5 months ago by

Koitenshin

New activity in Qwen/Qwen3-30B-A3B-Instruct-2507 12 months ago

An Improvement, But Q3 30b Still Has Very Little General Knowledge

❤️👍 3

#2 opened about 1 year ago by

phil111

New activity in DontPlanToEnd/UGI-Leaderboard 12 months ago

Any news on the awesome new benchmarks?

🚀👍 3

#373 opened about 1 year ago by

VOIDER

New activity in Orenguteng/Llama-3.1-8B-Lexi-Uncensored-V2 12 months ago

Not Uncensored

#1 opened about 2 years ago by

SicariusSicariiStuff

New activity in oobabooga/accurate-gguf-vram-calculator about 1 year ago

Possible Improvements / Future Options

👍 1

#1 opened about 1 year ago by

Koitenshin

New activity in tensorblock/Salesforce_xgen-small-9B-instruct-r-GGUF about 1 year ago

Q4_K_M is Very Unstable

#1 opened about 1 year ago by

Koitenshin

New activity in prithivMLmods/Luna-f32-GGUF about 1 year ago

Breaks once it hits Context Limit

#1 opened about 1 year ago by

Koitenshin

New activity in mradermacher/Qwen3-30B-A3B-YOYO-GGUF about 1 year ago

Let there be light !

#1 opened about 1 year ago by

phoebdroid

New activity in apple/FastVLM-7B about 1 year ago

Comparison?

#5 opened about 1 year ago by

yukiarimo

Koitenshin

AI & ML interests

Recent Activity

Organizations

Koitenshin's activity

This model is insane.

Gemma 4.2 26B A4B?

Adding `safetensors` variant of this model

Prescence penalty 1.5

The high-quality is actually pretty bad

Extremely slow, but amazing

Le Cirque du Raisonnement: 27B Parameters, 262K Tokens of Hot Air, and a Packaging Bug That Silently Truncates Every Prompt)

not particularly smart

Thanks - it works well in opencode

The model suffers from ADHD

What about uncensored/abliterated version?

Image Dimension Test Results

An Improvement, But Q3 30b Still Has Very Little General Knowledge

Any news on the awesome new benchmarks?

Not Uncensored

Possible Improvements / Future Options

Q4_K_M is Very Unstable

Breaks once it hits Context Limit

Let there be light !

Comparison?