Working with Cline

#14
by Explify - opened

Can confirm that IQ4_XS works perfectly with Cline on 60k context windows. Running on rtx 4090 (24GB) with 24% offload. I am using Cline memory bank (https://docs.cline.bot/prompting/cline-memory-bank) feature to have a detailed documentation available for the model so that most tasks start with it reading it. It is not ideal, but fully local. Doing some uncomplicated work on Django project.
<3 qwen and Unsloth

Unsloth AI org

Thanks for the feedback! :)

@Explify have you ever tried the Q4_K_XL one?
I'm personally still getting response formatting issues when going paste ~30k
I'm on a bandwidth limited plan so I quant download the IQ4_XS version to test out

Sign up or log in to comment