4 ms·
It says they can be "fine tuned," but it looks like the agents are all using the same model with different system prompts? This would be more intriguing if they
by mortsnort 1y ago
It says they can be "fine tuned," but it looks like the agents are all using the same model with different system prompts? This would be more intriguing if they trained a debugger model from the ground up that could be used for the debugger agent. I suspect we'll get there eventually.