search

crazyyanchao / llmcompiler

LLMCompiler is an Agent Architecture designed to speed up the execution of agent tasks by executing them quickly in the DAG. It also saves the cost of redundant token use by reducing the number of calls to the LLM.
Apache License 2.0
18 stars 3 forks source link

Trying with previous version such asGPT-3.5 Turbo #9

Open nooralahzadeh opened 6 days ago

nooralahzadeh commented 6 days ago

Did you try your code with GPT-3.5 Turbo? It seems cannot provide a proper plan and fails in execution!

crazyyanchao commented 5 days ago

@nooralahzadeh When using a less powerful model, you may need to provide more few-shot or better tool descriptions. If the results are still poor, it is recommended to switch to a more capable model.