Wolfram Ravenwolf's picture

Wolfram Ravenwolf

wolfram

·

https://ko-fi.com/wolframravenwolf

AI & ML interests

Local LLMs

Recent Activity

liked a model 7 days ago

zai-org/RealVideo

new activity 23 days ago

OpenEvals/evaluation-guidebook:fix: typo

liked a Space 23 days ago

OpenEvals/evaluation-guidebook

View all activity

Organizations

New activity in OpenEvals/evaluation-guidebook 23 days ago

fix: typo

#2 opened 23 days ago by

New activity in mlx-community/Qwen3-30B-A3B-4bit 8 months ago

Jinja chat template error on lmstudio

#1 opened 8 months ago by

New activity in wolfram/Athene-V2-Chat-4.65bpw-h6-exl2 8 months ago

Improve language tag

#1 opened 8 months ago by

New activity in blog-explorers/README 12 months ago

[Support] Community Articles

#5 opened almost 2 years ago by

New activity in turboderp/Llama-3.1-70B-Instruct-exl2 over 1 year ago

The tokenizer has changed just fyi

#2 opened over 1 year ago by

New activity in mistralai/Mistral-7B-Instruct-v0.3 over 1 year ago

no system message?

#14 opened over 1 year ago by

New activity in nvidia/Llama3-ChatQA-1.5-70B over 1 year ago

Concerns regarding Prompt Format

#1 opened over 1 year ago by

New activity in wolfram/miquliz-120b-v2.0 over 1 year ago

Strange observation: model becomes super horny in ST's MinP mode

#7 opened almost 2 years ago by deleted

New activity in wolfram/miquliz-120b-v2.0-GGUF almost 2 years ago

Upload folder using huggingface_hub

#3 opened almost 2 years ago by

New activity in wolfram/miquliz-120b-v2.0 almost 2 years ago

VRAM Estimates

#3 opened almost 2 years ago by

Merge method

#4 opened almost 2 years ago by

New activity in wolfram/miqu-1-103b almost 2 years ago

Can't wait to test

#4 opened almost 2 years ago by

Kindly asking for quants

#2 opened almost 2 years ago by

New activity in wolfram/miquliz-120b-v2.0 almost 2 years ago

GPTQ / AWQ

#2 opened almost 2 years ago by

New activity in wolfram/miqu-1-120b almost 2 years ago

Guidance on GPU VRAM Split?

#3 opened almost 2 years ago by

New activity in wolfram/miqu-1-103b almost 2 years ago

Upload folder using huggingface_hub

#1 opened almost 2 years ago by

New activity in wolfram/miquliz-120b-v2.0-3.0bpw-h6-exl2 almost 2 years ago

Very interesting that miqu will give 16k context work even only first layer and last layer

#2 opened almost 2 years ago by

New activity in wolfram/miquliz-120b-v2.0-GGUF almost 2 years ago

iMatrix, IQ2_XS & IQ2_XXS

#2 opened almost 2 years ago by

New activity in wolfram/miqu-1-120b almost 2 years ago

Performance

#2 opened almost 2 years ago by

KnutJaegersberg

New activity in wolfram/miquliz-120b-v2.0-3.0bpw-h6-exl2 almost 2 years ago

VRAM requirements

#1 opened almost 2 years ago by

sophosympatheia