Thanks, great model.
#2
by
ThijsL202
- opened
I like the model, it's very good for roleplay and I have a better time getting it to work overall in Silytavern compared to EVA-Qwen2.5-32B-v0.2-Q6_K (which gave me a lot of headaches.)
I'm using Q6_K_L with context size set to 24576
hardware
RTX 3090 (24GB) + RTX 4070 ti super (16GB)