Tool Calling
#1
by MAsad789565 - opened
Can you fine-tune a version with improved agentic tool calling caps.
Also, do you have any idea about the compute requirements for 120b?
Sure. I am actually testing different datasets and have few checkpoints that perform quite well.
120b model can run on a device with 64GB or so memory as per my current testing with MXFP4 quant.