These are my own quantizations (updated almost daily).
The difference with normal quantizations is that I quantize the output and embed tensors to f16.
and the other tensors to 15_k,q6_k or q8_0.
This creates models that are little or not degraded at all and have a smaller size.
They run at about 3-6 t/sec on CPU only using llama.cpp
And obviously faster on computers with potent GPUs
Due to the optimization of the preferred model, its performance is excellent across the range of 2000-1048K. Personal usage scenarios, such as 8186, 32K, etc., are insufficient. My primary role involves managing virtual idol Twitter accounts and assisting with singing, etc. A good conversation can be very lengthy, and sometimes even 32K is not enough. Imagine having a heated chat with your virtual girlfriend, only for it to abruptly cut off—that feeling is too painful.
What kind of graphics card is needed?
For investing in a girlfriend, would a 4090 or higher be more of a question?
Meet Layla,Layla is an AI chatbot that runs offline on your device.No internet connection required.No censorship.Complete privacy.Layla Lite https://www.layla-network.ai/
To use the multimodal capabilities of this model and use vision you need to load the specified mmproj file, this can be found inside this model repo. Llava MMProj
You can load the mmproj by using the corresponding section in the interface:
image/png
Thank you:
To the authors for their hard work, which has given me more options to easily create what I want. Thank you for your efforts.
