search

QwenLM / Qwen2

Qwen2 is the large language model series developed by Qwen team, Alibaba Cloud.
7.45k stars 458 forks source link

微调7B模型function calling的能力大概需要标注多少条数据 #817

Closed lifucong closed 3 days ago

lifucong commented 1 month ago

微调7B模型function calling的能力大概需要标注多少条数据

yangjianxin1 commented 1 month ago

可以尝试先标注两三千条数据，训练完后，模型会有一定的function call的能力。也可以先使用一些开源的function call数据做一些尝试，例如modelscope-agent开源的数据。

yangjianxin1 commented 1 month ago

MSAgent-Bench数据：https://modelscope.cn/datasets/iic/MSAgent-Bench

Linuxstyle commented 1 month ago

微调时，需要多少GPU呢，我目前用了4块3090微调，提示我显存不足。

github-actions[bot] commented 1 week ago

This issue has been automatically marked as inactive due to lack of recent activity. Should you believe it remains unresolved and warrants attention, kindly leave a comment on this thread.