Quick and dirty finetune of Apriel, using an improved dataset produced by scoring all replies with a Reward model, then discarding scores <5/5.

GGUF

Model size

15B params

Architecture

llama

Hardware compatibility

4-bit

Inference Providers NEW

This model isn't deployed by any Inference Provider. 🙋 Ask for provider support

Model tree for ConicCat/Apriel-R1P-Q4_K_M-GGUF

Base model

Finetuned

Quantized

(3)

this model