What toolchain are you going to use with the local model? I agree that’s a Strong model, but it’s so slow for be with large contexts I’ve stopped using it for coding.
I have my own agent harness, and the inference backend is vLLM.
I have my own agent harness, and the inference backend is vLLM.